There is a disease spreading through organizations right now, and it has nothing to do with AI or recession or interest rates. The disease is this: doing the thing that looks like solving the problem instead of solving the problem. It is everywhere. And if you are running a real business, it is probably costing you more than your worst employee.
Three stories. One pattern.
Amazon pledged to eliminate carbon emissions by 2040, then spent the last several years watching those emissions go up. Bezos has the press releases. He has the commitment. He does not have the results. That is not an environmental story. That is a management story about what happens when the announcement replaces the work.
France has arguably the most beloved and diverse national soccer team on earth, and it has not slowed the rise of ethnic resentment in the country. The argument was: put brilliant, inspiring men of every background on the biggest stage, and hearts change. It is a beautiful theory. The problem is that representation as a policy proxy, deployed without the harder structural work underneath, does not move the needle. Symbols do not substitute for systems.
Anthropic is turning Claude Code's autonomous mode on by default, which means more code will be written with less human review. This is being framed as progress. Maybe it is. But "progress" that removes human oversight without first establishing reliable accountability structures is not efficiency. It is a wager dressed up as a feature launch.
The thread connecting all three: someone, somewhere, chose the gesture over the outcome. And then called the gesture a win.
You do this too. So do we.
We have watched this pattern across every engagement we have run for the past two decades. A founder hires a marketing agency and starts counting deliverables instead of revenue. A product team ships a feature because a roadmap demanded it, not because a customer asked for it. An ops lead implements a new tool because the old process felt embarrassing, not because it was actually broken. The form of action substitutes for its function.
The reason it happens is not laziness. Lazy people do not build $2M businesses. It happens because action is visible and outcomes take time. Because the press release lands this week and the emissions data lands next year. Because a diverse roster makes a beautiful photo, and the structural reforms that would actually change a country's politics are slow, unglamorous, and politically expensive. Because shipping autonomous AI mode earns a TechCrunch headline, while the question of who is liable when it ships garbage sits comfortably unanswered.
Performance theater is not about deception. Most of the people caught in it believe they are doing real work. That is what makes it so dangerous. The energy is real. The motion is real. The results are not.
What it costs you specifically
At the $1M to $5M stage, you cannot afford the luxury of theater. A Fortune 500 can run a sustainability program that moves emissions in the wrong direction for five years and survive it. You cannot run a sales process that feels active but closes nothing for five quarters and survive it.
Here is what performance theater looks like at your scale. You do a weekly team standup that everyone treats as a status update instead of a blocker-clearing session. You build a CRM nobody logs into because the real pipeline lives in a founder's head. You hire a social media manager and call it marketing. You write a set of company values and post them on the wall and wonder why culture still feels off. You build a feature because a loud customer asked for it and you needed to feel responsive, not because the unit economics of keeping that customer justified the build cost.
None of these things are dumb decisions made by dumb people. Every single one of them is a reasonable-looking proxy for the harder thing. That is the tell. When something feels like progress but cannot be measured against an outcome, that is theater.
The Albedo test
There is a project called Albedo, a dead-simple single-file document database written in Zig, built by someone who got tired of the mismatch between the weight of existing tools and the actual problems they were solving. SQLite with migrations felt like overkill for a microservice that just needed to hold its own state. JSON loaded the whole file at startup. Neither fit. So the builder made something that did exactly what was needed and nothing more.
That is the opposite of theater. That is someone asking: what outcome do We actually need? And then building precisely that, without dressing it up. No announcement. No whitepaper. A small tool that fits the real problem.
We use this as a mental test now in client engagements. Before any decision, any hire, any build, we ask: are we doing this because it solves the problem, or because it looks like solving the problem? We call it the Albedo test. Is this the right tool for the actual job, or is this the tool that signals we are the kind of company that uses this kind of tool?
How to stop
First, get honest about your metrics. If the thing you are measuring cannot be directly tied to revenue, retention, or risk reduction, you are probably measuring effort instead of outcome. Effort metrics are theater metrics. They make you feel productive. They do not tell you whether the work is working.
Second, kill one proxy this week. Pick one activity in your business that everyone agrees is important but nobody can explain what changes if you do it better. Your weekly report nobody reads. Your social presence in a channel where none of your customers live. Your feature backlog grooming session that moves no code. Cut it. See what actually breaks. Usually: nothing. That is the diagnosis.
Third, and this is the hard one: stop rewarding motion. In early-stage companies, the founder's energy is contagious. Whatever the founder celebrates, the team repeats. If you celebrate shipping a feature, your team will ship features. If you celebrate keeping a customer who was about to churn, your team will prioritize retention. The culture follows what you actually cheer for, not what you write in the values doc.
The billionaires betting everything on the AI boom are not immune to this either. The bets that will actually pay off are the ones built on real infrastructure and real accountability, not on the narrative that AI is transforming everything and therefore any AI spend is justified. The ones who will lose are the ones who mistook the announcement for the arrival.
The honest version of this
We have been guilty of this ourselves. Early in client engagements we have produced beautiful technical architectures that solved problems the client did not actually have yet, because the architecture was elegant and we were proud of it. We have written documentation nobody read because documentation felt responsible. We have held retrospectives that felt productive and changed nothing because we named problems without assigning owners.
The difference between us then and us now is that we catch it faster. We have built enough scar tissue to recognize the feeling of motion without traction. It has a specific texture: lots of meetings, reasonable explanations for why results are not here yet, and a growing distance between what the dashboard shows and what the founder's gut says.
When you feel that distance, trust your gut. The dashboard is measuring theater. Your gut is measuring reality.
Results do not care about your intentions. Neither do your customers. Neither, eventually, will your bank account. Do the actual thing, or stop doing the thing. There is no third option that feels like action but counts as progress.