# When to Kill Good Experiments The hardest part of GTM isn’t knowing what works—it’s knowing what to stop. Startups don’t die from bad ideas. They die from **too many decent ones**—each eating a little attention, a little focus, a little calendar space that could’ve gone to something great. The paradox of growth is this: As your experiments start _working_, they get harder to kill. And if you don’t kill them fast enough, they quietly strangle the next breakthrough. ### Why “Good” Is the Real Threat Bad ideas are easy to spot. They flop, they stall, they get zero traction. But “good” ideas? They tease you. They give you _just enough_ signal to justify one more round. In GTM, that’s dangerous. Because every “good” experiment that stays alive consumes bandwidth that your “great” one needs to scale. Most founders don’t fail from lack of wins—they fail from lack of pruning. ### The Diagnostic: 3 Questions Before You Keep Going When a campaign, channel, or playbook shows mild success, pause and ask: **1. Is it compounding, or just repeating?** If you have to feed it constantly to see results, it’s not compounding—it’s manual labor. **2. Is it teaching you something new?** An experiment that doesn’t generate _fresh learning_ isn’t an experiment anymore—it’s a hobby. **3. Is it displacing a bigger bet?** Every hour you spend optimizing a B-tier play is an hour stolen from the A-tier one you haven’t launched. If you can’t say “yes” to at least one of those, it’s time to kill it. ### How to Kill Without Chaos Founders often fear that ending something means _wasting effort._ But the goal isn’t to prove a campaign wrong—it’s to free attention for something righter. The best GTM operators treat projects like **disposable scaffolding**: use it, learn, dismantle it, and build higher. Here’s a clean kill framework: **Step 1 — Review weekly.** Short feedback loops reveal which experiments have plateaued. **Step 2 — Log your learnings.** Every kill should leave behind a one-line lesson (“cold DMs convert 2× only when tied to live events”). **Step 3 — Celebrate the kill.** High-performing teams treat endings as progress markers, not losses. Every experiment you end deliberately sharpens your focus. ### What “Good” Looks Like When You’re Doing It Right A founder in motion doesn’t run _more_ experiments—they run _fewer_, faster. You can tell when a GTM motion is mature because: - The number of _active_ experiments goes down. - The number of _learned principles_ goes up. - Every new test builds on a prior truth. The output looks calmer, but the learning velocity skyrockets. Killing isn’t the opposite of progress—it’s how you make room for it. When something’s merely good, it’s time to stop optimizing and start reallocating.