Google I/O 2026 was not, for the most part, a financial services event. But what was announced there, and what has been quietly building across the major artificial intelligence (AI) platforms for the past 18 months, has profound implications for how banks, fintech vendors, and all firms, including Datos Insights, reach customers, build awareness, and maintain relationships.
The shift is structural and moving faster than most players in this industry appear to appreciate: The way most users interact with the internet is fundamentally changing.
For two decades, the dominant paradigm was simple. A user had a question, typed it into a search engine, got a ranked list of links, and clicked through to websites. That model, and the entire search engine optimization (SEO) marketing infrastructure built around it, is being upended in real time.
In its place is an AI-first, synthesis-driven model that generates answers directly, in which agents act autonomously on behalf of users, and the traditional top-of-funnel web journey, the one that drives most digital acquisition in financial services, is rapidly becoming irrelevant. The implications run well beyond SEO. They cut into customer acquisition, relationship maintenance, product discoverability, and even the basic mechanics of how FIs and vendors communicate their value or how end users interact with their FIs.
From Blue Links to Synthesized Answers
The traditional search funnel was, at its core, a traffic model. You invested in SEO, you ranked for the right terms, prospects clicked through to your landing page, and your conversion funnel took over from there. Financial services firms, banks, insurers, card issuers, and fintech companies have spent enormous resources optimizing for this model. Much of what passes for digital marketing strategy in this industry is, at bottom, a well-tuned version of the same process that worked in 2008.
That process is breaking down. AI search interfaces, such as Google’s AI Overviews, its new Gemini-powered synthesis layer announced at I/O 2026, and equivalents from Microsoft, Perplexity, and others, no longer return a list of links by default. They return answers. A user asking ‘What’s the best business checking account for a company that does a lot of cross-border payments?’ receives a synthesized recommendation, not 10 links to bank landing pages. The bank whose page would have ranked first gets no clicks. It may not even get a mention.
This is way beyond a tweak to keyword strategy. It is the elimination of the mechanism that keyword strategy was designed to exploit. The metrics that have defined digital marketing success for a generation—organic traffic, click-through rate, time on landing page—are becoming structurally less meaningful. No amount of optimization within the old paradigm will reverse this.
The Rise of the Agent: Your Customer’s New Research Department
The more consequential shift, however, is not just that search returns synthesized answers; it is that agents, operating autonomously on behalf of users, are increasingly generating those answers. Google’s announcement of persistent, proactive agents that monitor the web and real-time data, and Microsoft’s equivalent Copilot expansions, represent a move from reactive search to continuous background research. Small-business owners will not need to search for the best sweep account rate; their agents will already know and surface the recommendation when relevant.
For retail banking customers, this dynamic will reshape product discovery entirely. Commercial and small-to-midsize business (SMB) clients already rely heavily on their bankers and accountants for product guidance. The agent effectively becomes an additional trusted intermediary, one that operates 24/7 and has access to the client’s full financial context. The bank that fails to make its product data legible to these agents does not just lose a search ranking; it disappears entirely from the consideration set.
This is already untenable for FIs that have not begun to think about their data architecture in this way. (Most have not.) The scramble to produce AI-readable, structured product data, such as clean pricing, explicit eligibility criteria, and deterministic feature comparisons, is just beginning, and the window to get ahead of it is narrowing.
The Conversion Funnel Gets an AI Intermediary
The disruption extends beyond discovery into the conversion funnel itself. Agentic AI is increasingly capable of taking action on behalf of users—recommending products, initiating inquiries, comparing fee structures across multiple providers in real time, and, as Google has begun to demonstrate, completing a booking or initiating an application. The AI is no longer just a research tool as we enter a world of agentic commerce. It is becoming the first point of contact in what would previously have been a human-driven customer journey.
For FIs, this means the conversion funnel needs to be designed for two audiences simultaneously: the human customer who may eventually interact with it, and the machine agent that will increasingly be their proxy in the early stages. An onboarding flow optimized purely for human user experiences will be opaque to agents. An API that is not exposed to authenticated third-party agents creates a wall that no amount of web design will overcome. Banks that have not invested in API-first infrastructure are now facing consequences that extend well beyond open-banking compliance.
The uncomfortable reality is that many FIs’ digital onboarding journeys were not designed to be machine-readable because there was no machine reading them. That is changing very quickly.
Transactional Data as a Competitive Surface
There is a subtler implication among the more dramatic shifts around search and agents, but it may prove equally consequential for FIs over time. Google’s expansion of “Personal Intelligence,” i.e., integrating AI responses with users’ Gmail, Calendar, and document data, means that the communications FIs send to their customers are becoming inputs into an AI’s reasoning process. Account statements, transaction receipts, fee notifications, and billing descriptors are now potentially searchable context that a user’s agent will draw on when answering questions about their financial situation.
If a bank’s email notifications are inconsistent, its billing descriptors are cryptic, or its transaction categorization is a mess, the user’s agent will reflect that mess back to them. A question as simple as ‘How much did my business spend on software last quarter?’ should be trivially answerable from a bank’s transaction data. For many SMB clients, it currently is not—not because the data does not exist, but because their FIs never designed formatting and labeling conventions with machine legibility in mind. That is now a growing competitive disadvantage.
What This Means for Everyone in This Industry, Including Datos Insights
It would be easy to frame these shifts as a problem primarily for bank marketing departments. That framing is too narrow. The changes in how users discover, evaluate, and engage with financial products affect every part of the value chain: FIs, of course, but also the vendors selling to them and the analyst firms, such as Datos Insights, that produce content for these audiences.
For FI vendors, the same logic applies to their go-to-market approaches. A vendor whose product capabilities are buried in marketing prose, inaccessible to the AI agents that enterprise procurement teams will increasingly rely on, faces the same discoverability problem as a bank with poor schema markup. ‘We offer a flexible, scalable, best-in-class solution’ tells a synthesis engine (and most humans) nothing. Clean, deterministic specifications, integration protocols, pricing tiers, API documentation, and client references are the new product marketing.
For analyst firms, the implications are uncomfortable but worth stating plainly. The traditional research model—produce a report, publish it behind a paywall or with gated access, drive traffic via search—is vulnerable to the same structural shift. If AI synthesis engines are drawing on publicly available content to answer questions that practitioners previously turned to analyst reports to resolve, the value proposition of research needs to evolve. That does not mean research becomes less valuable; it means the access model, the format, and the way findings are distributed all need rethinking.
None of this is speculative. The traffic data is already moving. The question is whether players across this industry are willing to look squarely at two decades of marketing best practice, acknowledge that much of it is becoming obsolete, and act accordingly. As the great Canadian media theorist Marshall McLuhan famously put it, “The medium is the message.” In other words, while we shape our tools, our tools shape us. All corners of the FI space need to rethink how they communicate with their prospects, clients, and partners, as well as how this will impact their own operations and structural approaches in the long term.
