Verified against ChatGPT · 2026-08-03
Pick a GTM channel by testing cheaply across many, not betting on the obvious one
Scores candidate go-to-market channels on fit, cost, and speed to signal rather than raw potential, then builds a pre-committed pass/fail test plan for the top two before any real budget gets spent.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are a GTM strategist using the "bullseye" approach to channel selection: test cheaply across many channels before committing real budget to the one that looks obvious. CONTEXT Product: Shift-scheduling and labor-cost forecasting for independent restaurants Target customer: Restaurant owner-operators, who mostly find tools through their POS vendor marketplace or other owners in local groups Budget/resources available: $800/month, one founder doing GTM part-time alongside product work Channels being considered, if any: POS-vendor marketplace listing, local restaurant-owner Facebook groups, cold outreach to multi-location groups What's already been tried, if anything, and the result: Ran Google Ads for 2 weeks, $300 spent, 2 signups, both churned within a week CANDIDATE CHANNELS List candidate GTM channels — use POS-vendor marketplace listing, local restaurant-owner Facebook groups, cold outreach to multi-location groups plus any others clearly relevant to how Restaurant owner-operators, who mostly find tools through their POS vendor marketplace or other owners in local groups actually discovers solutions like this, given $800/month, one founder doing GTM part-time alongside product work. Do not repeat a channel from Ran Google Ads for 2 weeks, $300 spent, 2 signups, both churned within a week without directly addressing why it's worth reconsidering or ruling it out for good. SCORING Score each channel on three axes only: fit (how closely it matches where Restaurant owner-operators, who mostly find tools through their POS vendor marketplace or other owners in local groups already looks for solutions), cost/effort to run a real cheap test, and speed to a usable signal. Do not score on scale or long-term potential — that is a later-stage question, and scoring for it now is exactly what leads teams to over-invest in an unproven channel before it's actually worked at small scale. TOP-TWO TEST PLANS Select the top two channels by that scoring and design a 2-week test for each, with a pass/fail metric stated in advance, before any test result exists to argue with or rationalize around. POOR-FIT CALLOUT Explicitly name any channel from the list that's a poor fit given the customer's actual buying behavior, and say why — so it doesn't get tried later just because it's popular, cheap-sounding, or "what everyone in this space does." OUTPUT FORMAT A scored table across all candidate channels, then two test plans with pre-committed pass/fail metrics, then the poor-fit callout as a short closing note.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
This is the Bullseye framework from Gabriel Weinberg and Justin Mares' Traction: rank many channels cheaply on fit, cost, and speed to signal before committing real budget to any single one, because early-stage teams reliably default to whichever channel is most visible — usually the one their competitors are already running — rather than the one that actually fits how their specific customer buys. Explicitly excluding scale and long-term potential from the scoring matters because that's precisely the axis that makes an unproven channel look attractive on a spreadsheet before it's been tested at all; potential is a real question eventually, but optimizing the first test for it is how a team ends up with a large, expensive bet on a channel nobody actually validated at small scale. Pre-committing the pass/fail metric before running the test is what stops a team from quietly redefining 'success' after a lukewarm result comes in — a documented and common pattern in the {{channels_already_tried}} data most founders already have sitting around from a channel they kept funding past the point the numbers had already answered the question. Requiring the model to directly address any previously-tried channel, rather than silently ignoring it, also closes a specific gap: a channel that already failed once needs an explicit reason to try again with a different test design, or it needs to be ruled out for good — leaving it unaddressed is how the same underperforming channel quietly gets re-tried every quarter under a slightly different name.
Verified against
ChatGPT GPT-5.1 · 2026-08-03
Changelog
- 2026-08-03 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — custom software, built and maintained for you — that's Scult's day job.
EXPLORE CUSTOM SOFTWARE
