We almost gave every cold lead a free, full sample website. That would have been a mistake, and it's an easy one to make.

The plan was tidy: a stranger messages us, the bot works out what they need, and we build them a sample site — free, no call, here it is. The sample is the pitch. It's also, by a wide margin, the most expensive thing we do: minutes of model work, a whole multi-page site assembled from scratch. We were one approval away from running that for every lead a marketing campaign sends us. Then we changed it.

The expensive thing shouldn't be the first thing

A cold lead from an ad is a maybe. Some are real businesses ready to move; plenty are curious, mistaken, or gone after one message. If your funnel auto-delivers your most expensive output to all of them, you're spending your best work on the people least likely to want it — and you find out who was real only after you've paid for everyone.

The instinct to lead with your strongest offer is right. The instinct to make that offer the first, automatic step is the trap. You want the strong offer to land on someone who's leaned in even slightly, not on a number that messaged "how much?" and vanished.

Lead with the artifact that's cheap for you and valuable to them

So we put a cheaper thing first: an audit. Before we build anything, we look at the lead's existing web presence and produce a short report — what's working, what's missing, what we'd change. It costs us a fraction of a full build, and it's genuinely useful to the business owner on its own. They can read it, ignore us, and still have gotten something real.

Then — and only then — do we offer to build the sample. The flow went from:

qualify  →  build the full sample  →  "here's your site"

to:

qualify  →  run the audit  →  send it  →  "want the free sample?"  →  build

The build still happens. It just sits behind one more door, and the door is cheap to open.

The reply is the signal

Here's the part that does double duty. When we send the audit and ask whether they want the sample, the reply is the qualification. A lead who answers — even to say "just show me" — is real enough to build for. A lead who reads the audit and goes quiet told us something too, and it cost us an audit instead of a whole website.

We didn't add a budget question or an "are you serious" form. The audit plus a simple follow-up does the qualifying on its own, because a reply is the clearest signal we have that someone wants the thing. We let the answer decide what we spend next.

Two details matter for it to feel human, not like a gate:

  • "Skip" is a yes. If they don't want to answer questions, that's fine — we build from what we have. The point isn't to extract more data; it's to confirm someone's still there.
  • Silence isn't a no forever. A lead who goes quiet goes into a gentle follow-up, not a void. They just don't get the expensive build until they come back.

The audit makes the build better, too

There's a quieter benefit. By the time we build, we've already studied the business and surfaced the gaps — the things the audit couldn't find. So the questions we ask are specific ("you don't have a services page anywhere — what do you actually offer?") instead of generic. The audit isn't just a filter — it's the brief. The sample we eventually build is better because the audit came first.

What we'd tell you

If you hand every lead your most expensive output automatically, look for a cheaper thing that's still genuinely valuable on its own — a report, a teardown, a sample of the input rather than the whole output. Put that first. Let the reply to it decide whether you invest in the big thing.

We don't have the conversion numbers yet — we just shipped this, and we'll know more once real campaign traffic runs through it. We think the logic is sound, and now we'll let the data check us: give a small valuable thing, watch who leans in, and spend your best work on them.