Why Real Builders Leave and Mediocrity Gets Promoted

I’ve been talking to more and more professionals in decision intelligence, AI, and software development lately. Different companies, different industries, different roles. But they’re all noticing the same patterns.

Their best work gets ignored. Expensive consultants get celebrated for obvious recommendations. Leaders champion “AI strategy” without understanding what AI actually is. Projects that make millions get less recognition than PowerPoint decks.

These aren’t isolated incidents. This is an industry-wide problem.

The Latin root of “innovation” is innovat- — “renewed or altered.” Innovation isn’t about inventing new things from scratch. It’s about creating value today that you weren’t creating yesterday.

Most companies claiming to innovate are doing neither. They’re performing innovation for an audience that doesn’t know the difference.

What Innovation Theater Looks Like

You know it when you see it:

The aesthetic is perfect. Open offices. Unlimited PTO. Innovation labs with whiteboards covered in sticky notes. Executives in sneakers instead of suits. Hackathons that produce nothing. “Digital transformation” initiatives announced with great fanfare.

The substance is missing. Products that never ship. Pilots that never scale. PoCs that prove nothing except that someone got a consulting contract.

The tells are everywhere:

  • $10M consulting contracts while ignoring internal experts who already know the answer
  • Leaders championing “AI strategy” who can’t define what AI actually is
  • Teams running “1,600 ML models” whose managers can’t explain a ROC curve
  • PowerPoint presentations getting the same recognition as shipped products that make millions

This is innovation theater. It looks like innovation to people who have never done it. It’s obvious to anyone who has.

That’s why we call it theater.

Why Innovation Theater Persists

If it’s so obvious, why does it keep happening? Three reasons:

Career Safety

Calling out theater is career-limiting. The person championing the expensive AI initiative that produces nothing often gets promoted for “strategic thinking” and “driving innovation.”

The person who ships a straightforward solution using mean, median, and mode that makes $100M? They get told they “don’t understand the business” or “aren’t thinking big enough.”

Theater rewards performance, not results.

Legibility Bias

Theater looks impressive to non-technical leadership. Complexity beats simplicity in PowerPoint every time.

“We’re building 1,600 ML models with agentic flows” sounds infinitely more impressive than “we automated the manual process with basic statistics and saved $50M.”

The second one is innovation. The first one is a consulting bill.

But guess which one gets the budget?

Collective Action Problem

Everyone mid-level knows this is happening. But no individual can fix systemic dysfunction.

So builders either leave, become cynical, or learn to perform the theater while doing real work on the side. The system perpetuates because calling it out individually costs more than it’s worth.

The only solution is collective action. More on that later.

The Real Tests of Innovation

Want to know if you’re actually innovating or just performing? Ask three questions:

1. Does it ship?

Not PoC. Not pilot. Not “in testing.” Production, at scale, with real users relying on it.

If your “innovation” never makes it past the demo stage, you’re not innovating. You’re experimenting on the company’s dime without accountability for results.

2. Does it create measurable value?

Revenue up? Costs down? Time saved? Risk reduced? Be specific. Show numbers.

If you can’t articulate the value it creates, you probably didn’t create any.

3. Can the person championing it explain how it works?

If your executive is pushing AI but can’t explain linear regression at a fundamental level, you’re not innovating. You’re consuming buzzwords.

If the leader running your “digital transformation” can’t articulate what problem you’re solving and why your approach will work, you’re doing theater.

Real innovation requires understanding. Not PhD-level depth, but enough to distinguish signal from noise, substance from hype.

Innovation and Good Engineering Are Inseparable

Here’s something that needs to be said clearly: you cannot separate innovation from good engineering.

Good engineering practices breed innovation and experimentation. You can’t innovate without good engineering, and good engineering attracts talent capable of innovating.

Real innovation isn’t about novelty for novelty’s sake. It’s about novel implementations that create value. Creative uses of existing technology given unique constraints.

Some examples of actual innovation:

  • Using Haversine function to calculate distance between zip codes in an air-gapped system. Not a new algorithm. Novel implementation that solved a real constraint.
  • Combining Bayesian inference with Monte Carlo simulation to understand customer preferences at scale. (Patent pending.) Existing techniques, creative combination, massive business value.
  • Building a MILP product that added one step to the process and made hundreds of millions of dollars. The innovation wasn’t the MILP. It was understanding the problem structure deeply enough to know where MILP’s capabilities could traverse huge solution spaces profitably.

None of these are “breakthrough technology.” All of them created substantial value that didn’t exist before.

That’s innovation.

The Technical Literacy Problem

Let me say this plainly: if you run a tech company and don’t understand your core competency at even a fundamental level, you’re either lazy or incompetent. Choose one.

In the age of information, ignorance is a choice.

You don’t need a PhD. You don’t need to write code. But you need to understand what you’re building and why it works.

If you make decisions about AI, you should know what AI actually is. Not vibes. Not buzzwords. A real definition.

If your supply chain depends on semiconductors, you should know semiconductors depend on gallium. When China limits gallium exports, you shouldn’t be surprised.

If you’re investing millions in ML models, you should understand Type I vs Type II errors at minimum. You should be able to explain what a ROC curve tells you.

When executives lack this literacy, predictable things happen:

  • They pay McKinsey and Deloitte millions of dollars for explanations dumbed down to high school algebra. This is the Explainability Tax — paying consultants to translate what your own data scientists already told you, because you refused to learn enough to understand them directly.
  • They can’t distinguish signal from noise in their own teams. They optimize for things that sound impressive over things that work. They reward theater over substance because they literally cannot tell the difference.
  • They make catastrophically bad technology decisions because they’re operating on faith, not understanding. And when those decisions fail, they blame “execution” instead of strategy.

You wouldn’t hire a CFO who doesn’t understand financial statements. Don’t pretend you can run a tech company without understanding technology.

Novel Implementation > Novel Solutions (For 99.9% of Companies)

Here’s an uncomfortable truth: most companies should focus on creative implementation of existing solutions, not inventing new algorithms.

Why?

A) Your problems aren’t novel

All banks are solving the same problems: loan calculations, interest rates, default risk, fraud detection. Different metrics and weights, same fundamental decisions.

All retailers are solving: inventory optimization, demand forecasting, price optimization, supply chain logistics.

All manufacturers are solving: production scheduling, quality control, maintenance optimization, supply chain management.

The problems have been solved. You need to implement the solutions well for your specific constraints.

B) You can’t afford people who invent novel solutions

People who create genuinely novel algorithms make more than most of your VPs. They work at actual tech companies, research labs, or start their own ventures.

Your $165K total comp engineer isn’t inventing breakthrough genetic algorithms. And if they somehow do, they’re leaving immediately for 3x compensation elsewhere.

Don’t build your strategy around lightning striking.

C) Basic tools with good automation print money

There is so much value to be captured with mean, median, and mode combined with solid automation and good engineering.

But executives don’t consider this “innovative” enough, so it gets ignored while teams spend months architecting agentic flows for processes that don’t even exist yet because no one ever automated the manual workflow.

You’re trying to run before you can walk. You’re chasing novelty instead of value.

What Happens to the Builders

When PowerPoints get the same recognition as shipped products, builders leave.

Smart, hardworking people want to:

  • Learn and grow
  • Build things that matter
  • Get recognized and compensated for actual results

When nepotism and mediocrity dominate, when theater consistently beats substance, you lose your best people. They go somewhere their work actually counts.

I’ve seen this pattern repeatedly across organizations. The best builders quietly exit. The performative stay. Leadership wonders why execution is getting worse while they keep “hiring top talent.”

You’re not hiring top talent. Top talent won’t work for you anymore. They left when you promoted the person who made the flashy presentation over the person who shipped the product.

The stories are depressingly similar: optimization systems that return hundreds of millions in profit get less recognition than initiatives that produce slide decks. Data scientists who solve problems with elegant solutions get told they “don’t understand the business” by managers who can’t explain basic statistical concepts. Engineers who ship working products watch less capable colleagues get promoted for talking about products they’ll never build.

This isn’t about one bad manager or one dysfunctional company. This is a pattern across the industry. And it’s why the best builders are increasingly selective about where they work.

What You Can Actually Do About This

If you’re a builder stuck in innovation theater, you have exactly two options:

Option 1: Get Better at Translation

Learn to dumb down your complex solutions to pull on the emotions of decision-makers.

This isn’t about lying or oversimplifying. It’s about speaking the language of business value in ways non-technical leaders can grasp.

Can you explain why your solution matters without using jargon? Can you tie it directly to revenue, cost, or risk? Can you tell a story that makes executives care?

This is a real skill, and many talented builders fail here. They’re right about the solution but can’t communicate why it matters. I cover this extensively in “You Got the Data Job… Now What?” (Chapter 1) because it’s often the difference between getting support and getting ignored.

If you can master this, you can sometimes break through the theater. Sometimes.

Option 2: Leave

Don’t tolerate this shit.

We, the builders, need to collectively tell these executives and companies that we’re not going to put up with theater anymore.

If your company rewards PowerPoints over shipped products, leave.

If your best work gets ignored while expensive consultants get celebrated for obvious recommendations, leave.

If you’re building things that return hundreds of millions while watching mediocrity get promoted, leave.

Go make more money somewhere else. Or find a place that actually supports building real things. They exist. Not many, but they exist.

The only way this changes is if talent stops accepting it.

Companies doing innovation theater can only continue if builders keep showing up. When the best people consistently leave, when the talent gap becomes undeniable, maybe, maybe leadership will figure it out.

Vote with your feet. It’s the only vote that counts.

For Leaders Who Want to Stop the Theater

If you’re an executive reading this and thinking “oh shit, this might be us,” here’s what you can do:

  • Invest in technical literacy or get out of the way. You don’t need to code. But you need to understand your core competency well enough to distinguish substance from bullshit. If you won’t invest the time, delegate technology decisions to people who actually understand technology.
  • Measure shipped value, not initiated projects. Stop tracking how many AI initiatives you launched. Start tracking how many created measurable business value. Be ruthlessly honest about what’s working.
  • Reward results, not presentation skills. The person who shipped the boring solution that made $50M should get promoted over the person who pitched the flashy initiative that produced nothing. If your incentives don’t reflect this, fix them.
  • Stop tolerating mediocrity. Nepotism and tolerance for underperformance drive out your best people. When PowerPoints get the same recognition as products, builders leave. Full stop.
  • Experiment over committee. Want to know if something works? Run a small experiment. Get data. Decide based on results. Stop making technology decisions in conference rooms by consensus.

Conclusion: Innovation is Renewal, Not Performance

Innovation comes from the Latin innovat-: renewed or altered. It’s about creating value today that you weren’t creating yesterday.

Theater is about looking like you’re creating value while producing nothing.

The difference should be obvious. But in many organizations, theater has become so sophisticated, so embedded in the culture, that people have forgotten what real innovation looks like.

It looks like shipped products. Measurable results. Engineers who understand the problem deeply enough to solve it creatively. Leaders who understand technology well enough to distinguish signal from noise.

It doesn’t look like open offices and innovation labs. It doesn’t sound like “we’re leveraging AI to transform the business.” It doesn’t come from consultants who bill by the hour.

Real innovation is harder than theater. It requires technical literacy, patience, good engineering, and tolerance for short-term failure on the path to long-term results.

But it’s the only kind that actually matters.

John Brandon Elam is a Product Owner specializing in decision intelligence systems. He co-authored “You Got the Data Job…Now What?” and focuses on cutting through organizational theater to build tools that actually work.

Follow John and Adam DeJans Jr. at Bit Bros LLC for more on digital transformation, decision systems, and building software that doesn’t suck.

These observations reflect patterns I’ve seen across the industry and conversations with professionals in multiple organizations. All opinions are my own.