Welcome to The Visionary’s Guide to the Digital Future. If the best way to predict the future is to invent it, then let’s get you ready to do just that. This podcast is created for the visionaries of today who are charged with creating the customer experiences of tomorrow. I’m your host, Paul Lima, the managing partner at the Lima Consulting Group. From Wall Street to the Pentagon and Fortune five hundreds alike, I’ve been a part of some of the largest digital transformations ever done.
We promise three things here: a strategic perspective, content geared for decision-makers, and actionable insights into the real problems that digital visionaries can apply within minutes.
The Digital Pulse
Welcome back to The Digital Pulse. If you’ve been listening to this show over the past few years, you know that with my background in finance, I like to start with the financials and the vital stats of recent acquisitions and then overlay that with my business insights to give you my perspective on what I believe the investment thesis was, what those conversations were like in the boardrooms, and provide you the immediate takeaways that you can use on the job as you’re settling into your 2026 plan this quarter.
So let’s look under the hood to see what the boards of the acquirers were really betting on and what that means for the customer experience leaders like you who listen to our show. I’m also gonna give you a new axiom for the relationship economy that I call the Law of Relationship Fidelity, inspired by BCG’s experience curve. I’m gonna let you know how Google may respond to these acquisitions, and I’ll finish off the video by sharing my opinion about who else is gonna get acquired and by whom. So these acquisitions tell a story about the megatrend CX professionals need to know about where our industry is going.
And this week, the trajectory of the entire marketing industry just shifted in a big way with two acquisitions of companies that you may not know a lot about, SEMrush and Qualify. Adobe has entered a definitive all cash agreement to acquire SEMrush for 1.9 billion, coinciding with Salesforce’s acquisition of a company called Qualified for an estimated 1.5 billion.
Let’s look at the vital stats. Adobe paid 12 dollars per share, a nearly 70 percentage premium over where SEMrush was trading just weeks ago.
Market Trends and Investor Sentiment
Ina market where SaaS valuations are being crushed, and let’s face it, Adobe shares are down by about 50% from their all time high of 688 back in 2021. Or if we’re just looking at 2025, they’re down about 36%.
So why would they pay double the sticker price? Well, investor sentiment seems to agree that, quote, prompting is the new Photoshop. And perhaps in Salesforce’s case, Agentic AI is the new SDR. Creatives and executives who manage them, and salespersons know that those little sound bites aren’t entirely true.
But that didn’t stop investors from punishing Adobe’s and Salesforce’s stock price. With each new Sora and Midjourney release or enhancement, it seems to chip away at the Adobe stock price. But the pick and shovel companies, say in NVIDIA and AMD, and the data lords like Google and Meta, have hit all time highs this year. But application layer companies like Adobe and Salesforce and HubSpot, they aren’t rising with the tide of AI.
Like I said, Adobe is down 38%, Salesforce is down 31%, and HubSpot down 51% over the last 12 months. They wanna shift their business models from selling subscriptions. They are SaaS companies after all, software as a service, to a company that has a defensible, unique set of data that can power what I’m calling the relationship economy. In other words, they wanna put high octane data in the hands of Agentic AI and sell value and potentially have customers pay for compute based on the value that they can create for them.
I’ve been saying since episode 1 that data is really the most valuable corporate asset in determining a sustainable competitive advantage for most firms. Those who are investing in collecting their own customer data are best positioned to win in the future. Let’s take a look at the vitals on the Salesforce acquisition. They just spent an estimated 1.5
billion to acquire Qualified, a company that claims to kill the contact us form. So why should you care? Well, it’s because both companies are eager to help you improve your customer relationships, and they’re going about it by predicting identity and intent. Adobe and Salesforce are concerned about generative AI and Agentic AI.
In the case of Adobe, Gen AI can create content without the need of tools for some, but not all use cases. And in the case with Salesforce, Agentic AI can do a lot of the tasks related to marketing and selling and servicing prospects and customers without a human even needing to log into Salesforce. Now,
AI’s Effect on Software Revenue Models
when you’re in the business of selling licenses and seats, that’s a frontal assault on your revenues and your profits, consequently — your stock price. So, there’s a corollary to the phrase, “AI isn’t gonna take jobs.
It will be people who know how to use AI that will.” When talking about software, the corollary is this. And I would be remiss if I didn’t credit that this came out of a webinar that Scott Brinker and Frans Riemersma did when they released their comprehensive Martech report for 2026 earlier this month. It’s not that your tech stack, your marketing stack, or your ad tech stack are gonna be replaced by AI.
It’s that the technology platforms that embed AI within them will replace those platforms that don’t. But it doesn’t end with those snazzy sound bites. It’s the data underlying those stacks that will be the competitive differentiated advantage. And how the data gets there, be it through your company’s data collection or renting it.
Well, that’s where we are today. That’s probably the issue that is in front of you for your 2026. And we’re advising our clients to look beyond the rental model. And I’m going to take you there step by step.
Now on the surface, the headlines are boring. And you might think that these were grabs to acquire the software of these companies, the unique ways that they do reverse IP lookup, or maybe that they spider websites, or maybe the way they capture and report on keyword search history. Maybe it was about acquiring their customer lists, or perhaps they wanted to absorb these companies, their SaaS revenue streams from SEMrush and qualify into their financial portfolios. But I would say that those aren’t anywhere near the real reasons that these corporate development teams at Salesforce and at Adobe, that they use those arguments to convince their boards to pay aggressive multiples for these two companies.
The investment thesis here is about relationships. It’s about better understanding intent and better understanding identity. So why did they spend between the two companies 3.4 billion dollars to acquire these niche companies?
The Arms Race for Customer Data
Here’s where I’m gonna pop up the hood and show you why. It’s because CX platforms are in a high stakes arms race to buy high octane data about customer intent and customer identity. This is about powering AI, be that AI math or generative AI or agentic AI or machine learning models with clean data. It’s also about what’s happening with AEO, answer engine optimization, and GEO.
And I don’t think the industry has settled on what to call the next generation of SEO. But here’s the nuance about these that you need to know to understand as you build out your 2026 roadmaps, because you don’t need these data sets to do the same within your own data set. They’re totally taking different approaches. And you can do both of these approaches within your own platforms if you understand a strategy.
In short, it’s about combining identity and intent. Keep those two words in your hip pocket. We’re gonna come back to these, and I’m even gonna give you an axiom that you can use at the end of this. And I think you’ll find it useful in your day to day professional role as a CX executive.
First, let’s look at Adobe and SEMrush. Adobe purchased SEMrush to capture probabilistic marketing intent. Even though they would like to, SEMrush doesn’t have visibility on every Google search of every person across every device, across every geography, but that’s essentially what they’re mining. They own a massive click stream panel of around two hundred million anonymized users across the globe.
They record what these people search and click and visit, and then they use neural network algorithms to mathematically estimate what the rest of the world is doing. It’s basically a confidence interval. It’s a statistical extrapolation model of global demand based on what the two hundred million panelists are doing. It’s no different than how Nielsen does television ratings.
Take a small panel and extrapolate that to the rest of us. Right? So Adobe already does real time identity resolution back to that first I word identity. And so they bought SEMrush to contextualize the user’s intent.
SEMRush doesn’t know that Paul is on your site, but SEMRush does know that the keyword, let’s say I’m looking for enterprise ABM strategy, is highly correlated with say a user’s role, their timing from the first time they look at it to what they’re gonna do later across the journey, and their search history. So what does this unlock? It informs the next best action or next best offer, NBO, NBA. If I land on your website and Adobe Realtime CDP identifies me as a known user, well, that’s deterministic that I’ve landed on your blog and they already know me because I’ve authenticated.
So let’s say that I’ve authenticated before, so the identity stitching is a breeze and given Adobe’s heritage in web analytics, this is something that they’ve been doing for quite some time. SEMrush acts as a new layer of intelligence and tells Adobe, hey, the topic this guy is reading is currently trending up four hundred percent in the last week. And people who read this topic usually search for pricing models next. People with Paul’s profile also make decisions in six months from their first search and have a tendency to search in a buying group that has these five specific multidisciplinary roles that we’ve already identified.
And we believe Paul to be in role number one, which is the economic buyer role. With Adobe Journey Orchestrator, the Adobe GenAI tools can instantly generate a personalized banner that says compare pricing models instead of a generic contact us and place them in an economic buyer campaign sequence and cohort for paid media and email marketing activations starting today. So Adobe’s going to be able to go beyond matching identity. They’re going to be able to use regression analysis to match a specific user to a benchmark of intent.
It’s going to be awesome input into conducting personalization at scale and improving customer lifetime value. Along with their content supply chain offering, it will enable better personalization. Keep in mind that SEMrush has the data that Adobe needs to do this math, but neither Adobe nor SEMrush do this today without having a systems integrator like LCG set it up for a specific client. But essentially, Adobe bought a macro view of intent data that will be the high octane data to power personalized experiences, the content supply chain, and Agentic AI orchestration.
Salesforce’s Acquisition of Qualified
It’s where their platform’s going. Contrast that with Salesforce’s acquisition of Qualified. If Adobe bought SEM rushed to capture the macro view and assign a user an intent score using probabilities, then Salesforce went the other way and they purchased Qualified to be able to stitch a user’s identity. To be clear, Adobe and other CDPs already do identity resolution, but Salesforce needed to get to parity and quite frankly to catch up.
So Qualified uses signals to make extremely specific resolutions of identity. They utilize a proprietary identity resolution graph that works on two distinct levels to resolve the one and only one correct answer. That’s why we call it deterministic. First, at the account level, they map anonymous IP addresses to corporate domains telling you someone from Coca-Cola is on your site right now.
We’re talking about hundreds of millions of corporate IPs mapped to roughly 350 million global businesses. Presumably, and I don’t know for sure, but I do know that generally most companies that do reverse IP lookups use a niche provider for that. This capability has been around since 2006, and I’m willing to bet that Qualified uses a third party solution inside their own platform. They probably use Clearbit or Sixth Sense or Demandbase.
Demandbase is actually they were the first company to come to market in 2006 with this capability. But suffice it to say that reverse IP lookups are a commodity. It is not what Salesforce was interested in buying. The real power comes in matching that information with the second point, which is the person level information.
This is what Qualify calls the golden thread, which is the actual stitch that connects all of the retroactive identities and breaks the case wide open to identify who this user is. It generally happens when someone completes a form or clicks on an email link to return to your digital properties. And then the unified profile that gets connected in the CDP is called the golden record, essentially the signal that we needed to match the unknown cookies to a known user in our system. And that’s the golden thread.
It’s the stitch that the cookies needed to connect the identities of the cookies, which are unknown, with a known person. That’s zero party data. All of those cookie IDs and data studio get matched or stitched together once the user reveals who they are. When a user clicks a link, say in a marketing email, Qualified performs a cookie handshake.
It instantly links that previously anonymous web session. So we’re talking about a cookie ID, could be that Salesforce and Adobe’s, Adobe calls theirs the Marketing Cloud ID, so that they can actually connect their exact Salesforce CRM’s record with those sets of cookies from those devices that the user used to connect and log in. This is where the identity resolution graph takes that raw company name, some from say Clearbit or Sixth Sense, and stitches it to your Salesforce instance and all of the associated cookies. And they connect all of these cookies retroactively across all of the person’s devices, previous web sessions, and even previous emails or logins.
Now let’s be clear. Many real time CDPs already do this. Salesforce is matching other provider capabilities. Companies like Adobe’s real time CDP, Telium’s AudienceStream, Treasure Data, Breeze, and InsiderOne to name a few.
There are lots of companies that are doing real time identity stitching. And it’s important to note that Qualify does not have a proprietary dataset on its users. Salesforce bought features and math here, not data. Adobe, however, they did buy data.
But either way, it’s a similar investment thesis. Adobe bought the macro view. What is the market doing now? And Salesforce bought the micro view.
Who’s on my site right now? See,
The Need for Intent and Identity Data
to be really good at providing value to your customers, you have to do both intent and identity. You need to fuse the probability of the user’s intent based on market trends with the certainty of the individual’s context and moment and demographic and firmographic circumstances, their identity. It was true in first grade with your friends on the playground, and it’s true here.
You knew their names. You knew what they were going through. And if you wanted a relationship with them, you were empathetic with what information you collected about them, and you used it for good. Pretty simple, no?
And as I’ve said since episode 1 of this podcast, data is a sustainable competitive advantage. It’s the IP that has a tremendous impact in valuing your future cash flows and consequently your stock price. If you’re on a board of a publicly traded company, you’re looking at your digital business models and wondering how to get those valuations of digital businesses to be able to be leveraged in traditional businesses and the rest of your business lines. It seems that just about every business is a digital business now, or at least wants to position itself to earn the multiples of digital first businesses.
And shareholders demand that. That’s because the multiples are rich, but even subscription multiples are under siege. Everyone wants to be paid for the value they provide. And as a proxy, it seems like what they mean is that enterprise software licensing may be trending towards having customers pay for compute.
And at the root of that strategy is the defensibility and uniqueness of your customer data. In the very first episode of the Visionary’s Guide to the Digital Future, I talked about the transition of the global economy from having the largest companies transition from oil based companies to companies that are using data as their currency. If oil powered the industrial revolution, what does data power? It powers decisions.
And that’s what Agentic AI needs to be able to create relevant and meaningful and valuable relationships at scale and usher in the promise of what I’m calling the relationship economy. I said that for most companies, their data would become their most critical asset on the balance sheet. Data is more valuable than money. It’s more valuable than intellectual property.
And even in certain cases, it could be more valuable than the staff. And with Agentic capabilities, well, that’s actually starting to happen. This is about powering AI, generative AI, Agentic AI, and just about all variations of machine learning, neural networks, rags and such with clean data about customer intent, be it using math and probabilistic estimations, or be it using census data and one to one historical identity matching, which uses deterministic methods. So it could be probabilistic or deterministic.
But either way, whether via a confidence interval or an actual assessment of the totality of the information a company knows about a digital visitor, he who knows the customer best should be able to develop more profitable relationships. So
The Experience Curve and Market Dynamics
let’s apply economic laws to what is happening here. If we look back in the nineteen sixties, Bruce Henderson, the founder of a company I admire a great deal, the Boston Consulting Group, famously coined something called the experience curve. Now if you took an economics or business class, you learned about this concept when you learned about something called the marginal cost of production.
His theory was simple and profound. The company that produces the most volume achieves the lowest cost of production and therefore commands the dominant market share. That was a law of the industrial age. Volume equated to victory.
And you ought to get smart at producing faster than the next guy if you’re producing the most. That’s the premise. But we aren’t in the industrial age or even in a services economy anymore. We’re in a relationship economy.
In this new economy, profit isn’t a function of production volume. It’s a function of the relationship’s fidelity. It’s a function of data. In audio circles, the term fidelity refers to how accurately a signal represents the source.
We like to say that something is lo fi or low fidelity. We say a recording could be fuzzy like an AM radio broadcast or that a hi fi copy of a song sounds like the digital master. In this case, we define fidelity by how accurately a company can map identity. Who are your customers?
Really, by name. And intent. What does each customer want within the context of their moment? It’s a scale.
On one end, have probabilities of what the models are predicting about identity and intent on a given user. As the relationship advances across the continuum, we enrich each cookie’s profile because the user reveals information about their identity and intent to us. We move each cookie, which marketers equate to a website visit, through a continuum of unknown to anonymous users, and then from anonymous users to pseudonymous users, in cases for particularly for regulatory issues we need to use, and then from pseudonymous users to known users. So I’ve updated Henderson’s law for our times, for what I’m calling the relationship economy, and I call it the law of relationship fidelity.
In the industrial economy, profit was a function of volume. He who produced the most won. In a relationship economy, profit is a function of relationship fidelity. Therefore,
The Law of Relationship Fidelity
He or she who has the highest fidelity of identity and intent data achieves the lowest cost to deliver value and should win the maximum share of wallet.
Just as market share once belonged to the lowest cost producer, it now belongs to the owner of the highest fidelity relationship. He or she who knows and understands the customer best has the right to capture the largest share of their wallet and take the largest market share within their industry. And they ought to, if they truly can provide the best value for the user in the moment. I have a comment to make about why the law of relationship fidelity doesn’t focus on just the cost to serve, but rather the cost to deliver value.
While it’s true that marketing is a service, the lowest cost to serve isn’t broad enough to capture that the winners will look beyond operational excellence and truly find ways to delight the customers by providing more value with each engagement. Doing things faster, cheaper, smarter, and better is good for margin. But I found in my consulting travels that those minds don’t tend to focus on delivering the most value to the customer. If you can’t make the world a better place with your knowledge about your customers, you really don’t deserve to be the market leader.
Companies that combine probabilistic intent with deterministic identity will earn the right to have the strongest customer relationships because they are best positioned to truly offer the most value along a relationship over time. They should also have the highest customer lifetime value. And that’s my perspective on the investment thesis that went to the boards when the corporate development people at Salesforce and Adobe sought approval for a total of 3.4 billion dollars worth of acquisitions.
There you have it. An axiom that you can use as an investment thesis in helping you plan your 2026 budgets and strategies. Maybe you can use that to help break the inertia and build a consensus to do things differently using data for good. How are you building your relationship fidelity by improving your company’s ability to determine identity and intent and create customer value?
Before I wrap up, I have two more points to make. How will Google respond, and who’s gonna be bought next, and by whom?
Google’s Position in the Data Landscape
Where’s Google in all of this? Do they care that Adobe and Salesforce are building these data fortresses inside of their platforms?
Well, the answer is yes, but probably not for the reasons you might think. Google isn’t worried about losing the data battle. Google has all the data, like all of it. They own the browser, Chrome, which has about 65% market share.
They own the search bar with about 91% market share. They have the digital analytics locked up with Google Analytics, which is on about 87% of all websites, and they own the operating system, Android, which powers 71% of the world’s smartphones and even more, the percentages of Chromebooks. So what will Google do? Well, I’ll tell you what they won’t do.
They won’t respond by buying a competitor. The US Department of Justice and antitrust regulators have them in a chokehold right now. They can’t go shopping and they don’t need to. Their data coffers are already overflowing.
But they can respond by making it more difficult for Adobe and Salesforce to be successful with their new toys within their ecosystems over at Google. So expect Google to accelerate whatever next version is of their privacy sandbox and tighten the screws on third party signals in Chrome, Android, and Google Analytics. Their play is to make the data that Adobe and Salesforce just bought or that it’s working to stitch together harder to act on outside of the Google ecosystem. They wanna force these new agentic models to pay a toll to cross the bridge of the Google ecosystem, the Chrome browser, Android, and the like.
So friction will probably increase across the tech stack related to Google’s ecosystem and your own company’s data flows. It might not be a good time to double down on Google Tag Manager. Look for Google to respond to OpenAI’s new operating system for agents called ChatGPT Atlas. And Google has Project Mariner, which enables Agentic AI capabilities inside of Chrome.
And you might check out Dia, a browser that searches for you using LLMs that are embedded within the browser itself. If Atlas or Dia start taking off, Google begins to lose the interface. The browser wars will be back in twenty twenty six because whomever controls the browser controls what the agent can see. But in my opinion, that’s gonna be a topic that will start to heat up in the third quarter.
And as for who’s gonna be bought next and by whom? Well, if the new law of the land is no data equals no relationship or agentic advantage, then every other major platform needs to either collect or acquire unique datasets. And you should be doing this already in your company. So get your bingo card ready.
Here are my predictions for the response acquisitions that we’ll see in the next 12 to 18 months. Prediction
Predictions for Future Acquisitions
one. Now the obvious buyer here looks like Microsoft. Strategically, it makes perfect sense.
Microsoft has a browser that competes with Google. It has a search engine. It has its own LinkedIn capability, which has created an identity graph with 1.2 billion professional profiles and 70 million companies.
And now if they bought Sixth Sense, they would plug in the Sixth Sense Signalverse, which processes one trillion intent signals every single day and holds deep data on 450 million B2B contacts and 65 million companies. Think about that combination. You’d have LinkedIn’s who are they fused with Sixth Sense, what do they want? Identity and intent.
It would be the ultimate monopoly in B2B data, essentially allowing them to triangulate on what is probably 90% overlap between those two communities of users, between the two companies would create incredibly accurate views on both the intent and identity of both companies and their users. And that’s exactly the problem. Antitrust regulators would take one look at those numbers, merging the world’s largest professional network and world’s largest intent database in the B2B space, and they would just short circuit. It’s a category killer in the eyes of the FTC.
It’ll never happen. At least that’s my two cents. So if Microsoft were blocked, who swoops in? How about HubSpot?
Think about it. HubSpot is in a knife fight with Salesforce and HubSpot stock, just like Adobe’s, is also down. And all three solutions are viewed as applications and are not being rewarded with the multiples of the picks and shovel guys that I mentioned earlier or the data lorts that I mentioned earlier. HubSpot’s desperate to move upmarket and win more enterprise deals.
But right now, their new agentic solution was to launch a series of capabilities that they embedded under solution called Breeze AI. And essentially what they’re selling are tokens, but they have very limited data behind it. It’s empty except for what the business knows about the user. So essentially they’re like Salesforce and they need to do what Salesforce did and get into the identity stitching business and enable deterministic matching.
They might develop that capability themselves, but they still don’t have much intent data. By acquiring Sixth Sense, HubSpot gets intent data and solves their enterprise problem too. They wouldn’t just be in the CRM and marketing automation landscape business. They would be selling the data to illuminate The Dark Funnel.
If you’re not familiar with that term, The Dark Funnel, think of it like an iceberg. The top three percent that you could see above the water are those people filling out forms on your website. You can see them because they self identify and register and reveal themselves to you. Again, that’s what zero party data is.
The other ninety seven percent is underneath the water. It’s The Dark Funnel and it’s what you cannot see. And their journey begins well before they make it to your digital properties. It’s the research that buyers do before they identify themselves.
It’s reading reviews on websites like G2, browsing competitor websites, asking questions in Reddit communities, and reading articles on industry blogs. HubSpot could easily finance the acquisition of Sixth Sense, which in 2022 raised money at a 5.2 billion dollars valuation. It’s a privately held company, but estimates in the secondary market indicate that its current valuation is probably in the1.5
to say 3 billion dollar range. And HubSpot’s eager to improve their attractiveness to Wall Street since their growth slowed from about 40% a year to about 20%. And the market sees them as leaving a hyper growth stage and entering into a mature growth posture. Their stock went down by about forty five percent in twenty twenty five.
HubSpot has about 1.5 billion in cash and could easily finance a mix of cash and stock to pick up six cents. Prediction number two. Oracle or SAP or maybe a third party, and I’ll tell you who, maybe they buy Demandbase.
The legacy enterprise players are in trouble. They have massive install bases, but are losing the innovation narrative. Demandbase invented the ABM category back 2006. So for Oracle or SAP, this is a defensive buy to inject intelligence into their aging CX stacks before their customers migrate to another solution.
But I would say that if they bought it, it would be where demand base goes to die because I just don’t see the big investments from Oracle or SAP in the customer experience solutions that they’ve bought in the past. So that would really leave innovative pure plays like CX Orchestration platforms, like Breeze, InsiderOne, Treasure Data, Bloomreach, or even Infobip as dark horses. My third prediction is a wildcard. The Trade Desk is on the table.
Watch for AdTech giants to start getting cozy with Martech giants. These lines have been blurring for the past few years now, as the walls between advertising platforms who historically have chops in dealing with unknown users and marketing platforms whose heritage is in dealing with known users. Well, those walls are coming down. You can start combining pre click data, like an ad engine, with post click data, meaning the properties you know using tools like web analytics, with post conversion data, meaning things that are ERP systems like ERP or CRM systems.
A company like The Trade Desk is a great acquisition for a massive PE firm or a tech giant or a CX orchestration platform looking to connect ad inventory and programmatic media capabilities with digital experience platforms. It’s a terrific way to transverse the entire customer experience from pre click data to post click data to post conversion data. But the bottom line for you as a CX professional, the available squares left on the chessboard are shrinking. The independent companies that are collecting unique data about users are being swallowed up predominantly by big tech.
If you’re a Lima Consulting Group client, we’re already advising you on how to own your own data so you aren’t held hostage by those tech titans or those left trying to play catch up or trying to stitch together a Frankenstein of tech platforms who ultimately bought the data you wish to use. But if you’re sitting on the sidelines, here’s my take. The era of renting your audience is over. You need to own your client data.
You need to own your own relationships, and that means knowing their identity and intent. It was true in first grade with your friends on the playground, and it’s true here. You knew their names. You knew what they were going through.
And if you wanted a relationship with them, you were empathetic, and you used that information for good. No one else can do that for you. And why would you want them to?
Decision Makers Advange
We’ve entered 2026 with a lot of feelings about AI.
There’s anticipation. There’s FUD. Maybe some of you have entered the trough of disillusionment or confused, and it’s a mixed bag. And at about this time, most marketers and CX professionals are looking at the annual plan and the goals that came down from the mountain from the sea level with KPIs around revenue, margin, and risk mitigation, and so forth.
So you have your budget, you’ve got your team, you’ve got your mandate, and the hour to execute has arrived. And everyone’s talking about Agentic AI, but how many of you actually have a plan to operationalize it and take advantage of the data and the platforms you already have? Well, you’re in luck because today, we’re going to give you these insights for your 2026 roadmap. I have the unique opportunity to look under the hood of two global software giants who are combining forces to reshape the future of customer experience orchestration.
We’re exploring the world of Agentic AI, a leap forward where our systems don’t just assist us, but they help us automate rules and they can autonomously decide when, how and why to engage within the guardrails that we can provide them. I’m delighted to bring two of the best minds on the planet to help us unpack this. First, have Eric Matisoff, Adobe’s Global Evangelist for Data and AI. Now, I’ve had the pleasure of knowing Eric since his days at search discovery long before he took the global stage.
And speaking of stages, if you’ve been to Adobe Summit, you’ve likely been entertained watching Eric hold his own during sneaks alongside comedic legends like Wayne Brady and Kate McKinnon. But what you may not know about Eric is behind that incredible stage presence is a serious practitioner and operator. He’s literally crisscrossing the planet carrying the flag for how enterprises had used data to fundamentally change the customer experience. And I’ve been delighted to watch his journey over these years, I truly feel fortunate to have him on the show.
And joining him as a leader who’s quite literally engineering what I would call something of the nervous system of the Internet, Ervin Jagatić. Ervin is the product director and the owner of all things AI at Infobip. And when we talk about AI, it can often feel theoretical until you look at what Ervin is building. He is leading the charge on orchestrating billions of conversations, turning raw noisy signals into meaningful autonomous interactions.
I wanted him here because while Eric brings the decisioning engine, Ervin brings the global reach of communications platforms. And he’s the architect ensuring these agents don’t just think, but actually do, and do that securely and at scale. Eric, Ervin, welcome to the Visionary’s Guide. Great to be here, Paul.
Thanks for having us. Thanks for having us, Paul. We’ll start with you, Eric. And and maybe what you could do is just talk to us a little bit about Adobe, talk to us a little bit about your role there, and then kind of your perspective on Agentic AI at the company.
So at Adobe, our goal is to help people around the world to enable people to create, whether that is through our Creative Cloud products that everyone knows and loves, whether it’s Photoshop or Illustrator, Premiere Pro, After Effects, or our Document Cloud capabilities in Adobe Acrobat and Acrobat Studio or Adobe Sign. And finally, my love within my heart at Adobe is Adobe Experience Cloud. Adobe Experience Cloud is all about helping experience makers and marketers and business users get more value out of their data, out of their content, and out of their journeys. And so we’re enabling customer experience orchestration, which is exactly what my job at Adobe is all about, is helping our customers get the most value out of Adobe Experience Cloud and doing that at scale.
As you mentioned, Paul, whether that is through events around the globe I think I saw you in Mexico City earlier this year. Even earlier this year, I was in Sao Paulo at Summit Brazil. Next year, we have –my goodness — I think we have 32 different AI forums around the world, whether it’s in, Chicago, New York, Riyadh, Sao Paulo, Mexico City, London, Paris, Germany, you name it, and we’ll be there. And so the content that we’re creating and sharing and evangelizing, my team, oftentimes designs, develops, and delivers.
And so, that is my role at Adobe in a nutshell. And with my role being so long, the principal evangelist for customer experience orchestration, it just — thankfully, nobody has business cards anymore, so it doesn’t fit there — but, it does mean that I have to explain to my parents quite often what it is that I actually do. I get that. So Eric, I’m gonna start with you.
I’ve got a question, then Ervin will come back to you here for you to introduce yourself. But let’s go to Eric’s first question. So in my Digital Pulse segment, I talked about the high stakes arm race for data. And we saw that Adobe acquired SEMrush for just about 2 billion dollars last week, or at least the announcement, That’ll go through in the first quarter supposedly.
But that was a pretty significant premium, about 70% over, the stock price. So I analyze this as maybe Adobe buying a macro view, and to speak in internet, or what I would say digital marketing terms really with the intent to purchase something that would help Adobe do probabilistic intent, right, to figure out the intent of a particular user, at least in the longer term, right? It may not do that today, but that database is pretty powerful. So I mentioned that Adobe’s interest had probably very little to do with their software at SEMrush, their user base, or even their revenues.
I think you bought 20 years of human intent data. So your agents don’t have to guess what the world is searching for. So how does this massive influx of intent data change the game for Adobe’s Agentic AI capabilities? Are we moving from a world where we just create content to one where the agents know exactly what content to create before we even ask, or to realize the promise of creating content at scale or doing personalization at scale?
Yeah. That’s so I love that question, though there’s only so much that I am legally allowed to talk about until acquisition closes. First of all, we’re very excited for the acquisition of SEMRush to close. It’s a tool that I’ve been hearing about and working with since my days in the agency life.
So I’m really excited for the opportunity to incorporate data coming from SEO, coming from SEM, and coming from GEO, generative engine optimization, as well. What’s really interesting is, from an Adobe perspective on the Experience Cloud, we already have been playing in some of those worlds. So just this year, we released Adobe LLM Optimizer, which if you haven’t heard, the whole idea is tapping into the data that’s available and isn’t quite available from LLMs like ChatGPT, from Perplexity, from Google Gemini, and trying to identify when brands are being cited, when they’re being mentioned in prompts within these LLMs, what are the responses?
So if I ask, you know, what’s a great consulting group in Latin America that I can work with, is Lima Consulting Group actually being a part of their response? And if not, then what types of content can we create so that the LLMs that are traversing through Lima’s websites and blog posts and content actually starts to become a part of that conversation. And Paul, you’ve been in the industry a long time just like I have, and it feels very SEO like, right? And that’s not by accident.
And so the opportunity for SEMRush to get integrated into the Adobe ecosystem is very exciting. I think there’s a great opportunity for that data to be incorporated, for our agents to then learn from that data, for customers of SEMRush to get value from Experience Cloud capabilities, applications, and agents. And so it’s an exciting world. There’s not much more that I can say about the SEMRush acquisition until it does close.
Hopefully, it closes very soon because I can’t wait to get my hands on it and for Adobe customers too as well. Yeah. And then the other segment, I talk a little bit about, you know, the I mean, that data today is kind of macro intent data, but it wouldn’t surprise me if where you’re going is to be able to use that macro intent data to do probabilistic intent, say this cookie, this user is going to want to do XYZ. And of course, Adobe is in better place than anybody to be able to orchestrate that with the content supply chain and personalization scale features.
So anyway, maybe that’s something that you’ll release in the future, but we’ll get there in the first quarter when that puppy closes, right? What’s really interesting about Adobe LLM Optimizer today is there’s a lot of really powerful features that customers are already getting value from. The product was released at the end of Q4, and we’ve got customers lining up asking questions about it. And one of the really cool features that was just released at the end of December is called Auto Optimize on the Edge.
And this is the whole idea. If you think about the way that websites have been built for the last twenty plus years is a lot of the content that’s on them is personalized, right? That’s been, all three of us know, we’ve been, you know, that’s the goal, right? Personalize the content on the edge or personalize it at scale.
Use JavaScript. Use Adobe Target. Use whatever you want in order to customize that content so that when the visitor, when the consumer, or when the prospect gets to the web page, they get the message that is ideal for them. The problem is it’s, these agents or these crawlers that the LLMs are using aren’t running the JavaScript to enable all of that personalization.
So if you go to a website, maybe, you know, when a human views it, there’s ten thousand words. But when an agent or the LLM crawler views it, there’s 15 words or 20 words because everything else is being added dynamically using personalization. And so we have first of all, we have a free plugin that you can go to that is llmo.page to get to it.
It’s a Chrome extension that’s free for anyone. And what you can do is load it on any web page in the world and see the difference between what humans see on that page in terms of content and what LLMs see on that page in terms of content. And you’ll see sometimes there’s a pretty significant difference between those two. And so this Auto Optimize on Edge page, what that does is it takes that same information and makes recommendations or basically says, How do we get that page accessible on the edge to these LLM crawlers so that they can now we’re at one hundred percent between the human view and the AI agent crawler view.
And so that’s a really exciting feature that I’ve toyed around with just a little bit. And what I love about it is it’s able to do it at scale. So it recognizes, you know what? There’s tens of pages, dozens of pages, hundreds of pages, thousands of pages that are missing, that aren’t AI readable.
And click this button to deploy, and it basically says, Okay, we’re going to push all of that personalized content to the edge so that when AI agents crawl it, then they’re able to see it. And even cooler is it works for any CMS and any CDN that customers are using. So totally agnostic in that world as well. Anyway, you can tell I’m excited about it.
And I just love the opportunity that customers have to just simply make their brands more visible to these LLMs and their agentic crawlers. Well, it makes a lot of sense. Right? I mean, when you talk to SEO professionals, you know, they usually talk about two work streams.
You’ve got a technical SEO, and then you have the content SEO, activities. Right? And so what you’re talking about with the LLMO, if I got it right, for the optimizer would be kind of a technical SEO tool to help drive parity between what humans and machines see. And I think the other thing that I that I was really excited about, and and again, our subscriber base is mostly, CX Visionaries executives.
And so I think there’s the LLMO and also the, just the, the features that are in, other parts of Adobe Analytics and such can help drive the content and editorial calendar, right, and know what to write about and what is really happening with the drop and the decline that all of us are experiencing with regard to a decline in our internet traffic due to the fact that Google isn’t the first place that a lot of our searches are coming from anymore. Right? So how would it impact the editorial calendar? Any any features there, Eric?
Yeah. Absolutely. So also as a part of LLM Optimizer and Adobe Sites Optimizer, which is a similar set of capabilities but is focused on more of the SEO side of things. So in terms of content recommendations, technical recommendations, redirecting 404s, all of those kind of things.
Another piece of those technologies is making recommendations for new content to create, whether it be blog posts or articles or FAQs, even tapping externally into sources like Reddit and LinkedIn, which are oftentimes influencing the responses that these LLMs are providing. The takeaway there is to use some of the features that you may already have to influence your editorial calendar this quarter. Let’s pivot. Ervin, I wanted to pivot to you.
First, I’m going to try to introduce who is Infobip, But you might have to come back over the top and, and, and, and, and clarify it here. But I, I classify Infobip as a company in the category of software, or maybe even in the industry as a CPaaS, communications platform as a service. And that enables businesses to build connected experiences across all stages of the customer journey. So Infobip has over 800 direct carrier connections.
So we’re talking about, carriers in the telecommunications world as Verizon and Comcast. Also, believe it or not, Google. This is the world’s most connected CPaaS provider. They orchestrate over 530 billion interactions annually.
And through their platform, businesses can orchestrate communications via email and marketing channels, just about every communications app you could think of, starting from say WhatsApp, Apple Messages for Business, SMS, something new that needs to be on your calendar for this year is Rich Communication Services. For those of you attending or listening in, in North America, and about 16 other additional channels. So, Ervin, before we go off of that, or go from there, is there so this is kind of a pause here. Is there anything that I misquoted or anything that that I that we need to clarify about who is Infobip?
Thanks, Paul. So this is pretty much really accurate, and thanks for the for this intro. So pretty much Infobip solves the problem when the business needs to communicate with the end customer. So we orchestrate around 550 billion.
I think this year will be like around 600 billions of interactions. So pretty much where business application is to send the message and deliver this message to a user, pretty much Infobip is there to orchestrate. And in our portfolio of messaging channels, like besides, like, SMS, MMS, there’s, like, email, WhatsApp, like, any over the top channels or any, like, other channels, like, Instagram Messenger or pretty much any way how to communicate with the end user. Perfect.
The Role of AI in Communication
So if Infobip is orchestrating billions of messages globally, you must have a unique view into, I would say, the signal data that most SaaS platforms may not. Right now, in the case with, Adobe, which is just, I think, the richness of this communication, they are doing web collection. In other words, data collection methods using Adobe Analytics so they can see and observe. But most of the SaaS platforms in the market, don’t have that capability.
Right? So in this, as the way that you do, especially with tools like RCS and rich communication apps like WhatsApp and Viber and the rest, you have really rich analytics. What did my thumb touch on my mobile device and the carousel of the four products it recommended? Right?
So in this agentic era, how is Infobip thinking about AI and AI agents? How do you ensure your agents have fidelity that are the mapping between identity, they’re mapping intent so that they’re really creating value? So there are, like, a couple of points. One of those points is like this what Eric mentioned, like, this high hyper personalization.
So pretty much personalization that that messaging is actually relevant at the relevant point of time around the relevant topic. And also the the other thing is that messaging is like a two way. So that kind of the user can get the message, user can reply it. And then like by utilizing the channels, as you said, like high fidelity channels, like where user can send pictures, they’re like product catalogs, they’re like multiple functionalities embedded in in the communicational channel.
So combining hyper personalization and this richness of the communication is pretty much how it will be how communication will be shaped in in the future. Because prior prior to AI, there was like a need for really comprehensive UI tools, like applications, reach mobile apps, so which will need like a lot of different like features that are kind of these features were curating user and customer experience. But with the kind of with the adoption of AI and the adoption of AI agents, this kind of rich feature rich experience will go more to kind of conversational rich experience. And then we’re partnering with companies like Adobe where kind of the CDP and marketing data lies, and then pretty much it will be kind of enabled to curate the customer experience so that it can be like two way conversational and that this delivery of the communication will be go through the kind of the channels that Infobip is providing.
So to kind of to summarize is, like, to hyper personalization and using the communication channels, and then agents will be kind of curating based on the actual user need communication. You know, in preparation for this, we had a conversation the other day and you had mentioned to me a little bit about how you were thinking about AI. Right.
Human to Human Interaction in Marketing
And, and you, you mentioned to me that, you know, and I think nowadays the, the language of B2C or direct consumer, I mean, a lot of times people are just talking human to human interaction, right?
So if we just start with that basic of understanding that a human being on the customer side, whether that’s in a B2B purchasing decision of a team on the left of where the, the purchasing is dealing with maybe a team on the right if they’re buying enterprise software, or maybe it’s just a consumer buying a single product from a brand off of a website. In that regard, I would still say it’s, you know, direct to consumer or human to human interaction. But how are you starting? You you you started taking us down a path where you started introducing the AI agents into that mix.
So tell me a little bit about how you’re thinking about AI and the evolution of human to human marketing. Yeah. So pretty much what what kind of we see from the data and from the initial kind of use cases that we are launching in the production with AI agents is that customer service industry is pretty much underserved. And there’s like there are like a lot of phenomenal tools today on the market, but there are like so many much possibilities with AI agents.
So the kind of the the end goal is that every user, every customer has, like, their own representative with the brand who actually under so that brand understands the this particular user, why is using this brand, why the what kind of products this user wants, What what will be, like, the next best step? Because today, pretty much, it’s not it’s you always, there there are, like, some cohorts, and it’s, like, kind of the not personalized way. And pretty much, AI agents and digital workers will enable that. We did that.
They will enable us to have, like every every one of us will have, like, their own personalized banker, for instance. Regardless of how big our portfolio is today, this is, like, enabled only for first for the people for with a significant portfolio. But tomorrow with AI agents, everyone will have, like, their own personal AI agent who mean, for instance, like, from the Fintech sphere, like, who will understand the the financials and the kind of the perspective of the of the end end user. And this is really pretty pretty significant and pretty powerful, and I think that that will be, like, in terms of, like, when people are talking about AI agents and AI, like, really will be, like, jobs for humans.
The Future of AI and Job Market
There will be, but the the horizon will significantly will be broadened with the capabilities that we’ll have. Yeah. I mean, this is a trend that we’re gonna see start to see in 2026 with Atlas coming out of OpenAI. Google has their project Mariner that they’re embedding into Google Chrome.
And then there’s a new browser that if you haven’t heard it yet, you gotta download and check it out. It’s called Dia, and it’s an Agentic AI kind of enabled solution. I don’t know if if, you know, Eric, I saw you respond to that. I don’t know if you have anything more to add on on kinda what’s happening with our our own, our own assistance, if you will, with with Agentic AI capabilities to help us engage and get jobs done as individuals or as employees at a company.
It’s been interesting playing with some of these new Agentic browser capabilities. So what’s the test that you test the robot to see if it has human What’s that called? The Turing test? Yeah, the Turing test.
So I have my own little Turing test that I go through. Oh, do tell. This would be interesting. So every time I try a new Agentic AI capability in a browser, I tried it with ChatGPT Atlas.
I tried it with Perplexity Comet. I’ve tried it with Copilot. And what I find is, so far, I’ve yet to get one of them to pass. And it’s such an easy thing that would save a total of about 18 seconds of my day every single day.
But one of the things that I do every day is I As you know, I’m an analytics nerd. And so I am oh, I’m always checking every single morning how many learners I have for my three LinkedIn learning courses. So LinkedIn provides this subpar set of data and visualizations in, like, their for instructors where you can go see for your courses how many learners have taken them. So I’ve got one course on an intro to Adobe Analytics, another which is like an intermediate Adobe Analytics course, and another which is an intro to customer journey analytics.
And so the problem with it is it’s not very good at providing historical data. So you can look at data from 7 days ago or 14 days ago or twenty one days ago, but nothing more granular than that. So what I do is every morning, I’ve got an Excel spreadsheet that I go through and I say, Okay, well, how many learners do I have today? And I manually data entry and how many learners I have, and then I can see it trended and I can see, Oh, okay.
You know what? There’s been one hundred over the last day, whereas it would have been a little harder to tell that from the built in tools. Anyway, my idea was using Agentic Browser so that I can just click a button, and it automatically goes through and records it, puts it into an Excel spreadsheet or a Google Sheet or whatever, and does the very basic math for me. So, like, really something that anyone with a little bit of Python knowledge could do or someone There’s a hundred different ways to do this, and I do it manually, but it seemed like a perfect opportunity for an agentic browser, right?
All of the data is in one tab, and then the destination for that data is another tab. And I’ve gotten really close to being able to train one of the agentic browsers on it, but I’ve yet to succeed. And so that is my own little personal Turing Test to make sure, is it actually going to be useful? Is it going to save me time or not so much?
And it generally seems to trip up on multi tabs and multiple domains is where it tends to have trouble. I’ve gotten close enough that I was able to get some of the data copied over, but then the next day running it, for example, it’s like, Okay, well, I’ll just create a new Google Sheet. Adding to it is where it becomes problematic. What I’ve found in general with these new AI technologies, these new Agentic technologies is if you wanna kick their tires, have a use case in mind.
Figure out a way like, spend time away from your computer thinking about, well, what are the mundane tasks that I do daily, weekly, monthly that maybe I could train an AI agent to handle? And that’s the way that we’re also thinking about AI and agentic technology at Adobe as well, is where is there an opportunity for us to take things off of the plates of the people that are in our tools every single day so that they can then elevate the time that they’re spending? I can spend less time copying and pasting values from a browser tab to Excel and instead start thinking about my next course.
Same thing with Adobe is you can you spend less time manually dragging and dropping things and more time thinking about what you should be manually dragging and dropping.
Preparing for AI Implementation
There’s some factors in order to prepare your use cases that you need, right? Do I have the people? And that, that people might be that you might have the people, but do they have the training and the enablement and the support?
Do I have the processes? Right? If you haven’t documented it, you can’t digitize it. And documenting generally means you need customer journey documentation and business process modeling notation written in BPMN 2.0,
so that when you give it to a builder, they can build. Right? And those are things that we do a lot of at Lima Consulting Group is CX Journey Orchestration. Those are kind of pretty pictures and then the the data mapping and then the BPMN compliant process maps.
So, and then the last part, so people, the processes, obviously the platforms and our whole point in today’s podcast is to be able to help folks understand that they really should be taking advantage of the marketing tech platforms that they may already own or that they can extend into in the case if they don’t already have something like an Infobip that it’s an Adobe partner that can turn it on. But the point is that I don’t think that it’s gonna be AI that takes people’s jobs, right? We’ve heard this phrase a million times. In the case, Scott Brinker and, and, and Franz the other day published out, you know, I don’t know, two hundred pages of reporting.
And I got the opportunity to sit down and listen to them talk about that master class in the Martech report that they do annually. And the key takeaway that they said, what is it’s not that AI is going to replace marketing technology. It’s that the marketing technology that effectively uses AI is going to replace those that don’t. So I have a new axiom I’m working on and I, I’ve been on the board for the, editorial board and a writer on, a journal called the Journal of Applied Analytics.
And I’m writing a paper about, the, relationship economy. Right? And and I don’t think it’s the agentic economy. To me, agent and autonomous and AI, it’s like a beeper.
It was here today and gone tomorrow. It’s like a fax machine. It’s here today, gone tomorrow. What are we trying to do?
I think we’re trying to improve our ability to do relationships. And what we just learned from Ervin is that human to human interactions are going to be maybe human interactions regarding their personal agentic capability or internet or their their solution interfacing and and bet and it it it might move from human to agent. And then on the agent from that individual or that company, they may, that agent may work with the agent of the other side of the brand that they’re working with, who then in turns worked with the human. So there might be any permutation.
The human agent could be agent to human, could be human to human, or could be agent to agent, but doesn’t really matter. If you can’t create value, we’re, we’re going down a hole that doesn’t need to, there’s no value there. So what I’ve been working on is a paper about the relationship economy. And what I’ve been able to kind of summarize it in a single sound bite is that profit is a function of relationship fidelity.
And relationship fidelity, meaning how accurately we identify who someone is. So who is your customer? And correspondingly, how accurately we can understand their intent or what do they want? So think about your best friend from kindergarten, first grade.
And when you walked up to them in the playground, what’s your name? What do you like? That’s identity, and that’s intent. And he who has the best relationship fidelity of understanding identity and intent should be able to produce that value cheaper than anyone else.
That’s the marginal cost of production. And so in the industrial era, he who had the lowest marginal production cost should have the most quantity and they should have the best market share, the most market share. And so in essence, he who has the best understanding of relationship fidelity about intent and identity should be able to create relationship value the most economically and should therefore take the most market share. So that’s the gist of the paper.
Enhancing Relationship Fidelity
You both sit at the intersection of this with the data that you both collect and your ability to do customer orchestration. So how do your two platforms work together to increase what I’m calling relationship fidelity? Can you give us any real world examples of learning about identity and learning about intent and activating that? So let’s start with you, Ervin.
Yeah. So pretty much what Infobip how is Infobip partnering with Adobe is providing the tools, communication tools so that Adobe can, like, deliver the messages across the CPaaS stack that we have in our portfolio. So the power of Adobe Adobe marketing products can be connected with the communicational channels. So this is pretty much where the partnership aligns with between Adobe and Infobip.
I’ll add to that, that there’s two sides that you mentioned, Paul, which is identity and intent. And it’s really interesting to see how we can kind of use the channels that Infobip makes available to Adobe customers in order to activate based on the identity that we have of consumers or that brands have of their consumers, And then the intent, the opportunity to actually go ahead and drive action, drive conversion, drive whatever the next goal is that you have within your campaign or what you’re building. And what’s really interesting about this new world that OpenAI and ChatGPT has unlocked for us over the last 3 years is that consumers are being more proactive in telling us what their intent is than ever before.
Prior to three years ago, pre ChatGPT life, it’s like you’ve got BC and a not AD, but I don’t know what that a you know, before ChatGPT and after ChatGPT. Right? Prior to that, one of my favorite analyses as a web analyst to do was analyzing internal search terms. It’s like the quickest bang for your buck that exists is when l I used to say this.
I’ve said this one hundred thousand times. I’ve said this like, you can Google this sentence, and you will see a picture of my face pop up talking about it, is there’s no other time that a consumer tells you exactly what they’re looking for than when they use internal search on your website. They type in, you know, sneakers or they type in, you know, blue zip up or they type in, you know, B2B software or CPaaS or whatever. They’re telling you exactly what they want, and then that’s the perfect opportunity for you to, number one, provide them with the information they’re looking for.
But number two, analyze the failure or success of whether or not you succeeded. So we’ve got a whole set of data in Adobe Analytics, in Customer Journey Analytics, across Adobe Experience Platform that provides that information of what was the search term, how many results were provided, what results were provided, did the end user actually end up clicking one of those results, did they click one of those results and then convert? All sorts of really interesting opportunities for analysis there. What if you filter all of the internal search terms on your website to just those that have zero search results?
Like, is there a more actionable set of data that you could provide with one table of data? Right? It’s like one dimension and one metric. Show me the search terms that have no search results.
But what’s really interesting in this After ChatGPT world that we’re in is now a lot of brands are starting to take advantage of adding LLM capabilities to the front page of their dot com, of their websites. At Adobe, we have a technology called Adobe Brand Concierge, which enables that type of functionality where you can train it on your product data, on your marketing messaging, on your campaigns. You can train it on the tone of voice, how it should be responding. And then you can pull in other Adobe agents and non Adobe agents to actually help support the response.
And again, talk about a goldmine of data to optimize that experience. We now finally have a second way in which consumers are proactively sharing what exactly they’re looking for. We’re personalizing that response, but, oh my gosh, such a great opportunity to improve it, to personalize it, to optimize it for those consumers. So it’s something that has been sort of a second order effect of the release of ChatGPT, is there’s these new sets of data that we can analyze in order to further improve the experience that consumers and that prospects have on our owned properties.
I love that, right? And I think another way that I’ve characterized that, and I, like how you said it actually more clearly than I do. But I I think about the phrase, if you’re not here, raise your hand. And in the case with okay.
Somebody typed in, you know, orange sneakers on Nike.com, for example. And for whatever reason, nothing showed up. I have a pair of orange, Nike sneakers, by the way.
My my, orange is my favorite color, as you can see from my water bottle here. And the, my indoor soccer sneakers that, I wear are orange, bright orange, Nike sneakers. Well and and I like orange too. You’re part of our corporate brand.
Right? And I think, orange and blue, I guess, Denver Bronco colors. And, but in any event so so we share that. Either that or you’re not.
Orange and blue are the Lima Consulting Group, colors because, so my dad, who’s now retired, founded a community bank in Philadelphia, I don’t know, 10 or 15 years ago, like 15 years ago. And as he was building the brand for it, he was doing a ton of testing for what colors, to make the branding color to make the for the branding. And worked with an agency, they came up with a bunch of different designs and different colors. And this is a total aside.
This has nothing to do with AI or agents or anything. But he would show to my friends, to my brother’s friends, to people in the neighborhood the different options and just get their gut feeling. And the one that consistently was preferred by people that were younger, like our generation and younger, was the orange and blue combination. It’s sort of like a burnt orange and like a navy blue, and it really looked good.
And that was the one he ended up going with. So it’s really funny to hear that that’s the Lima preferred set of colors as well. And it’s great to mention that also Infobip brand color is orange, so we have a full circle. That’s really I wonder, let’s I’ve got all the things.
I wonder if Adobe needs to change our red. I don’t I don’t think that’s gonna happen. Oh, no. No.
The the chief chief marketing officer there would would you’re gonna short circuit them. I know I know Anne’s no longer the chief marketing officer, but, Lan, Anne would would go would go nuts. But I I’ve seen your really cool red sneakers on stage, Eric. So, you know, when you and then you had the fellow from, the the Dutch fellow who would always wear, I think he was from Neil the Leo Neil Lane acquisition, and he always wore the, you know, I mean, the the the shoes that he he needed to plug in because they were like this electric orange.
Right? They they were the top. Isn’t that Bertle that would that would always do that? Bertle.
Yeah. Yeah. Yeah. He was, very proud of his Dutch background. But in in, so that’s that’s a fun aside.
And and what I was thinking, though, is is, you know, if you’re not here, raise your hand. And so in the case where you have basically a search that has a null response, in other words, you type something in, you don’t get anything back on your website. At least you can see that there was an old search. In other words, there’s no content.
I think the thing that is brilliant about what you were saying is it’s an argument to bring in LLM on the dot com homepage. Because if that null search happens on ChatGPT, guess what? You don’t even know. So if you’re not here, raise your hand.
Well, you don’t know that that search took place. And, of course, that nothing and it might have been even an absolutely super specific search. Right? For Nike.com,
I need you to, you know, go on their website and ask for and find me orange sneakers. And they may have 70 of them, but if they don’t show up, off your website, you would never know that. So you’re gonna wanna bring those yeah. It’s an argument to bring LLM capability into your own search capability on your own website.
In the case, Ervin, with what you’re talking about, know, your approach for AI, tell me about some of the features and capabilities and benefits of what you’re busy working on at at Infobip?
Infobip’s Communication Tools for Agents
Yeah. So pretty much what we’ve been working on and what we will be releasing this year is providing our platform for the agents. So our platform was tailored for builders with so we which we’re integrating with info before APIs.
So different applications were integrating info before APIs and then sending the messages through the APIs. Now basically, the next step and the next evolution is integrate our communication stack over the for the agents. So over the MCP protocol. So this is something what we already launched last year.
And again, like support a lot of features that will enhance the genetic experience when an agent, for instance, like when an Adobe agent needs to communicate, we want to create the best possible experience for this Adobe agent to kind of initiate communication anywhere in the world, like, globally. And this is something what we are working on, and this is something what we are kind of obsessed with, like how to enable agents to communicate, how to enable agents so that they can out autonomously use the communication channels and remove the friction, the human friction in the barrier in terms of compliance and monitoring of the communication, so that it is safe compliant, but also that agent can be autonomous in terms of the execution of the communication for the end goal and that’s like meaningful communication, meaningful interactions for the end customers.
Well, you mentioned it, so let’s go there. And Eric, I’ve got the same question for you about what you’re working on there at Adobe. But before we go there, let me because Ervin and being a product leader and and a builder, it’s not surprising that Eric, or Ervin put on the table the, MCP or basically model context protocol. For those that don’t know what that is, Irvin, let’s give us a little master class quickly.
What what is that, and why should is it important? And why should customer experience professionals care about MCP? So MCP is pretty much a protocol similar as APIs. So we know how the application is communicating with other application over the API over the API protocol.
MCP is standardized protocol, but for the agents. So how agents are using different tools. So pretty much they are using over the m MCP model context protocol. And this is something that kind of the unification part is kind of in increasing the pretty much in enabling agents to to integrate more tools in in a easy easier way.
So that’s like in short terms. It’s like just a protocol how agents are using different tools, different that they can when they when they’re executing a task so that they can leverage leverage these tools in order to done make the task successfully done. I heard a really nice I heard a really great metaphor for explaining what MCP is, and I will not take credit for it. But what I really I I thought it really kinda explained it well is is it think of it as, like, u USB.
You’ve got all these different devices that have USB, USB C, or whatever it is, and when you then plug them into your computer, they all just kind of work. MCP is similar to this USB protocol in that it enables you to just plug your agent into this protocol and it’ll effectively just work. I thought that was a good metaphor to explain it for customers that are unfamiliar. Yeah, and I think it’s important, particularly why some of those words are there, particularly the word context.
A lot of what happens with AI isn’t like other technologies. It’s regenerative in many cases. It’s doing something for the first time every time you ask it, even if you asked it to do it, you know, now and then in 30 seconds from now, the exact same thing. And in order to be able to work, it doesn’t use columns and tables the way an API would.
And so in order to be able to plug in, to your data, particularly unstructured data where you’re training it on, you needed a new set of APIs, right? And that’s why it’s a different acronym. That’s why it’s different. So we, we said we were going to deal with, or at least talk a little bit about mon, model context protocol as kind of as middleware for Agentic AI, particularly, to be able to train it over your data or to be able to plug in your own data sets, be they CRM, ERP, your content, your product data model, or master data model, your PIM, whatever other enterprise data sources you have, and to be able to have this middle layer that that it’s able to to engage with, the way that a LLM or a, it could even be a rag, but an AI solution thinks and works.
It doesn’t think and work the same way as, ACR rules engine where we do if then else, it works a little bit differently. So you gotta have a little different middleware. And in order to plug that in, I mean, you know, let me go back to, Eric. What are you working on?
Adobe’s Agentic AI Developments
I I know when I saw you, last speak in México, you had a slide with a handful of new agentic capabilities that you had released. Some you released early in the year and some which you’re releasing almost every couple of weeks, it seems like.
Yeah, that’s true. We are keeping busy on the Agentic AI side of things.
So in March at Adobe Summit in Las Vegas, we announced the pending release of, I think, 12 or 11 agents that are built on top of Adobe Experience Platform. And then on top of that, just in December, we released an additional 5 Experience Platform agents that are meant specifically to be used within Adobe Experience Manager as well. So those are helping with deployment, helping with governance, helping with experience production, helping with discovery, and more. And what’s interesting about the way that we’re building them is we want to make sure that it’s a better experience for our customers.
And so the way that we’re making that happen is by building these agents on something that we call the Experience Platform Agent Orchestrator, which is kind of similar, if you think about it, to the conversation we’re just having about what an MCP is and that making sure that these agents are orchestrated and accessible regardless of what pane of glass you’re actually interacting with them. So if we if you want to, you can, log in to Adobe Experience Cloud and utilize the AI assistant and engage any of these fifteen, eighteen different agents without specifically saying, hey, data insights agent, I want to work with you right now.
Or hang on, switching gears. Now I want to work with Journey Agent because I want to build a journey. Instead, it’s just simply orchestrated literally by the agent orchestrator. It’s a good name that we came up with so that you can say, you know what?
I wanna analyze this data, and it pulls in the correct agent from there. Okay. Great. Here’s an interesting set of data.
I wanna build an audience behind the scenes without the user having to care. Is this the audience agent? Is it the journey agent? Is it the data insights agent?
Is it the workflow optimization agent? What have you. Instead, it’s just all a single pane of glass and available to you. Now, of course, if you’re in AEM, you’re going to Adobe Experience Manager, you’re going to tap into the agents that make sense to be using in that context.
And that’s all made possible through this agent orchestrator. And so each of the agents that we’ve released within Adobe Experience Cloud sit on top of this Experience Platform Agent Orchestrator to ensure a number of things. It ensures that the right agent is getting pulled in at the right time. It ensures that there’s a strict level of governance so that the data isn’t leaking out or that there’s a minimal amount of hallucinations, all of those types of things.
So that’s really what I’m most excited about. I’ve kind of played around with some of these new ones and we continue to iterate with them as well. So when they were first released, when we saw each other in Mexico City, a number of the agents were focused on extraction, right? Extracting data, extracting information about your consumers, about your audiences, about your journeys, about your assets.
But now they’re starting to move into this world of creation, where we’re no longer just saying, How many audiences focus on cart abandonments? But instead, we can say build an audience of consumers who have recently abandoned their cart or get even more specific of have abandoned their cart or more specific, abandoned their cart on a specific set of products in a specific region, who have interacted with a campaign or who have previously purchased and keep getting more and more granular, now we’re getting to a world that regardless of where you’re building that audience you’re building it in Adobe Experience Platform.
You’re building it in customer journey analytics. You’re building it in journey optimizer. You’re building it in non Adobe tools. You’re building it with SQL.
You’re building it with Python. If you’re able to really get kind of granular with the audience that you’re building just through natural language, it really can actually save a significant amount of time. And that was what’s most exciting to me as, like, a seasoned practitioner. When I first saw, Okay, great.
I can build a visualization with a prompt. And I’m a fast typer, but I’m an even faster click and dragger. Like the rest of us. So you’re human like the rest of us, Eric.
Yeah. Like, I can use Analysis Workspace really fast. I would know the keyboard shortcuts. I can drag.
I can drop. I’ve been using it since the day it was released. But if I’m able to build complex things, not just simple things, then that’s where the Agentic capabilities start to really unlock some opportunity for better productivity. Yeah.
I mean, to kind of sum that up, I mean, as a practitioner and a marketer, you’re able to start using the agent orchestrator to help basically manage, you know, faster, cheaper, smarter, better. It’s a marketing function. It’s a customer experience function, and you’re probably doing it in a way that you may not be able to do manually. And it may even open up new, experiences that you’re able to use to do personalization at scale and accelerate your content supply chain.
Right? So how do we get there and how do we use the brand concierge to produce more content? All those use cases are basically for the marketer and the practitioners at the organization. Ervin, I want to go back to you and, and, and give you an opportunity.
I know that in a lot of what we’re, what I think you’re developing at Infobip, particularly around this promise of conversational commerce and having agents work, maybe on the consumer side. Do you have any real world examples? What are you working on to basically make either the company, the marketer, the practitioner’s life easier? Or I know in a certain cases, maybe the improving the customer experience for on behalf of the end customer.
Real-World Applications of AI Agents
Yeah. So what we kind of released is the first production use case of AI agent who is involving customer support use cases. This AI agent lives in the WhatsApp and is serving for digital insurance. And, basically, it’s kind of the evolution.
So it’s company first released on the WhatsApp, like, two way conversational AI assistant, and then into and then we have evolved into with AI agentic functionality so that kind of this AI assistant is kind of has these sales capabilities and can sell travel insurances over over the WhatsApp. And this is pretty much what kind of resonates with our strategy, and also that kind of what we see is that marketing data and customer support data will be unified at some point of time in order to kind of so that AI agents can then serve like a full customer journey. So for instance, like, if I as a user have, like, a problem with the product and I’m, like, interacting with the customer support so that I can easily share this info also with the marketing data and then kind of adjust the communication with the marketing part.
So pretty much today, these are the two separate departments at company. So it’s a marketing department, and then there is a support department. And usually like there are like two different softwares, like there is a marketing software, customer support software, and they’re like data services, so there is like one profile with more around a user in customer support, and then the other one is in the marketing. So pretty much with the evolution of AI, we see that companies and that that, like, consultants, companies will help play a huge role in terms of making this awareness.
So if you knew that the unification of these two data sets is really, important. And then the tools for the activation of this data, kind of how to activate this data in terms of, like, to serve in the the for the best use case. So that that’s something how what we kind of released on this support sphere, support use case. It’s like the digital insurance, and we’re it’s really, really great to see that Agentic AI capabilities are working, delivering with the end customers safely.
And then and serving the purpose. And the on the other hand, we need to be see that kind of the consolidation of the data between marketing and support is needed in order to activate this data properly. Well, I had just a comment, and then I wanna another follow on on that. So I think when we look at ad data, that’s the third piece of these silos that isn’t really connected.
And I think that the walls breaking down between ad tech and the data there, marketing tech and the data that you own, you know, that you have the pre click data, post click data, and then post conversion data. And each one of those has different sets of data. Pre click is your, your Google AdWords data. Maybe you’re using Trade Desk programmatic tools.
Post click date data is your marketing data. That’s your MQLs, your, it has your, your, your anonymous users and your web analytics. And then post conversion is what they bought. And so they’re in CRM, their, their margin is known, their ERP profile is known.
In most cases, sometimes it’s not the case in retail organizations. So as you start to, as you said, in your case, you’ve got, and then you’ve got basically the service data, but these silos are breaking down. And so how does that manifest for Infobip? Is that happening in a chatbot?
Is it happening in your case, you talked about WhatsApp. So, like, what is that use case, that, you know, someone asks a question, I have a how does it break down and how does it manifest? So the cool thing with Infobip is that kind of it can manifest in in any communication channel. And that’s like the where kind of Infobip orchestrates and owns, like, multiple communication channels.
And it can be, like, in WhatsApp. It can be in live chat, or it can be, like, over the email or over the SMS. But pretty much when we kind of with the brand, when we kind of when brand orchestrates the data from the info web, this can be, like, orchestrated properly with our conversational customer data platform, which then composable can be added on, for instance, like, on Adobe tools. So that Adobe marketing platform, again, has, like, all this visibility, like, what’s kind of and what’s happening across the communication channels and maybe from the support side of of the tools.
So that that’s kind of the value proposition of Infobip, like a really broad set of channels, analytics in these channels, and then composability part that it can be like composably added to the Martech platforms or to the to different different platforms that that have like really deep niche of the of the functionalities. You know, as a consumer, I’m excited for for improved customer support whatever in whatever channel I’m heading to. So glad to hear about this great work that you’re doing, Ervin. Thanks, Eric.
Yeah, I mean, ultimately, if you can’t create value with these tools, then what are we doing? Right? We ought to be making the world better through improved data. If we can’t add that value, then we’re really not adding to what I would say the, you know, the GDPs of our respective countries and our stock prices, the companies we work at.
And just professionally, it’s not very rewarding to, to, to work if you’re not adding value, to the end, the end user, right? And really helping people. So I think having empathy, which is like the first part of this is just it it goes back, and that’s something I look at when I’m hiring is, you know, I ask questions, and I I just watching a a a silly little TikTok thing, but the guy that he what he’s saying is, you know, hey. We’re gonna go and get a cup of coffee in the corporate cafeteria if you don’t ask to throw the coffee out or ask what to do with the cup when you walk out.
Apparently, in his world, you fail the interview. That’s empathy. You’re looking for empathy. Is that a silly thing?
I don’t know. But empathy certainly isn’t silly, and we need to get really good at it, using these tools. Because rather than distancing ourselves from our end customers, we need to be closer than ever with understanding better our relationships through identity and through intent. Right?
I, our time’s coming to a close. I have basically two more questions. One’s a question and one’s a little fun little thing we’ll do at the end to wrap it up. So and and, Eric, I don’t know if you got a chance to talk about any, you know, global deployment.
So maybe we just finish out. We we basically, Ervin talked about a use case. Did you have a use case you wanted to talk about too? Yeah,
Adobe.com as a Use Case for AI
I could talk about how adobe.com
is using some of our tech. Yeah. Adobe on Adobe. That’s a fantastic use Let’s do it.
So what’s really exciting at Adobe is that the very first customer that uses our Adobe Experience Cloud capabilities, often times before they’re even released, is Adobe.com. It’s one of the most visited websites in the world, billions of users, billions of visits, billions of hits. And it’s a great opportunity for basically our new technologies to have their tires kicked before any other customer in the world gets access to it. It’s great because we receive feedback, we receive opportunities to improve it before these technologies actually get released.
And not surprisingly, Adobe.com has been utilizing our capabilities for Adobe Brand Concierge, for Adobe LLM Optimizer. In fact, it’s the biggest Adobe Experience Platform implementation in the world. Over one thousand campaigns and journeys, tons and tons of data and profiles and opportunities to analyze and engage.
And recently, the Adobe.com team has been using Adobe Sites Optimizer and Adobe LLM Optimizer to enhance the content, just like we’ve been talking about, to make it more accessible by LLMs and make it more accessible to search engines. And I think they, right before it got released at the end of December, utilized our auto optimize on edge capability so that over ten thousand pages had more content that was accessible to these AI agents and the LLM crawlers on the edge. And that was all made possible by this new capability that basically said, hey, you know what?
There’s a ton of content that’s missing that these agentic crawlers can’t see, so let’s make it available. And to the human eye, nothing changed at all. It’s all done on the edge. And instead, it recognizes, oh, you’re an AI agent.
Let me make sure that everything is available to you so that you can then pull that information into responses from prompt LLMs. And so it’s been really kind of cool to see how the Adobe.com team has been using Adobe Experience Cloud applications and agents in order to just simply make their content do more work and more accessible. So it’s really, really cool to see.
Oh, Adobe on Adobe. I mean, that’s it’s always one of the most interesting use cases to me because I think a lot of the you know, you guys do not only sell software. You have product. You have learning.
You have instructions. You have events. And, you know, some of those a lot of you know? I mean, I think I remember one time attend attending worldwide sales kickoff, and I think at that time, it was maybe pre Covid.
So probably 8 years ago, 7 years ago, there was a comment that made at that time only 20 companies in the software category were over a billion dollars in annual revenue. And I think at that time, you had two or three divisions that were meeting that number. You know, with with, your Adobe Experience Cloud and Adobe Creative Cloud. So Adobe.com,
you know, is is is just to your point, it’s I think it’s a lot bigger than folks understand. Right? I guess that’s the point I’m making. Yeah.
We we have an Adobe Brand Concierge deployment on Adobe.com somewhere. I’ll have to find it. But there it’s really kind of fun to see how that team uses our tech so that we make sure that before it’s in the hands of Adobe Experience Cloud customers, that it provides value, that it works the way it’s supposed to, that it’s integrated the way it’s supposed to be integrated, that it does all the things, that it works with Infobip for customers as well.
Brilliant. Brilliant. So alright. So I’ve got basically one more question and a little exercise we’ll do.
So, you know, the three of us work in industry, we obviously, in the case with Ervin, your career has been around building AI agents.
The Role of AI in Customer Experience
It’s what your academic background is and your pursuit as a professional. Eric and I kinda came from similar and, you know, just the the digital analytics space and kinda grew up in that space. So we’re data guys.
We’re technology folks, AI people. No. No doubt. Some folks may look at the three of us and say, well, you’re the part of the world that’s optimistic about AI.
And, know, obviously, I mean, I even said it, we ought to be using data and AI to make the world better. But you know, a lot of the people that, that listen to this podcast and that are customer experience professionals are trying to break inertia. They’re trying to build consensus. They’re trying to tell this story and they’re hearing the other side of it.
Oh, this could ruin our brand. This could destroy our reputation. This can hurt us. There’s ethical issues.
There’s governance issues. What do you say to the folks that think that AI may, you know, detract from what they’re doing? And and how do we help and enable them to get going and and using this? I mean, there’s the competitive imperative.
Well, your competitor’s doing it. Okay. Fine. But let’s let’s move beyond, you know, we’re gonna do this because everyone else is doing it.
We learned in first grade. That’s not a great answer to mom when she asked, well, why did you make that mistake? Right? We we because everybody else did it.
Well, no. We we have a vision, and I think this is, you know, the purpose of of what I am doing with my career is to apply innovation and help people to do that responsibly. But what guard guardrails, what ethical issues should they be considering as they move into 2026? I can I can take it?
So pretty much what kind of, like, the couple of things, like one thing is, like, always keeping human in the loop. So always keeping, like, human overseeing, like, what AI is doing in in in in most most, like, simple simplest way. So the second thing is, like, what kind of we are doing as a company is, like, rigorously monitoring rigorously, like, monitoring that the date that the communication and the data is, like, safe, and it’s kind of compliant based on the specific geographies. So if we as a global company is running communications like globally and pretty much this needs to be compliant.
And one insurance to in order to ensure this compliance, we have, like, a lot of set of tools in order to make make it compliant and trustworthy for the end end user. So and and, again, like, monitoring the data, testing the data, testing the solutions, double checking it, verifying it, and also, like, being aware what’s kind of with the human, like, keeping human in the loop and being aware what outputs of AI we are getting. Like, it’s really, really important, and it’s really important to use tools so to make it, like, efficient and and safe to this, like, in short terms. Yeah.
I mean, humans in the loop, you know, testing. I mean, sounds like, you know, most of the things we’ve been doing in our careers, probably even different stage architects, right, where we have a staging environment and so forth. So all the things sound so sounds like it’s very incremental. Eric, how about for you?
What do you think?
I think that regardless of what we do, consumers are already there or already getting there. I think I read recently that for 80%t of people that are I may have the numbers wrong. I’m not a data guy.
But something like 80% of consumers that are searching on Google or searching on chat platforms that have AI overviews, something like 40% of their searches are answered now just by the results of an AI overview. They are not clicking through. They’re just getting the information they need and
saying: Great, I’m good. So what I find most interesting about that, aside from the opportunity to influence those AI overviews and what have you, but what I find interesting about that is that consumers are there already. Consumers are, You go into any room and ask everyone to raise their hand if they’ve used ChatGPT or Perplexity this week, and you’re going to see a lot of hands. And so regardless of what some of the ramifications are, the consumers are there.
So we need to meet them there. They’re expecting these enhanced experiences, and we need to make sure that we’re providing them while also aligning with everything that Irvin just shared, whether it is legal ramifications, governance ramifications, data quality concerns. We need to make sure that those are built into our process because the consumer is there. And so we need to meet them there, or we simply get left behind.
At Adobe, from the creative side of the house, Adobe Firefly, in Adobe Creative Cloud, It’s been interesting to see how our organization is kind of thinking about all the different models that are out there, whether it be models that we’ve built at Adobe with Adobe Firefly. And we’ve got, I think, five different models that are Adobe Firefly on Firefly, Adobe.com that consumers can use to build images, to generate videos and sound effects and all these really cool things. But we also want to make sure that consumers or creatives have the opportunity to use partner models, models from Google, models from OpenAI, models from Runway and Luma AI and all of these different partnerships so that they can use what they want.
And so we’re basically saying, we wanna give you the tools to build what you need to because it’s it’s your end consumer experience that you’re building. It’s not ours. And so I think it’s important to realize that there’s a number of different opportunities out there. There’s a number of different capabilities that are out there and to test and learn the same way we’ve been testing and learning in our roles for the last twenty plus years.
You know, I think I kind of poo pooed the answer around, you know, hey, we’re going to do this because everyone else is doing it, which is a competitive imperative. And I love the way that you contrasted that with, hey, we’re going to do this because our customers are doing it. And that’s probably the best answer, right? Because it’s what not only do they expect, but if we think about it and do this well, it can be a competitive differentiator.
And I go back to that law that I was putting in that journal, the academic journal and naming about this concept of the law of relational, you know, friction where the organization has the best relationship should win, take the most market share. And because you’re in a position to have the best, to provide the best value for who they are and what folks need. So I appreciate that. Put the customer first and you won’t go wrong.
I think that’s the takeaway.
Upcoming Events and Product Announcements
don’t know if there’s any other burning questions or any, or, or, or other comments that one of you had for the other. Right? But it, what I’ll do is, is there anything that you wanted to promote or talk a little bit about, with any of your annual events, research, product feature releases, anything like that?
You mentioned at Adobe Summit, that I get to host the Summit Sneak Session. I’ll be hosting again in April this year in Las Vegas, which I’m super excited about. I can’t share any information about who the celebrity will be, but I can share that we have already started digging into the hundreds of brilliant ideas from across the ecosystem at Adobe of people coming up with new technologies, and there are some really, really cool ones. It’s been interesting to see how much better and more forward looking and more exciting some of the ideas are in this world of AI coding assistance whether it be Replit or Cursor or whatever folks are using for their vibe coding technology, it’s been kind of cool because it’s enabling more people that maybe previously would have just said, Oh, my idea is a PowerPoint slide.
Now their idea is some working technology. And so, Adobe Summit Sneaks this year is gonna be better than ever because the tech is cooler than ever too. Yeah. I am going back to an earlier point you made, I had seen a bank that indicated that if you’re not using AI twenty times a day, that you’re not meeting the corporate expectation, right, in the way that they expected their employees to be using the licenses that they bought, for all of their employees across, the globe.
That was a global bank that had that that that very specific guidance, which I thought was interesting. Ervin, anything, else that you wanted to to share? Yeah. From my end, pretty much, I would like to highlight that people stay tuned in because we’ll announce the new product from our end, which will kind of combine multiple of our products.
This will happen, like, during the April, and that’s it. Super. And I’m I’m super excited for the strengthening of the partnership also with Infobip and Adobe. We’re we’ve been working with some large organizations to really make sure that those integrations are happening.
We’ve already integrated the Adobe and Infobib platforms in Latin America, just given your strength with WhatsApp and Adobe strength across the customer experience industry at at large. We’re
Rapid Fire Word Association
Gonna do a little, a little thing here to wrap it up. I’m just gonna say a word. It’s kind of a single word.
And, we’ll go with you, Eric. We’ll start. And what I’m looking for you is for you to do a counterword. It could be, you know, the first word that comes to mind as a counter response.
So if I said Mickey, you might say mouse. But, you know, and and there’s gonna be four words, and we’ll just see how we go here. Okay. So let’s start off.
My first word is the word Adobe. Create. Ah, that’s good. Agentic AI. That’s two words, Paul.
Yeah. I broke the rule. And you can use a phrase if you want. I’ll open it up.
Yeah. Productivity.
Data. My love. Eric Matisoff. Enable. Okay. Fantastic. All right. And so, Ervin, let’s go to you.
Infobip. Connect. Agentic AI. Possibilities. Endless possibilities. That’s awesome. Data. Power. Yeah. I love love that fuel, like the oil.
Right? And then lastly, Ervin Jagatić.
Product. Product person. Yeah. Engineer, hands on this, you know, like the Eric, Irvin, thank you for helping us invent the digital future.
This has been Paul Lima with my guests, Eric Matisoff from Adobe and Ervin Jagatić from Infobip. We hope you’ve enjoyed today’s deep dive into the agentic era of AI as much as we did making it. If we’re helping you accelerate your business objectives, please let us know. Subscribe to the show, message me directly on LinkedIn, or email me at [email protected].
You’ve been listening to The Visionary’s Guide to the Digital Future. I’m your host, Paul Lima, and I’ll see you in the digital future.
