The 1% Fix — Frank Shea, Jr.

I texted my buddy Tommy a long message about British Cycling one night. Not because he asked. Because I'd just spent an afternoon watching a coworker manually enter data into a spreadsheet that I'd already automated for him three months earlier.

He didn't use it. He went back to the old way. And I needed to tell someone about Dave Brailsford before I lost my mind.

The Story I Keep Telling

In the early 2000s, British Cycling was a joke. No Tour de France wins. Poor Olympic results. Their reputation was so bad that a major bike manufacturer refused to sell them bikes — afraid it would damage the brand.

They weren't lacking effort. They were lacking systemic improvement.

Then Dave Brailsford took over as performance director and introduced a deceptively simple idea: "If you break down everything that goes into riding a bike and improve each element by just 1%, the cumulative effect will be extraordinary."

He called it "the aggregation of marginal gains."

They didn't chase one big breakthrough. They improved hundreds of small things. Tire compounds. Rider positioning. Nutrition plans. Massage gels based on absorption rates. They brought their own pillows to hotels. They tested mattress types. They painted the inside of the team truck white so they could spot dust that might cause infections.

Painted the truck white. To spot dust.

The results? 2008 Beijing Olympics — British Cycling won 7 of 10 gold medals on the track. 2012 London — repeated dominance. First-ever British winner of the Tour de France. Then another. Then another.

They didn't outspend everyone. They out-systemed them.

The Accountant in the Room

Here's why this story lives in my head rent-free: because I've spent 30 years watching companies do the opposite.

I texted Tommy one afternoon from my desk:

"I sit here and watch the other accountants do shit that I don't want to do while I contemplate better ways of doing it so nobody has to do this stupid shit on repeat every day."

That text is the most honest career summary I've ever written.

I've worked in ERP systems — Microsoft Dynamics GP, jBASE, legacy platforms that look like something out of a Cold War movie. Green and black screens. Overnight batch jobs that extract data into SQL so someone can open it in Excel the next morning and pretend they're doing analytics.

I described one system to Tommy like this:

"They have jBASE which is their 'live system'. Think 1980s movies. Green and black screens. Then overnight they have a SQL server that extracts data from that system and loads it into more Microsoft friendly format."

And the kicker:

"They load snapshot data into SQL. They take a snapshot of inventory yesterday and load that today. So data is a day old. Find an error. Fix it. Wait until tomorrow and see if it worked."

That's not a system. That's a prayer with a 24-hour lag.

No Guardrails. Total Lawlessness.

Every company I've worked for has some version of the same disease: a system that was stood up years ago, configured badly, and now runs on inertia and fear.

"And a million different people at a million different places are allowed to enter transactions into the system. No guardrails. Total lawlessness."

I sent that text and meant every word. Operations teams creating duplicate orders. Warehouses shipping without documentation. People entering journal entries with no approval chain. EDI transactions failing silently because nobody set up the error handling.

At one job, the ops team was picking and choosing to route orders to a different warehouse for a customer set up as FOB — creating duplicate orders in the system, breaking the connection to the original EDI document, which meant subsequent invoices didn't go out automatically. By the time anyone noticed, it was a mess of manual corrections and angry customer calls.

This is what most businesses look like under the hood. Not the polished dashboards. Not the strategy decks. The actual plumbing — where transactions enter the system, flow through processes, and come out the other end as financial statements — is usually held together with manual overrides, tribal knowledge, and one person who knows where the bodies are buried.

That person is usually me.

He Won't Change

The hardest part isn't finding the problem. It's getting people to accept the fix.

"He's pretty much useless. He won't change. I've shown him a better way to do what he's doing. I built it and gave it to him. He doesn't use it."

I've had this experience more times than I can count. You see a process that takes someone two hours a day. You build an automation that does it in thirty seconds. You hand it over. You explain it. You document it. And three weeks later, they're back to the spreadsheet.

It's not stupidity. It's identity. People don't resist better tools — they resist the implication that what they've been doing was wrong. The spreadsheet they've been maintaining since 2014 isn't just a spreadsheet. It's proof they've been working hard. You can't just replace it. You're replacing their evidence of effort.

Brailsford understood this. He didn't walk into British Cycling and say, "You've all been doing it wrong." He said, "What if we made everything 1% better?" That framing matters. It's not a revolution. It's a refinement. Nobody had to feel stupid. They just had to be open to slightly better.

Most corporate environments aren't that generous.

The Spreadsheet, the Duplicates, and the Nap

I sent Tommy this one after a particularly long day:

"The spreadsheet had a bunch of shit that needs to be refreshed and viewed in the right order. Then the second half I manually override certain things for missing data. Repeat the process."

That's the job. For a lot of people, that IS the job. Open file. Refresh. Scroll. Override. Save. Email. Wait. Repeat.

Later:

"I've got no time for her to sort that out manually so yeah I spent the afternoon figuring out how to get duplicates out automatically."

That's the other version of the job — my version. Someone else creates the mess, and I spend an afternoon building a system so the mess can't happen again. Not because I was asked to. Because I can't stop myself.

Eventually I texted:

"I am down to like two accounting tasks at work I do sort of manually. Once Claude figures that out I'll be looking for Jerry's carpenter to build a napping area under my desk."

That was a joke. Mostly.

The Aggregate of Incremental Gains

Here's the thing about the British Cycling story: people hear it and think it's about bike racing. It's not. It's about how change actually works.

It's not the big ERP migration. It's not the management shakeup. It's not the consultant's six-figure transformation roadmap. It's the person who notices that the team truck is dirty and paints it white. It's the person who realizes the snapshot data is a day old and builds a real-time query. It's the person who sees a spreadsheet that takes two hours and builds a script that takes thirty seconds.

I used the phrase with Tommy about something completely unrelated — switching cell phone providers:

"Pain in the ass to be sure. But you'll save money in the long run. The aggregate of incremental gains."

I used it about clothes — after Stitch Fix started working for me at the office:

"This is all about incremental gains."

It's not a business strategy. It's a way of seeing. Once you start looking for the 1% fix — the small, obvious thing that nobody's bothered to improve — you can't stop. Every process becomes a puzzle. Every workaround becomes an opportunity. Every "that's how we've always done it" becomes a dare.

GP Is Going Away

"GP is going away so they aren't supporting it further but companies will hang on to their systems until they're literally dead."

I sent that to Tommy about Microsoft Dynamics GP — the ERP I've spent more time inside than any human should. Microsoft stopped developing it. It's end-of-life. And companies will run it until the servers catch fire because migrating is expensive, terrifying, and nobody wants to own the decision.

This is the opposite of marginal gains. This is the compounding of marginal neglect. Every month you don't improve the system, the debt gets bigger. Every workaround you accept becomes permanent. Every manual process becomes sacred tradition.

British Cycling painted the truck white. Most companies won't even sweep the truck out.

What I've Learned

I'm 48. I've been inside more broken systems than I can count — ERPs, accounting teams, data warehouses, Excel nightmares, green-screen terminals. Here's what three decades and one obsessive re-telling of the British Cycling story have taught me:

1. The 1% fix is always available. You don't need a budget. You don't need approval. You don't need a project plan. There is always something small you can improve today. A query that runs faster. A report that's easier to read. A process that has one fewer manual step. Start there.

2. Systems beat heroics. The person who stays late to manually reconcile a spreadsheet is not the hero. The person who builds the system so nobody has to reconcile it again is the hero. One of them gets praised. The other gets ignored. But only one of them actually fixed anything.

3. The resistance is personal, not logical. When someone rejects a better process, they're not rejecting the process. They're protecting their identity. Understand that, and you'll stop being frustrated. You'll start being strategic.

4. Compounding works in both directions. Small improvements compound into extraordinary results. Small neglects compound into extraordinary failures. Most companies are compounding neglect and wondering why nothing works.

5. Paint the truck white. The strangest-sounding improvement is often the most important one. The thing that makes everyone say, "Why would you bother with that?" is usually the thing that makes the difference. Bother with it. Every time.

James Clear put it best: "You don't rise to the level of your ambition — you fall to the level of your systems."

I've been building systems my whole career. Some of them got used. Some of them got ignored. But the ones that stuck — the ones that actually changed how a company operates — were never the big flashy projects. They were the 1% fixes. The white paint. The better query. The automation nobody asked for but everybody needed.

That's the work. It's not glamorous. Nobody puts "painted the truck white" on a résumé. But it's the work that compounds.

And compounding is the only thing that actually wins.