Advertising Into a Decision in Progress
When your ad lands next to a live decision someone is working out with ChatGPT, the job of the creative changes. It has to be the next useful step, not an interruption and not a slogan. Break the conversation on the click and you paid to frustrate someone at the worst possible moment.
Picture someone who just spent 12 minutes with ChatGPT working out the difference between FHA and conventional for a first home. Your ad shows up under the answer. They click.
And the page says: “Welcome to Smith Financial. Integrity. Service. Experience.”
You just broke the conversation. The person was mid-thought, you offered to continue it, and then you dropped them into a brochure. Advertising next to a decision in progress is a real opportunity, and I think it punishes the usual habits harder than any channel I have used.
The context sets a higher bar
When your ad lands next to a live decision, the job of the creative changes. The person is already reasoning. They do not need to be stopped, because they were not scrolling, and they do not need a slogan, because they are in the middle of a thought.
Generic branding says nothing to someone in that state. “Your dream home starts here” is noise to a person who, 90 seconds ago, was asking an AI about down-payment assistance. Specificity is the price of entry here, not a nice-to-have. The ad has to read like the next useful step in the thing they were already doing, or it reads like every other ad and gets ignored by someone who is, for once, actually paying attention.
Write the next step, not a billboard
The weak version is the one that could belong to anyone. “Discover luxury, reimagined.” It names no situation and answers no question the person was asking. It is a logo with a tagline.
The better version names the moment. “Relocating for work? Compare neighborhoods, homes under $700K, and a commute under 30 minutes with a local advisor.” It says what the person is trying to do and offers an obvious next move. The shape that tends to work is plain: name the scenario, add one honest qualifier, offer a low-friction next step. That is not clever copywriting. It is continuing the person’s train of thought instead of interrupting it with a brand. These are illustrations, not a formula to paste, because the whole point is that the ad matches a specific conversation rather than a template.
Do not break the conversation on the click
The click is the fragile moment. If the ad continued the thought and the landing page abandons it, you spent money to frustrate someone at the exact instant they leaned in.
The person who clicked an FHA-versus-conventional ad should land on a page about FHA versus conventional. Not a homepage full of adjectives about the company. This is old message-match advice, and the bar is higher here, because this visitor did not arrive from a keyword typed into Google. They arrived from a conversation that had already done real work for them. The page either picks that work up where they left it, or it throws all of it away and asks them to start over. Most homepages throw it away.
The line between relevant and creepy
There is a failure mode on the other side of generic, and it is worse. Too much relevance.
If the ad mirrors the conversation too precisely, it stops feeling helpful and starts feeling like someone was reading over the person’s shoulder. “Still worried about your 2019 tax return and that $18,000 in write-offs?” is not clever. It is unsettling. And the irony is you never actually got those details. You do not see the conversation. The platform matches your ad to the moment; it does not hand you the person’s tax return. An ad that guesses too precisely just makes it look like you did, and that impression alone makes the whole surface feel like surveillance instead of help.
The move is to be relevant to the decision without appearing to know the person. Speak to the situation a lot of people in that conversation share, not to the specific details this one just typed. “Self-employed buyers often qualify for more than they expect once the right income documents are on the table” speaks to the moment without claiming to have read their mind. It is the difference between a knowledgeable advisor who understands your situation and a stranger who somehow has your file. In housing, where the personal details are exactly the things you are not allowed to target on anyway, staying on the safe side of that line is not just good taste. It keeps you out of trouble.
Measure what happens after the click
This traffic can behave differently once it is in the funnel, so watch the right thing. A cheap $12 click that never becomes a real conversation is not a win, and I have watched cost-per-click flatter a campaign that produced nothing worth calling.
What matters is what reaches a real stage in your CRM, and how fast it gets there. Conversational traffic might enter the funnel further along, because the assistant already did some of the qualifying before the click. Or it might enter rawer and need more work than a search lead. You will not know which until you track it all the way to an actual outcome instead of stopping at the click and admiring the cost per click. Judge it by pipeline, not by traffic.
The edge is learning, not arriving
The temptation with any new channel is to plant a flag and call it a strategy. That is not the edge. Arriving first on a young ad platform mostly means paying to discover its bugs before anyone else.
The edge is learning first. Run one small, controlled test. Keep the budget honest. You will not see the conversations with the AI yourself, so pay attention to which scenarios and ad groups actually turn into clients, tracked through to a real CRM stage, and which ones just look busy in the dashboard. Do that, and by the time the channel is crowded and expensive, you already know which decisions-in-progress are worth showing up for and which are a waste. That knowledge is the real asset, and it is the one thing a competitor who shows up late cannot copy from you.