Google recently announced more than forty new innovations across its advertising products. Conversational AI inside Search Ads. Intent scoring. Journey-aware bidding. Predictive attribution. AI-generated creative at scale. The framing was straightforward: Google is solving the gap between generating leads and generating good leads.
It is worth reading that announcement alongside a different question: who benefits most from these changes and what does an organisation need to have in place before any of this works as advertised?
What the ads actually do now
The most structurally significant announcement is Business Agent for leads — conversational AI interactions built directly into Search Ads. The concept: a user searches for a product or service, an ad opens a dialogue rather than directing to a landing page and the prospective customer can ask questions about availability, pricing or suitability before clicking anywhere.
That is a meaningful change to how PPC works. The landing page has always been where qualification happened — the page either answered the question or it did not and the bounce rate reflected the outcome. Moving qualification into the ad itself means the conversation happens inside Google's interface, powered by content Google has read from your website, before the visit occurs.
Which makes the quality and accuracy of your website content more important than it has ever been — not less.
The content problem that AI cannot solve
There is a persistent assumption in coverage of AI-powered advertising that the technology compensates for weak inputs. It does not. AI systems optimise based on the data they receive. If a campaign tracks form fills Google optimises toward form fills regardless of whether those leads become customers. If a website describes products vaguely the conversational ad will describe them vaguely.
The fundamental requirement has not changed: a website needs to clearly accurately and specifically reflect what it sells and who it sells it to. Not for SEO in the traditional keyword sense — for the AI systems that are now reading that content and generating responses on your behalf. The organisation that has invested in product-specific customer-aligned content will get a more accurate and useful conversational ad than the one that has not. The gap between them will be visible in the conversation quality before it is visible in any reporting dashboard.
The announcement's emphasis on offline conversion imports and CRM connectivity is significant — and demanding. To make AI Max and journey-aware bidding perform well you need to feed richer signals back into Google: actual sales data not just form submissions, revenue quality not just volume, customer lifetime value not just acquisition cost. That requires integration between your CRM and your ad account that most organisations have not built. The announcement presents it as the solution without acknowledging that for the majority of businesses it is still the problem.
The impression share history
For years Google's automated recommendations and the agency ecosystem built around its partner programme consistently pushed brands toward higher impression share. The message was consistent: you are leaving commercial opportunity on the table. Capture every eligible query. The brands that followed that signal paid for dominant presence in the search results.
AI Overviews now occupy the top of those same results. Conversational ad formats sit below them. The organic result that impression share was supposedly protecting is further down the page than it has ever been. The inventory that brands paid to dominate has been restructured around Google's AI products — and the brands that spent that budget are now being asked to invest in the next layer of the same stack.
Whether that outcome was deliberate strategy or convenient alignment is not knowable from the outside. What is observable is that the advice to maximise impression share built the inventory base that AI Overviews and conversational ads now monetise. The brands paid. Google captured the data. The landscape shifted anyway.
The measurement problem
The announcement introduces predictive attribution — connecting ad exposure to downstream behaviours that may occur months later. Google wants to estimate what will happen next not just measure what happened previously. Features like Attributed Branded Searches and qualified future conversions are designed to close the gap between awareness and conversion in ways that traditional attribution cannot.
The challenge is that predictive models are not independently auditable. As more of the user journey happens inside Google's interface — conversational ads, AI Overviews, AI-assisted browsing — the data that describes that journey belongs to Google. Attribution becomes a story Google tells about what happened based on signals that advertisers cannot access in full.
This is not a new concern. It has existed since Smart Campaigns, since Performance Max, since every automated system that optimised for outcomes while obscuring the decisions behind them. What is new is the scale and the depth of the interface layer. If a user's entire pre-purchase conversation happens inside a Google ad, the advertiser's visibility into that conversation is entirely dependent on what Google chooses to report.
The organisations that will navigate this best are not the ones with the largest budgets or the most sophisticated AI tooling. They are the ones that understand which signals actually matter to their business, that have built the offline data connections to feed those signals back and that maintain enough independent measurement — their own analytics, their own CRM data, their own conversion tracking — to validate what Google's dashboards are telling them.
What this means in practice
The announcements are real and the direction is consistent. Conversational ads will change how qualification works. Predictive attribution will change how performance is reported. AI-generated creative will lower production costs and raise the baseline of what's accessible.
None of it changes the underlying requirements. A website that accurately reflects its products and its customers will produce better AI conversations than one that does not. An organisation with clean offline conversion data will get better automated optimisation than one without. A business that maintains independent measurement alongside Google's reporting will have a more complete picture of what is actually working.
The technology is advancing. The foundations that make the technology useful have not changed. Content that means something. Data that is trustworthy. Measurement that does not depend entirely on the platform being measured.
Ranking is still not data. It is a perspective. Attribution increasingly is too.