We've built 100+ trade websites in two years. We've also tracked conversion rates for all of them. The strongest predictor of which sites convert well in months 2-6 isn't the design, the SEO, or the budget. It's the answers to seven specific intake questions.
This is the scorecard.
Question 1: "What's the most common job you turn up to?"
What we're testing: specificity. Founders who answer "all sorts" or "general plumbing" tend to build sites that convert poorly. Founders who answer "leaking combi boilers, mostly Worcester Bosch and Vaillant" tend to build sites that convert well.
Why: specific answers produce specific copy, which matches specific Google searches. Generic answers produce generic copy, which competes against everyone.
Question 2: "Where geographically do most of your jobs come from?"
What we're testing: whether the client knows their patch. "All of Newcastle" is a weak answer. "Mostly Heaton, Jesmond, Gosforth and the city centre" is a strong one.
Why: specific suburb-level knowledge lets us build suburb pages that actually rank. See the service-area page rationale.
Question 3: "What do you do differently from the next van on the next street?"
What we're testing: commercial self-awareness. The killer answer is something specific and unfakeable: "I take photos of the job and send them to the customer the same day", "I quote in writing within 24 hours, always", "I offer a 12-year guarantee that none of the others do".
Why: the answer becomes the headline. Without a real differentiator, we're trying to win Google against 50 identical websites - which is a losing game.
Common bad answer: "good service, fair prices". Everyone says that. It's not a USP.
Question 4: "How much of your work comes from existing customers and referrals?"
What we're testing: whether they're starting from zero or from a base of social proof.
Why: a founder with 200 happy past customers has a review-velocity head start. A founder with 5 has to build the trust signals more deliberately.
Question 5: "What does your perfect customer look like?"
What we're testing: commercial focus. The best answer names a customer type ("homeowners 40-65, owner-occupiers, mid-renovation"). The worst answer is "anyone with money".
Why: the perfect-customer description shapes the tone of the entire site. "Anyone" leads to copy that appeals to no one.
Question 6: "Can you commit to 2 calls in the first 7 days?"
What we're testing: engagement. Subscription clients who can't make two calls in week one will be ghost clients in month three. (See the field note on firing clients.)
Question 7: "If a customer rings you tomorrow, what's the first thing you ask them?"
What we're testing: qualification instinct. The killer answer reveals the founder's actual sales process: "What's the postcode? Is it a leak you can see, or just lower pressure than usual?". A weak answer is "I'd just say I can come round".
Why: this answer goes directly into the FAQ section and the contact form's free-text placeholder. It pre-qualifies the customer before they fill anything in. Highest-leverage 30 seconds in the entire intake.
The scorecard
We score each answer 1-3. Across our last 60 builds, sites scoring 18+ converted at 7.4% (form fills / sessions). Sites scoring 14-17 converted at 4.8%. Sites scoring under 14 converted at 2.1%.
The bottom-quartile sites are the ones we now coach hardest at intake. If we can move a 12 to a 16 with a 30-minute coaching call, the conversion delta is worth more to the client than the rest of the build combined.
What this means for UK local web design
The vast majority of design and SEO work is hostage to the answers to these seven questions. A perfect site for a founder who can't articulate a USP will convert worse than a mediocre site for a founder who can. If you're commissioning a build from anyone - Built Local or otherwise - and they don't ask these questions, ask why.
If you want to score yourself, our intake form asks all seven. Or read the pillar guide to local web design for the wider context on what makes local web design for tradespeople actually convert.