How many experiments should you run per month
Experiment velocity matters more than any single experiment. Founders who run 4-6 experiments per month learn 10x faster than those who run one per quarter. Here's how to find the right pace for your situation.
Growth is a numbers game. Not in the sense of blasting out content or cold emails, but in the sense that more experiments equals more learning. Companies with high experiment velocity, like Booking.com (25,000 experiments per year) or Netflix (hundreds per month), consistently outgrow their competitors.
You're not Booking.com. But the principle scales down. A solo founder running 4 experiments per month will dramatically outperform one running 4 per year. This guide helps you find the right velocity for your situation and maintain it sustainably.
Why velocity matters more than accuracy
If you run one experiment per month with 90% confidence in results, you get 12 learnings per year. If you run four experiments per month with 70% confidence, you get 48 learnings. Even accounting for the lower confidence, you know far more by year's end. Speed of learning beats precision of learning at the early stage.
There's also a compounding effect. Each experiment informs the next. The fourth experiment of the month benefits from the lessons of the first three. Over a year, that compounding means your later experiments are dramatically better than your early ones. Higher velocity means you reach that quality level faster.
Brian Balfour, former VP of Growth at HubSpot, argues that experiment velocity is the single best predictor of growth team success. It applies to solo founders too, just at a different scale. The founder running 50 experiments this year will have a better growth engine than the one running 5, regardless of individual experiment quality.
The right number for solo founders
For solo founders, aim for 4-6 experiments per month, or roughly one per week. This is aggressive enough to generate real learning but sustainable enough to maintain alongside product development, support, and everything else you're doing.
The key is keeping experiments small. Each experiment should take no more than 2-4 hours to implement and measure. If an experiment takes 20 hours, break it into smaller tests. Instead of redesigning your entire onboarding flow (20+ hours), test changing the first screen (2 hours), then the welcome email (1 hour), then the default settings (2 hours).
Some weeks you'll run zero experiments because of a product launch, a bug emergency, or just life. That's fine. Some weeks you'll run two because you had a productive sprint. The monthly number matters more than the weekly one. Track it and aim for consistency over perfection.
How to maintain velocity without burning out
Batch your experiment work into a fixed time slot. Most solo founders find two hours on Monday morning works well: 30 minutes to review last week's results, 15 minutes to pick the next experiment, and 60-75 minutes to implement it. Batching prevents experiments from bleeding into your other work.
Keep a prioritized backlog of experiment ideas so you never waste time deciding what to test. When your weekly slot arrives, the next experiment should already be selected. Use ICE scoring to maintain the backlog, and you'll always know what comes next.
Automate measurement as much as possible. Set up event tracking once, create a dashboard once, build an email template once. Each experiment should require as close to zero measurement setup as possible. The less overhead per experiment, the more experiments you can run.
Scaling up when you're ready
If you find yourself consistently running 4-6 experiments per month and wanting more, you have three options: make experiments even smaller (test a single variable instead of a change set), hire a part-time growth person, or use tools that accelerate the experiment cycle.
Tools like Golden Gecko help by providing pre-built experiment templates, so you skip the hypothesis and method design phase. Instead of spending 30 minutes designing an experiment from scratch, you pick a relevant playbook, customize it for your product in 5 minutes, and start. This alone can double your velocity.
Another scaling strategy is running parallel experiments in different parts of the funnel. Test an acquisition change and an activation change simultaneously since they don't interfere with each other. A solo founder can realistically run 2-3 parallel experiments across different funnel stages. That gets you to 8-12 experiments per month without much extra time.
Tracking your experiment velocity
Add experiment velocity to your monthly metrics review. Track: experiments started, experiments completed, experiments with a clear result, and learnings documented. The gap between started and completed tells you about execution discipline. The gap between completed and clear result tells you about measurement quality.
Set a minimum viable velocity and protect it. If your minimum is 3 experiments per month, treat the third experiment like a deadline. When you're at month's end with only 2 experiments done, run a quick, small test to maintain the habit. Velocity is a muscle that atrophies quickly when unused.
Celebrate velocity, not just wins. If you ran 5 experiments this month and all 5 failed, that's still a great month. You eliminated 5 bad ideas and have 5 lessons to inform next month's experiments. The worst month isn't one where experiments fail; it's one where no experiments run at all.
Problems this guide helps with
Paying for every user when product should spread itself
You're spending money on ads to get every single user. Meanwhile, competitors seem to grow organically. Your product isn't spreading on its own.
Paying for ads while competitors grow free
Your competitors rank on Google and get free traffic. You're stuck paying for every click. SEO feels impossible and content marketing takes forever.
Put this into practice
Golden Gecko gives you proven playbooks matched to your goals, step-by-step guidance, and AI that tells you what results mean.