Most solo founders are waiting for the big break: the viral post, the anchor client, the product that finally takes off. Meanwhile the business is quietly leaking time and money in a dozen small places, such as the slow reply, the clunky proposal, the invoice nobody chases.
The “one percent” idea says to stop waiting and fix the leaks. It’s a good idea with an overhyped origin story. Here’s the real story, the honest caveats, and a practical way to apply it when you’re a company of one.
Where the One Percent Idea Comes From
The most famous version is “the aggregation of marginal gains,” tied to Dave Brailsford, who became performance director of British Cycling in 2003. The pitch: break down everything that goes into riding a bike, improve each piece by a small amount, and the improvements add up. Reported examples range from the obvious (training, nutrition, bike design) to the quirky (pillows, hand-washing technique). Under his leadership British cyclists won eight gold medals at the 2008 Beijing Olympics, and in 2012 Bradley Wiggins became the first British winner of the Tour de France. Brailsford has summed up the approach as, “Forget about perfection; focus on progression, and compound the improvement.”
James Clear later made the idea mainstream in Atomic Habits. His headline math is that if you get 1% better every day for a year, you end up about 37 times better. (It’s just 1.01 multiplied by itself 365 times, which comes to roughly 37.8.) He pairs it with the line, “You fall to the level of your systems,” meaning outcomes follow whatever repeatable process you’ve built. Our sister site The Growth Reads has a full Atomic Habits review if you want the whole framework.
The Caveats (Read These Before You Get Excited)
A tough-minded solo founder should hold the story at arm’s length.
- The 37x figure is a metaphor, not a forecast. Nobody improves their business by exactly 1% a day, and you can’t measure it if they did. It illustrates compounding; it doesn’t predict your revenue.
- Not everyone buys the cycling story. Wiggins himself has called marginal gains “a load of rubbish,” arguing that the real drivers were talent, hard work and getting the fundamentals right rather than the pillow you sleep on. A UK parliamentary committee report in 2018 also accused Team Sky of crossing an “ethical line” in its use of medical exemptions around Wiggins’s 2012 Tour win, which the team and Wiggins denied. Treat the cycling tale as a parable, not a proof.
- Small gains can’t fix the wrong business. A faster invoice process won’t help if nobody wants your offer. Tweaks work after you have something that sells.
The sensible reading: fundamentals first, then polish. Get the offer, the price and the audience roughly right. Then use small improvements to keep pulling the business forward.
Big Bets and Small Gains Are Partners
Pieter Levels is a good example of the combination. In March 2014 he set himself the challenge of launching 12 startups in 12 months, one a month, and said he’d call a project a failure if it showed no signs of demand within a month. That’s the big-bets side: try lots of things quickly, cut what isn’t working. A couple of those projects, Nomad List and Remote OK, were the ones that stuck. In a later post on how he builds, he explained that he reuses code snippets from earlier projects and keeps things simple. Each project made the next faster. That’s compounding in a very literal sense.
Consistency is the other side. Justin Welsh publishes one newsletter essay every Saturday, and his site listed Issue #250 by mid-October 2026. No single essay moved the needle on its own. The pile of them did. (If you’re more of a “burn the boats” type, go all-in on the big commitment first and use small gains to keep the engine tuned.)
How to Run a One Percent Practice in a One-Person Business
Step 1: Keep a friction log
For two weeks, write down every moment something felt slower, clumsier or more annoying than it should: a question you answered for the fifth time, a step you forgot, a client confused by your invoice. Don’t solve anything yet. You’re collecting evidence.
Step 2: Pick one fix a week
Every Friday, choose the single most irritating item, or the one closest to money, and fix it in under an hour. Examples:
- Write a standard reply for the question you keep answering.
- Add a clear next step and payment link to your invoice.
- Shorten your intake form by two questions.
- Set a reminder to follow up on every proposal after three business days.
- Rewrite the first two lines of your proposal template.
Step 3: Measure one number
Pick a number that the fix should move, and note it before and after: reply time, proposal win rate, days to get paid, hours spent on admin. Say your proposals win two out of ten, and you add a one-page “what happens next” section. A month later you can see whether it moved. If it didn’t, you’ve learned something, and you drop it. Without a number, you’re just redecorating.
Step 4: Lock it in
If the fix worked, make it permanent: update the template, the checklist or the saved prompt. A gain you don’t capture disappears the next time you’re busy. This is the same loop as building any habit; see developing great habits.
Where AI Fits
AI is good at the tedious parts of this process: finding patterns and drafting alternatives. It is not good at knowing what matters to your customers.
- Mine your own material. Paste in a month of (anonymized) client questions or feedback and ask for the recurring themes. That points to your friction log’s best entries. Check the themes against the raw messages before acting.
- Draft the fix. Ask for three versions of the new proposal opener, or an FAQ based on those repeat questions. You choose, edit and test.
- Review the numbers. Give it your before-and-after figures and ask what else could explain the change. Then be honest about sample size: three proposals is not a trend.
The human decides what’s worth fixing and whether it worked. The AI speeds up the options. Anything that goes out to clients still needs a check; see Get It Right, Not Just Fast.
What Counts as a One Percent Fix
The test is simple: it takes under an hour, it’s reversible, and you can name the number it should move. “Redo the website” fails. “Add a testimonial and a booking link to the top of the services page” passes. Small also means specific, so write the fix as a sentence you could hand to a contractor or paste into an AI tool: what changes, where, and how you’ll know it worked. Vague improvements feel productive and change nothing.
Guard Against the Grind
A one-percent mindset can turn into an excuse to tinker forever or to feel perpetually behind. Keep it light. One fix a week is plenty. Some weeks the best improvement is a day off, and the compounding that matters most may be your own energy (see how to beat burnout).
The Takeaway
Don’t wait for a breakthrough. Start a friction log this week, fix one thing every Friday, measure it, and bake in what works. Over a year that’s about fifty small upgrades to a business that you actually run. It won’t be exactly 37 times better, but it will be noticeably sharper. If you’d like a head start finding where AI can take over the tedious parts, grab the free “One-Person Business AI Stack” checklist.
Sources
- James Clear, article on marginal gains (Brailsford, British Cycling, 2008 and 2012 results; 1% math): jamesclear.com
- World Economic Forum, “This coach improved everything by 1%. This is the remarkable difference it’s made” (Brailsford quote): weforum.org
- James Clear, Atomic Habits book page: jamesclear.com/atomic-habits
- Sticky Bottle, “Wiggins: Marginal gains and inner chimp were a load of rubbish”: stickybottle.com
- CNN, report on the 2018 UK parliamentary (DCMS) committee findings about Team Sky and Wiggins: edition.cnn.com
- Pieter Levels, “I’m Launching 12 Startups in 12 Months” (March 2014): levels.io; “How I build my minimum viable products” (September 2014): levels.io; homepage listing his current products: levels.io
- Justin Welsh, newsletter page (weekly Saturday essay; Issue #250 listed October 2026): justinwelsh.me