Why this matters
The free PDF was the standard lead magnet for a decade, which is exactly the problem. People download it, never open it, and forget who gave it to them.
A free tool is different. A pricing calculator, a "rate my landing page" checker, or a generator that writes a first draft in your style gives someone a result about their own situation in under a minute. They use it, come back to it, and send it to a friend.
AI can now build a simple one-page tool from a description, so the hard part isn't the code anymore. It's picking the right tool.
What makes a good free tool
- Answers a question your customer already asks you: you know there's demand.
- Gives a personal result in under 60 seconds: the payoff comes before they lose interest.
- Uses your expertise in the logic: it is harder to copy and shows what you know.
- Points naturally to what you sell: the result creates a useful next step.
Examples: a freelancer rate calculator, a "how much is this costing you" calculator for a problem you solve, a scorecard that grades a resume, pitch, or listing, or a generator that drafts something in a proven format.
Prompt 01
Choose the tool
Start with real customer questions and the expertise you can build into the logic. This prompt helps you compare ideas before you start coding.
Help me choose a free tool to build as a lead magnet for my business.
MY BUSINESS
- What I sell: [product or service, and price]
- Who buys: [specific customer]
- Questions customers ask me before they buy: [list 5-10]
- What I know that most people don't: [your expertise, frameworks you use with clients, numbers you know]
STEP 1: IDEAS
Give me 8 free tool ideas. For each one:
- What the user puts in (keep it to 5 inputs or fewer)
- What they get out
- Why my customer would care
- How it leads naturally to what I sell
Mix calculators, scorecards, checkers, and generators.
STEP 2: SCORE THEM
Score each idea 1-5 on: how often customers already ask this, how fast the result is, how much it shows off my expertise, how easy it is to build as one page, and how naturally it leads to my offer. Show a table.
STEP 3: PICK ONE
Recommend the best idea and explain why. Then define it precisely:
- The exact inputs (with labels and example values)
- The logic: the calculation, rules, or scoring used to produce the result, written so a developer or AI could build it
- The result screen: what it shows, and the one next step it offers
- Where the email capture goes (before the result, after it, or optional to save it), and the tradeoff of each
RULES
- Use my real expertise in the logic. If I haven't given you enough to build real logic, ask me questions instead of inventing numbers.
Prompt 02
Build the first version
Once you have a clear spec, give this prompt to Claude as an artifact, Claude Code, Replit, or Lovable.
Build a single-page web tool from the spec below.
THE SPEC
[paste the definition from Prompt 1: inputs, logic, result screen, email capture]
REQUIREMENTS
- One page, works on a phone first, loads fast
- Clean, simple design with lots of white space. Match these brand colors: [colors, or "black and white"]
- The result updates as soon as the inputs are filled in, or on one button press
- The result screen shows: the personal result, a one-sentence explanation of what it means, and one call to action: [book a call / join the newsletter / see the product]
- Email capture: [where it goes, per the spec]. Connect it to [your email tool], or leave a clearly marked placeholder
- A "share your result" button that copies a short summary and link
- No data stored except the email, if they give it
WHEN YOU'RE DONE
Test it with 3 sets of example inputs, including one unusual case, and show me the results so I can check the logic is right.
Pro move: make the result shareable
A "my score is 72/100" line people want to post can bring in more people than anything you'll write about the tool.
Do this next
Write down the question customers ask you most before they buy. If the answer involves "it depends on...", you've found your tool: the inputs are whatever it depends on.