SKILL.md

Free Tool Strategy

A free tool wins when it solves a real problem your buyers Google, sits adjacent to the paid product, and does one thing well. Everything else is a side project wearing a marketing budget.

Workflow

Load context. If .agents/product-marketing.md or .claude/product-marketing.md exists, read it first. Then pin: core product, target audience, primary goal (leads, SEO, brand), and build capacity.
Generate candidates from pain. Four prompts: what does the audience Google, what manual process is tedious for them, what do they need to figure out before buying your product, what data do they wish they had. Each answer is a tool candidate.
Score every candidate. Use the 8-factor scorecard in references/tool-types-and-scorecard.md. 25+ is a strong candidate, 15-24 promising, under 15 gets dropped. Also read that file when choosing the tool type: calculator, generator, analyzer, tester, library, or interactive.
Check the economics. Lead value times expected leads must beat build cost plus maintenance. No plausible path to that: recommend a lead magnet instead and stop here.
Pick the gating level. Fully gated maximizes capture and kills usage. Ungated maximizes reach and SEO with zero direct leads. Default: ungated core experience, gate the full report or export behind email only. One qualifying question maximum.
Scope the MVP. Core function that works reliably, clear input, obvious output, works on mobile, email capture wired to somewhere useful. Explicitly skip: accounts, saved results, advanced features, edge-case polish.
Plan build vs buy. Custom build when the concept is unique and strategic. No-code (Typeform, Tally, Outgrow, Webflow) when testing the concept or speed matters. Embed something existing when the tool is not a differentiator.
Plan SEO from day one. Landing page targets "[thing] calculator" / "free [tool type]" queries. Supporting content targets "how to [use case]". Tools earn links because they are useful, unique, and shareable, so make the result page linkable.

Good vs bad

Bad: A CRM company builds a general "marketing maturity quiz" because quizzes are hot. No search demand, audience is anyone, results page says "you scored 62". Six weeks of build, forty leads, none qualified.

Good: The same company builds a "sales commission calculator". Sales ops people Google exactly that, the output is a shareable breakdown, the email gate is on the exportable spreadsheet, and every user is inside the buying committee for a CRM.

Verification

Before recommending a tool, present the completed scorecard with all 8 factors filled and the economics line (lead value x expected monthly leads vs build + monthly maintenance). Expect: score 15 or higher and positive economics within 12 months. If either fails, the recommendation is "do not build", say so and propose the lead-magnet alternative.

Completion checklist

[ ] Context loaded, goal and capacity pinned
[ ] Candidates generated from all four pain prompts
[ ] Scorecard completed with a number per factor
[ ] Economics check passed or the recommendation is "don't build"
[ ] Gating level chosen with reasoning
[ ] MVP scope lists what is cut, not just what is in
[ ] Build vs buy decision stated
[ ] SEO keyword targets named for the landing page

Any box unchecked: not done. Fix or say so.

Footguns

Building the product's little sibling. The free tool replicates a paid feature and cannibalizes upgrades. Fix: the tool solves an adjacent problem, upstream of the product, not a slice of it.
Gating everything. Full gate on an unproven tool means nobody uses it, so nobody links to it, so it earns nothing. Fix: ungate the core, gate the artifact (report, export, saved result).
Maintenance amnesia. Calculators with hardcoded rates and graders with API dependencies rot in months. Fix: score maintenance honestly in the scorecard and assign an owner before launch.
Skipping demand validation. A clever tool nobody searches for gets traffic only when you pay for it. Fix: check search volume for the tool query before scoping anything.

free-tools