The customer journey now starts before anyone searches on Google
You see zero sessions, but the decision has already been made
A buyer asks ChatGPT for three suppliers for a new CRM project. He reads the summary, maybe clicks on a few sources, and puts two names on his shortlist. Your company doesn't come up. In GA4 you see nothing. No session. No bounce. No form submission. And yet you've just been eliminated.
That is the blind spot many Google Ads accounts are running into right now. We measure the click as if it marks the beginning of the customer journey, while that click is increasingly the visible end of a much longer orientation phase. Not exactly a problem you solve by adding three more keywords…
The new pre-awareness layer
Exposure Ninja describes in the source video an extra layer above the funnel: the phase in which customers use AI tools to understand their options, compare suppliers, and sharpen their questions. I'd put it more plainly: people only search on Google after they've already formed a picture. Sometimes even after they already know who they do and don't take seriously.
SparkToro and Datos already showed in their zero-click research that a large share of Google searches end without a click to the open web. The exact percentages need to be checked per market and per year, but the pattern is clear enough: visibility is no longer the same as traffic. That's inconvenient for dashboards, but especially for business owners who assume that no drop in leads means everything is fine.
| Old assumption | What happens more often now | What you need to check |
|---|---|---|
| The search query starts in Google | The question starts in ChatGPT, Gemini, Perplexity, or an industry forum | Is your brand mentioned in explanatory articles, reviews, and sources that AI can summarise |
| A click is the first measurable signal | The click only comes after orientation without a website visit | Which search queries lead to branded searches or direct sessions |
| SEO and Ads are separate channels | AI summarises all visible sources | Whether your message is consistent and concrete everywhere |
| Only conversions count | Authority determines who makes the shortlist | Reviews, case studies, authorship, and external mentions |
Why last-click ROAS falls short here
Last-click ROAS is reassuringly tidy. You put money into Google Ads, revenue or a lead comes out, and the dashboard tells you whether it's working. The only problem is that this calculation works well mainly when the ad actually tipped the scales. In longer B2B cycles that's often not the case. There the ad is sometimes a reminder, sometimes a confirmation, and sometimes just the easiest thing to measure.
Kahneman described with the peak-end rule that people don't remember experiences as a neat average. They remember peaks and the ending. Something similar applies to marketing. The click you optimise for is often the end of an earlier series of impressions: a LinkedIn post, a review, an AI summary, a podcast, a case study, or a colleague who mentioned your name. If you only measure the last click, you make that final step look bigger than it really is.
- Split brand and non-brand traffic in your reporting so you can see whether Ads is capturing demand or creating it.
- Look in GA4 at assisted journeys, direct sessions, and returning visitors around lead moments.
- Use Search Console to see whether branded searches grow after content or PR activity.
- Label campaigns not only by channel but also by role in the customer journey: creating demand, deepening demand, or converting demand.
- Ask in sales conversations how someone first heard your name. That information is rarely captured neatly in your ad platform.
What AI tools need in order to mention you
AI tools don't mention you because you want to be mentioned. They build an answer based on sources, patterns, authority, and clear language. That means vague proposition pages, empty case studies, and thin blog posts help less than many business owners hope. 'We deliver total solutions for growth' is already foggy for a human reader. For an AI summary it's even less useful.
Rumelt wrote that strategy begins with diagnosis. The same applies here. First it must be clear what you want to be known for. Not 'online marketing agency', because there are plenty of those. Instead: Google Ads for lead generation with longer sales cycles. Or: Funnel Adviseur for companies that want to run Ads, GA4, automation, and follow-up as one integrated system. That is Direction and Focus. Only then do you Scale.
- Create a sharp page for each service that names the problem without marketing fog.
- Publish case studies with context: target audience, starting situation, intervention, measurement point, and limitations.
- Make sure reviews use the same language your ideal customer uses.
- Connect articles to each other with logical internal links, not just as standalone blog posts.
- Have external sources, partners, and industry pages name your specialism concretely.
“Strategy is not a list of actions. It is choosing which problem you solve first.”
The practical audit for this month
You don't need to launch a new 'AI search project' right away. In fact, that will probably just turn into another list of small tasks. Start smaller. Take five questions your ideal customer asks before requesting a quote. Put those questions to a few AI tools, check which sources come up, and note whether your brand, case studies, or explanations are logically present anywhere.
Then place those findings next to Google Ads. Which campaigns are mainly capturing existing demand? Which campaigns are trying to create new demand? Where is content missing that would make the ad credible? And where does sales keep encountering the same hesitation? This isn't a perfect science. But it's better than optimising for a slightly lower CPC while the real decision has already been made above your dashboard.
The companies that do this well don't treat Google Ads as a standalone cash machine. They use Ads as a measurable layer within a broader customer journey. That requires less dashboard religion and more clear-headed architecture. Or as I often tell clients: look beyond the end of your nose. Especially when your customer is already doing exactly that.
