Skill decays after 11 attempts—the third try feels easiest

You've probably felt it: the first attempt at something new is clumsy, the second is slightly better, and by the third go you're weirdly confident — like the thing has clicked. Then, somewhere around attempt eleven, it quietly falls apart. Not dramatically. Just enough that you notice you're worse than you were a week ago.

Is that real, or just a story we tell ourselves? And if it is real, what does it mean for the way we build and run websites for Australian businesses?

The third attempt feels easiest — and there's a reason for that

There's a well-documented pattern in skill acquisition called the power law of practice: performance improves steeply at first, then flattens. The curve is sharpest in the earliest repetitions. Your third attempt feels easiest because it is the point of maximum relative gain — you've shaken off the worst of the errors, but you haven't yet accumulated the bad habits that come with over-repetition.

This is why a junior developer often ships their third landing page faster than their thirtieth. The first two were reconnaissance. The third is momentum.

But the curve has a second act. Once the easy gains are gone, performance plateaus — and plateaus are dangerous, because they're where attention drifts. Kahneman's work on System 1 and System 2 thinking is useful here: early attempts run on deliberate, effortful System 2 processing. By attempt eleven, the task has been handed off to System 1 — fast, automatic, and blind to its own errors. You're no longer thinking about the work. You're just doing it. And that's exactly when quality slips.

What "attempt eleven" actually looks like

In web development, an "attempt" isn't always a fresh project. It can be:

• The eleventh time you've built a contact form the same way
• The eleventh client site using your default theme stack
• The eleventh month of running the same SEO checklist
• The eleventh revision of a homepage layout

The decay isn't in your ability. It's in your attention to the specifics. The third attempt was bespoke. The eleventh is a template.

Loss aversion is why we don't notice the decay

Here's the uncomfortable part: we're bad at spotting our own decline. Kahneman and Tversky's work on loss aversion shows we feel losses roughly twice as strongly as equivalent gains — but only when we notice them. Skill decay is gradual, so it never registers as a loss. It registers as "fine."

Meanwhile, variable-ratio reinforcement — the same mechanism that makes unpredictable rewards so compelling — keeps us coming back to the same approach. Sometimes the eleventh site performs brilliantly. Sometimes it doesn't. The unpredictability masks the trend. We attribute the good outcomes to skill and the bad ones to the client, the market, or the algorithm.

For Australian businesses, this matters because the decay is usually invisible until a competitor's site outperforms yours on something you stopped paying attention to eighteen months ago.

A concrete example: the Google core update pattern

Consider how many Australian agencies responded to the 2023 and 2024 Google core updates. The pattern was strikingly consistent: sites that had been built by teams on "autopilot" — same structure, same content strategy, same internal linking — saw larger ranking drops than sites that had been revisited with fresh eyes.

The technical work wasn't wrong. It was just stale. The teams had hit attempt eleven and stopped interrogating their own assumptions. The sites that recovered fastest were the ones where someone had gone back to attempt three — the deliberate, curious, System 2 version of the work.

Decision-making under uncertainty: why "best practice" is a trap

"Best practice" is a phrase that should make you slightly nervous. It's a snapshot of what worked under a specific set of conditions at a specific moment. By the time it becomes standard, the conditions have usually shifted.

This is where the behavioural economics literature on decision-making under uncertainty gets interesting for web work. Gerd Gigerenzer's research on heuristics shows that simple rules often outperform complex optimisation — but only when the rules are matched to the environment. Apply a rule from 2019 to a 2025 environment and you're not being efficient. You're being unlucky.

The practical implication for Australian businesses is uncomfortable but useful: the eleventh version of anything deserves more scrutiny than the third, not less. The third attempt is still being tested. The eleventh is being trusted.

How to break the cycle without starting over

You don't need to rebuild from scratch. You need to reintroduce friction at the right moments.

Rotate who does the work. If the same person has built the last ten contact forms, have someone else build the eleventh. Fresh eyes run on System 2 by default.

Set an expiry date on your defaults. Your theme, your hosting stack, your content template — give each one a review date. Not because they're wrong, but because you need to decide again rather than drift.

Measure something new every quarter. If your analytics dashboard hasn't changed in a year, you're measuring the same things and missing the same things. Add one metric, retire one.

Do a "third attempt" audit. Pick a process you've repeated more than ten times. Ask: what would this look like if I were doing it for the first time, today, with what I now know? The gap between that answer and your current process is your decay.

The competitive angle: why fresh teams sometimes win

There's a reason new agencies occasionally outcompete established ones on specific projects. It's not talent. It's that the new team is on attempt three, and the established team is on attempt eleven. The new team is still uncertain, still asking questions, still running System 2.

This is also why in-house teams sometimes outperform agencies, and vice versa — it depends entirely on which one has more recent practice with the specific problem at hand. Familiarity is not the same as competence. Sometimes it's the opposite.

For Australian businesses choosing a web partner, the question worth asking isn't "how many sites have you built?" It's "how recently did you change your mind about how to build them?"

What to do next

Pick one thing you've done more than ten times. Open it up. Look at it the way you looked at the third version — curious, critical, willing to be wrong.

Then change one thing about how you do it. Not because the old way was broken. Because the eleventh attempt always feels fine right up until it doesn't.

The businesses that stay sharp aren't the ones with the most experience. They're the ones who keep resetting themselves back to attempt three.