Google answers for you now. How a site still earns the click

Not long ago, a search for «how much does a landing page cost» showed ten links and people clicked a few of them. Now there is often a ready-made answer at the top, assembled by AI from several sites. People read it - and go nowhere. The click that sites competed for over years simply does not happen.

The change is real, and pretending it is not is a bad strategy. But there is no need to panic either. What is needed is to understand which clicks are disappearing and which are staying.

Which queries AI takes

The ones with a short, general answer. If your page can be summarised in three sentences without losing anything, AI will summarise it - and nobody will need to click.

• «What is...» and definitions of terms
• General comparisons with nothing specific in them
• Simple instructions of a few steps
• Questions whose answer is the same for everyone

Which clicks stay

The ones where a person needs something that cannot be summarised. Specific experience. Numbers from real projects. A tool they can use. A decision that depends on their situation. A conversation with an actual person.

And queries with an intent to do something: order, compare specific companies, book. There the person is looking not for an answer but for someone who will do the job. AI can suggest where to go, but they will still go to a website.

AI is good at retelling what everyone already knows. It cannot retell your experience - it can only quote it.

How to end up inside the answer

AI answers cite their sources. Being a source is a new way of being visible, and it works by familiar rules - just stricter ones.

• The answer to the question in the very first paragraph, not after three screens of preamble
• Clear headings, with one idea under each
• Text in the HTML rather than drawn in by JavaScript: many AI systems do not run scripts
• schema.org markup: who you are, what you do, where to find you
• Facts that can be checked, with sources for them

And above all - something of your own. If your page says what a hundred others say, AI has no reason at all to quote you in particular. Your own data, your own experience, your own position - that is what turns a page into a source rather than one of many.

What to change in your pages now

Reread your pages and ask honestly: what is here that is not in the first five results? If the answer is nothing, the page either gets something added or gets ready to lose traffic.

Service and case study pages suffer least: the intent there is to order, and people will come. The articles most at risk are the ones rewriting general advice. Commissioning that kind of text now means paying for something only AI will read.

That is why on this blog we write about what we see in our own work: specific bugs, specific fixes, things we had to measure and redo. It is the one kind of text AI cannot replace. It can only link to it.