Moment one: the research that used to be several searches
A customer with a problem used to run four or five searches — what is wrong, what does it cost, what are my options, who does it. Increasingly that whole sequence happens in one conversation with an assistant.
The consequence: they arrive at the provider-selection stage already informed, with a rough price expectation and a shortlist of options. Your ad meets a more educated buyer than it did three years ago.
What that changes about your messaging
Educational content in your ads is now largely wasted — they already know. What works is specificity: your price range, your availability, your credentials, your area, and the reason to choose you over the two other names they already have.
It also raises the cost of vagueness. A buyer comparing three specific propositions will skip the one that will not state anything concrete.
Moment two: the recommendation itself
Some portion of buyers now ask an assistant directly who to call. That answer is assembled from verifiable data — profiles, reviews, structured data, mentions.
It is not yet the dominant path for local commerce, and anyone claiming precise numbers is guessing. But it is growing, it is invisible in your analytics, and the work to be included is work you should be doing anyway.
What has not changed at all
Speed of response, answering the phone, showing up when you said, and having recent reviews. Every technology shift in the last decade has left those exactly where they were: decisive.
The businesses that win the AI-informed buyer are mostly the ones that were already easy to verify and easy to reach.
The practical list
State prices or ranges publicly. Keep profile and structured data accurate. Generate steady recent reviews. Respond in minutes rather than days. And stop paying to educate buyers who arrive educated.