Independent Analysis: AI, SaaS, and Revenue Growth Strategy

the best outcome for Adobe out of today's keynote is not the one the press release leads with. The headline is Adobe CX Enterprise, an end-to-end agentic system for managing the customer lifecycle. The real prize is quieter and larger. Adobe wants to be the operating layer that stays ready no matter who shows up at your website: a human, an AI agent shopping on behalf of a human, or a large language model crawling to answer a question someone typed into ChatGPT.

Same platform. Same brand controls. Same measurement. The question of who is visiting stops mattering because the answer is always, doesn't matter, we are ready. That is a bigger business than martech ever was.

It reframes Adobe from a marketing cloud vendor into the infrastructure layer for brand presence across every surface where commerce happens. Figures throughout are vendor-supplied and unaudited.

The keynote line that tells you where Adobe is planting its flag

Anil Chakravarthy walked the analyst audience through the architecture this morning in the kind of language that sounds modest on stage and reads differently once you sit with it. His framing:

"Now, we provide a comprehensive solution for the entire customer engagement lifecycle, built natively on top of the Adobe Experience Platform. Acting in concert, Adobe Real-Time CDP, Customer Journey Analytics, and Journey Optimizer bring together all of your customer data sources in a composable manner to deliver comprehensive, real-time insights and orchestrate connected, personalized experiences."

Read that slowly. The word that matters is natively. Adobe is not saying Experience Platform is one option among several. It is saying Experience Platform is the foundation, and every agent, every workflow, every new surface Adobe ships is assembled on top of it. The three products Chakravarthy named, Real-Time Customer Data Platform, Customer Journey Analytics, and Journey Optimizer, are the load-bearing walls.

That is a commercial strategy dressed as an architecture statement. Any enterprise not already on Experience Platform is hearing, this week, that agentic customer experience lives inside Adobe's foundation or it lives somewhere else. There is no middle path Adobe is offering.

Time on site is falling because the website is no longer the destination

Website sessions are getting shorter across most enterprise categories. Most marketing teams are treating that as a content problem or a user experience problem. It is neither. It is a destination problem. Users are getting their answers from ChatGPT, Perplexity, Gemini, and Claude before they ever click through. The website has stopped being where the decision happens. It is where the transaction closes, if it closes there at all.

That breaks the last twenty years of digital marketing measurement. Sessions, bounce rate, pages per visit, time on page, all of it assumes the website is the destination. When the large language model is the destination and the website is the footnote, Adobe Analytics and every competing product needs a different unit of measurement.

Adobe Brand Visibility and Adobe LLM Optimizer are Adobe's bet on what replaces session metrics when the session itself is disappearing.

The Brand Concierge announcement sits in the same place. Adobe is openly saying customers will increasingly find and evaluate brands through AI interfaces rather than through search engines or company websites. Most competitors in this category have not yet admitted that shift publicly. Adobe has, and it has product tied to the admission.

The Semrush acquisition does more than Adobe is saying publicly

Adobe announced intent to acquire Semrush in December 2025. Today's CX Enterprise launch is the first event where that acquisition stops being a line item and starts being a product capability. Semrush spent years building the data pipelines that tell brands where they rank, who links to them, what competitors are doing, and, increasingly, how they show up in AI-generated answers.

Adobe buying that capability means Brand Visibility ships with real data behind it instead of a roadmap promise. That shortens the gap between announcement and demonstrable value in enterprise evaluations.

The integration risk is real. Semrush customers are search engine optimization practitioners and agencies. Adobe enterprise customers are marketing operations leaders, chief marketing officers, and the digital experience teams reporting into them. Those two buyers do not talk to each other inside most companies. Adobe has to bridge that gap or the acquisition becomes two products sold to two different buyers under one logo.

The business value if you strip out the product names

A chief financial officer reading today's announcement cares about three things.

Speed. A marketing campaign that takes six weeks and ten people today compresses into days and a smaller team. More campaigns per year. Faster response to what is working. Faster kill decisions on what is not.

Cost. Fewer people needed per campaign. Not zero people. Fewer. The marketing operations team either gets smaller or gets reallocated to higher-value work that software cannot yet do.

Consistency. Every piece of content runs through Brand Intelligence, which checks it against brand rules before it ships. Fewer off-brand emails. Fewer legal escalations. Fewer awkward social posts that get screenshotted and ridiculed.

That is the business value. Whether it materializes depends on two conditions Adobe cannot fully control. First, whether the agents work reliably enough that marketing teams trust them without babysitting every output. Second, whether the customer data feeding those agents is clean enough to produce useful results.

Most enterprises have messy customer data. That is the quiet problem behind every agentic AI announcement this year, including this one.

The dependency that sits outside Adobe's walls

Adobe has built measurement and optimization products for a world in which OpenAI, Anthropic, Google, and Microsoft expose clean APIs for brand visibility metrics inside their models.

That world does not exist yet.

Today, no one knows exactly why a given large language model surfaces one brand over another in a specific answer. The model providers have not yet decided whether to expose that data to brands, monetize access to it, or keep it opaque as a competitive moat. Three different futures are live at once, and Adobe's product strategy has to work across all three.

If the large language model providers open up, Adobe wins because the platform is ready. If they keep visibility metrics locked behind their own advertising products, Adobe ends up scraping and inferring, which is a fragile business. If they monetize API access, Adobe pays a tax to measure what its customers need to see, and that tax flows through to enterprise pricing.

Adobe cannot control which path the model providers choose. That is the single largest risk sitting inside this entire agentic pivot, and it lives outside Adobe's walls.

The partner list Adobe published today (Amazon Web Services, Anthropic, Google Cloud, IBM, Microsoft, NVIDIA, OpenAI) is a pressure signal as much as a capability signal. Adobe cannot win this alone. The partners know it. The buyers evaluating CX Enterprise should know it too.

The history that tells you how long this takes

Google took roughly fifteen years to ship genuinely useful webmaster tools. Brands spent most of those fifteen years guessing. The model providers today are where Google was around 2005: indispensable, opaque, and not yet obligated to share anything about how their systems rank brands.

Adobe is betting that history compresses this time. Competitive pressure, enterprise procurement weight, and regulatory attention all push in that direction. None of them guarantee it.

I walked into Summit this week already holding a thesis about what Adobe is becoming. The sponsor list I wrote about last week pointed toward infrastructure layer ambition. The fiscal 2025 financials pointed toward workflow accumulation. The Workday case study pointed toward joint chief information officer and chief marketing officer governance as the adoption prerequisite.

Today's announcement fits that pattern. It also raises the stakes of my thesis, because the one dependency I did not fully account for sits with the large language model providers, not with Adobe. That is where I will be spending my reporting energy for the rest of the week.

If your renewal cycle opens in the next six months, the question is not whether Adobe's demo works. It will.

The question is whether the foundation holds.