Just been digging through Google's latest data dump on how people are using AI Mode in the US. Some useful directional stuff, but let's not kid ourselves - this is Google framing the narrative to suit its own ecosystem. So with that caveat, here's what stood out.
Searches are longer and more conversational - average query is about three times the length of a traditional search. That means keyword research now has to cover prompts, tasks, constraints, scenarios. Fun times for anyone who thought SEO was just about stuffing keywords.
Follow-up behaviour is growing fast - up 40%+ per month. Which means you can't just optimise for the first query anymore. A brand can be mentioned, dropped, compared or misrepresented across a whole journey. The unit of analysis is the journey, not the query. Pain in the arse for measurement, but that's where we're headed.
AI Mode isn't just for discovery - it's for decisions. 'Which' queries grew 40% faster than overall AI Mode queries. People want price, location, colour, brand, availability, size, material, style, type, quality. So if you're in ecommerce, it's not about piling on more product content - it's about accurate, complete, fresh product data across feeds, variants, reviews and attributes. No shortcuts.
Local and availability intent is screamingly obvious. Follow-up queries include 'near me', 'in stock', 'replacement parts'. AI needs to understand location, inventory, services. If your data is a mess, you're invisible.
And here's the kicker: planning queries grew 80% faster than overall AI Mode queries. AI Mode is becoming a task layer, not just an answer layer. The opportunity is to get into plans, shortlists, comparisons - not just rank for a single query.
My takeaway? Don't confuse Google's usage narrative with independent performance data. Yes, there's a behavioural shift, but Google has every incentive to make it look rosy. For us, the practical step isn't to ditch SEO fundamentals - it's to expand how we research, optimise and measure. From keywords to prompts and tasks. From rankings to presence and citation accuracy. From single queries to follow-up journeys. From content-only to entity, product, local and feed-level readiness. And from obsessive click-counting to a more nuanced view of visibility and influence.
Relying on traditional rankings and clicks to measure search visibility is getting riskier by the day.
