The short version: AI has made building an interactive calculator or quiz for lead generation cheap and fast, mine took a weekend and cost under £40 in tools, but it hasn’t made the traffic problem disappear, and that’s the part nobody selling you an “AI lead gen tool” wants to say out loud. I built one, ran it for six weeks, and I’m giving you the real numbers below.
Why I built it in the first place
I keep getting the same message in my inbox, on LinkedIn, and after every talk I give: “Do I need an AI consultant or can I just figure this out myself?” I’d been answering it one person at a time for months, which is a rubbish use of anyone’s evening. So one Friday in late 2025 I decided to turn that question into something people could answer for themselves in ninety seconds, and I’d capture their email address on the way through.
I called it the AI Readiness Score. Ten questions, mostly about how much time they spend on repetitive admin, whether they’ve got customer data sitting in spreadsheets nobody looks at, and whether they’ve already tried and abandoned a tool like Zapier or a chatbot plugin. At the end you get a score out of 100 and a short paragraph telling you whether you’re ready to bring in outside help or whether you’d waste money doing it right now.
I’d read the piece on how interactive calculators convert visitors years ago and always meant to build one. AI is what finally made it worth the weekend rather than a three-week dev job.
What it took to build
Here’s the honest build process, step by step, because most posts about this skip straight to “and then I had a beautiful funnel”:
- Saturday morning: I used ChatGPT to draft the ten questions and, more importantly, the scoring logic behind them. This is the bit that used to need a developer. I described the outcome bands I wanted (0-40 “not ready”, 41-70 “ready with support”, 71-100 “ready to go it alone”) and asked it to write the weighting rules for each answer. Took about 40 minutes of back and forth.
- Saturday afternoon: I built the actual quiz in Outgrow, which has a paid plan around £30 to £40 a month if you commit annually. Involve.me is a cheaper alternative if you want to test the idea first, they have a free tier that’s fine for a low-volume launch. Neither of these needed a line of code.
- Saturday evening: I fed the ten questions and my scoring logic into the tool, wrote the three result pages, and connected the email capture to my Mailchimp list. If you’ve never looked at how a proper welcome sequence should work after someone hands over their email, the breakdown of Mailchimp’s own marketing strategy is worth a read, because the calculator is only half the job, the follow-up emails are the other half.
- Sunday: I tested it on four people who had no context (my sister, two clients, my VA), fixed the two questions that confused everyone, and published it.
Total spend: £37 for the Outgrow plan, plus roughly three hours of my own time on the questions and copy that ChatGPT drafted but I rewrote, because AI copy still sounds like AI copy if you don’t touch it.
The six-week numbers
I promoted it exactly three ways: a LinkedIn post, an email to my existing list of around 6,200 subscribers, and one paid boost on LinkedIn of £150 aimed at UK small business owners. No blog SEO push, no ads on Meta, nothing fancy.
- 412 people completed the full quiz
- 318 of those gave a real email address (some clearly typed nonsense to see the result, which happens with every quiz, budget for it)
- 9% of completions, so about 37 people, booked a discovery call within the six weeks
- Cost per lead from the paid £150 spend worked out at £2.80, which is cheap compared to the £8 to £14 I usually pay for a cold LinkedIn lead ad on a static offer
- Of the 37 calls booked, 6 became paying clients within the six-week window, at an average first invoice of £1,850
On paper that’s a brilliant return for £187 total spend. But here’s what the number doesn’t show you: 74% of the traffic came from my existing list, people who already trusted me. The organic reach and the one paid boost brought in the rest. If I’d built this exact calculator with zero existing audience and zero list, I’d have had maybe 20 completions in six weeks, not 412. That’s the bit that gets left out of every “I built a quiz and made £10k” story you read.
Where I think everyone’s getting this wrong right now
AI has made the actual building of these tools nearly free. A year ago you’d have paid a developer £1,500 to £4,000 to build custom scoring logic into a lead capture tool. Now you can do it in an afternoon with ChatGPT and a £37-a-month subscription. That’s a good thing for small business owners who never had that budget.
But the barrier to building has dropped faster than the barrier to being seen. Everyone who reads this post and thinks “I’ll build a calculator too” is about to add to a growing pile of shiny, well-built, empty tools sitting on business websites with three completions a month, most of them from the business owner’s own team testing it. I see this constantly with consulting clients now: they’ve built the tool, it works fine, and then they message me confused about why it isn’t “converting.” It isn’t converting because nobody’s finding it. The tool was never the hard part. It just used to feel like the hard part, so building it felt like progress.
This is the uncomfortable bit nobody selling AI tools wants on their landing page: distribution was always the actual bottleneck, AI just removed the excuse of “I couldn’t build it anyway.” Now you can build it in a weekend and still get nowhere, which is arguably worse for morale than not building it at all.
What moved the needle for me
Three things mattered more than the tool itself:
- The list I already had. Six years of showing up, writing, and being on stages meant 6,200 people would open an email from me. Without that, this whole thing is a much smaller experiment.
- The specificity of the questions. I nearly built a generic “how AI-ready is your business” quiz and it would have flopped, because it’s vague. Once I narrowed it to the exact fork in the road people were already stuck on (“hire someone or DIY”), completions jumped. This is the same lesson Joe Pulizzi’s content approach teaches over and over: narrow beats broad, every single time.
- The follow-up sequence, not the quiz result. The result page converts nobody by itself. It’s the three emails after it, spaced two days apart, that turned quiz-takers into calls. The quiz is the icebreaker, the emails are where the trust gets built.
Should you build one?
If you’ve already got an audience, even a small one, a list of 500 to 1,000 engaged people, a calculator or quiz is one of the better uses of an AI-shortened weekend. It works particularly well for businesses selling a decision, not a product: “should I switch accountants,” “is my pricing too low,” “am I ready to hire,” that kind of thing. It works less well if you’re a straightforward product business, because a calculator that just tells someone the price of what you already sell is a pricing page with extra steps.
If you don’t have an audience yet, spend the weekend on something that builds one instead, guest posts, a proper LinkedIn posting habit, showing up on other people’s podcasts, and come back to the calculator once you’ve got a few hundred people who’d notice if you disappeared. I’d rather see a small business owner get their Instagram or LinkedIn presence to a decent size first than spend the weekend polishing a lead magnet that then sits unseen.
And if the idea of doing all this on your own feels like too much on top of everything else you’re running, this is exactly the kind of small, contained project an AI implementation coach can help you build in a day rather than a weekend, because they’ll already have the scoring logic templates and the follow-up sequence sorted.
What I’d change next time
I’d cut the quiz from ten questions to seven, the drop-off between question six and the end was 22%, which tells me it was one or two questions too long. I’d also test a shorter result page, mine ran to about 400 words and the average time on that page was 38 seconds, meaning most people skimmed the headline and score and left. And I’d build the second calculator sooner instead of waiting, because the second one always takes half the time once you’ve done the first.
Free resource: The Lead Magnet Funnel Build Checklist.
Frequently asked questions
How much does it cost to build an AI-powered lead calculator?
Realistically £30 to £50 a month for a no-code quiz builder like Outgrow or Involve.me, plus a few hours of your time using ChatGPT or Claude to write the scoring logic and copy. My build cost £37 total plus a weekend of my own time. Custom-coded versions from a developer still run £1,500 upward, which most small businesses no longer need to pay.
Do calculators convert better than a plain contact form?
For engagement, yes, quiz completion rates typically run far higher than form fill rates on a static page because people are curious about their own result. For actual sales, it depends entirely on the audience you send it to. A calculator sent to a warm list will outperform a contact form, sent to cold traffic with no existing trust it often performs about the same or worse.
What’s the best AI tool for writing the scoring logic on a quiz?
ChatGPT and Claude both handle this well, you describe your outcome bands and the weighting you want per answer, and it drafts the rules for you in minutes. The actual building of the quiz interface still happens in a dedicated tool like Outgrow, Involve.me, or Typeform, AI writes the brains, the platform builds the shell.
How long does it realistically take to build one from scratch?
One weekend if you already know your questions and have an email list ready to connect. Budget a full extra day if you’re starting from nothing, including working out what question your audience is stuck on, because a vague quiz gets ignored no matter how well it’s built.
If you would rather not build this yourself, here is how to build an AI workflow.
Related reading: Is ChatGPT Recommending Your Small Business? Here’s How to Check (and Fix It) and How I Built an AI Lead Calculator in a Weekend (And What It Told Me About My Own Marketing).
More on AI automation support in Liverpool.
