For decades, the production moat protected the incumbents.

If you had a sleek website, 4K video case studies and perfectly punctuated whitepapers, you were a tier-one player. The high cost of production acted as a signal. It told the market you had the capital, the taste and the longevity to invest in your image. In that world, production quality was a proxy for operational excellence.

AI has vaporized that moat.

Today, a two-person startup can produce a brand identity that looks as professional as a McKinsey report for the price of a Netflix subscription. When everyone can produce 10/10 quality for next to nothing, professionalism is no longer a signal. It's just the noise floor.

We call this the high-quality average. It's one of the most dangerous traps in modern business, because it feels like progress while it quietly erodes your competitive advantage.

The inflation of “nice”

When production becomes a commodity, the value of looking good trends toward zero.

We're watching a massive inflation of “nice.” Every LinkedIn post is perfectly structured. Every corporate video is color-graded to perfection. Every website has frictionless UI.

But because so much of this output comes from AI models trained on the average of the internet, it's statistically designed to be unremarkable. It lacks the rough edges of a real opinion. The result is a signaling crisis: customers can no longer use your aesthetics to tell whether you actually know what you're talking about.

If your content looks like everyone else's, the market assumes your service is like everyone else's.

You're being commoditized by your own quality.

The new scarcity: earned perspective

In a world of infinite, high-quality average, the remaining scarcity is earned perspective.

Earned perspective is the specific, often messy insight that only comes from doing the work. It's the uncomfortable truth about your industry. It's the data point from a failed project that changed how you operate. It's the opinion that makes half the room nod and the other half get defensive.

AI can't have earned perspective. It has never had a client yell at it. It has never stayed up until 3 a.m. debugging a systems failure. It has never lost a deal.

I have. I once flew a small team across the country to pitch organic growth to a large retailer. The best way to get there was a red-eye, and the trip turned into a multi-leg marathon. We spent two hours in their conference room walking through a carefully constructed presentation. Then they took us to a fancy dinner in Palo Alto. By then we'd been awake for about 24 hours and were all a little delirious, and that's when they told us they wouldn't have budget until next year. Our client lead asked for the check, and we flew home.

The presentation was polished. It didn't matter. What that trip taught me, painfully, is to truly qualify an opportunity before you spend the effort. It's not a sophisticated insight, but it's mine, and I paid for it with a cross-country flight. No model can hand you a lesson like that. It can only help you say it well once you have it.

The competitive advantage has shifted from how you say it to whether you have anything worth saying.

What this looked like for one of our clients

We saw this up close with a creative agency that was using off-the-shelf custom GPTs to simulate customer focus groups for a Fortune 500 client.

The output looked great. It was fluent, well-organized and confident. It was also useless, for two reasons. First, the models praised every concept put in front of them, because generic LLMs are trained to be agreeable. Second, the “personas” were stereotypes pulled from training data, the internet's average idea of a 30-year-old car seller.

The agency didn't need more polish. It needed its own perspective in the loop: years of first-party research, licensed audience reports and hard-won opinions about what actually drives its client's customers. So we built the tool around a point of view instead of a template. Psychographics carry 40% of a persona's weight and demographics carry 10%, calibrated to how the client's audiences actually behave. The personas reason from mindset, not age and zip code.

That ratio is the opposite of high-quality average. No model would arrive at it on its own, and no competitor could copy it by swapping in their own logo. It's earned. Once the system was grounded in that research and built to push back instead of agree, it became an asset the agency now deploys across its client portfolio. Read the full case study.

From content calendars to intelligence capture

Most companies treat content as a marketing task. They hire a social media manager or an agency to “create content.” When that person gets busy, they use AI to fill the gaps. The result is a feed full of high-quality average.

To escape the trap, stop thinking about content and start thinking about intelligence capture.

The most valuable things your company knows are trapped in internal exhaust:

  • The heated Slack debate between two engineers about a technical trade-off.
  • The email a founder sent explaining exactly why a standard industry practice is a waste of money.
  • The post-mortem from a project that went sideways.

At FM, we build systems that treat this internal exhaust as raw material. Instead of asking a marketer to brainstorm ideas, we capture real-world insight from the people doing the work, as it happens, so the firm's know-how becomes something it can use, not something that disappears into old threads.

The replacement test

Want to know if you're stuck in the high-quality average? Run your last three pieces of external communication through one question:

If I swapped my company's name for my biggest competitor's, would this still be true?

If the answer is yes, you're producing noise.

None of this means polish doesn't matter. Your work should look great. But looking great is now the price of entry, not the reason anyone picks you. The goal is to build a system that surfaces your most specific, earned insights and makes sure they don't get polished away by the machine.

Make it look great. Then make sure it says something only you could say