I've Been Watching Businesses Become Zombie Corporations. Here's How to Tell If Yours Is Next.

Your business is quietly dying - and your dashboards will never tell you.

Margins are up. Headcount is controlled. Every metric is green.

I've spent 14 months evaluating businesses to buy - fire safety, HVAC, building compliance, training, education. I talk to owners, review financials, and look for what makes a business actually worth acquiring. And I keep seeing the same pattern that nobody's discussing openly enough.

Businesses deploying AI aggressively look exceptional on paper for 2-3 years. Then something shifts.

The metrics stay green. But the organisation loses its ability to respond, adapt, or even notice when things go wrong.

I call it the Zombie Corporation. It walks. It produces output. But the intelligence that kept it alive has been systematically removed.

Here's exactly how it happens.

Year 0-2: The Win That Seeds the Problem

The business automates the obvious stuff first. Scheduling, compliance documentation, reporting. EBITDA improves. Leadership feels validated.

This part is correct - these are the right things to automate first.

What nobody's tracking: the roles being eliminated weren't just cost centres. They were the early warning system.

The scheduler who knew which clients were quietly frustrated before it showed in churn. The coordinator who noticed a student disengaging three weeks before dropout. The junior engineer whose questions caught assumptions seniors had stopped questioning.

That intelligence doesn't live in a process document. It lives in a person doing a repetitive job who also happens to notice things.

Year 2-4: The Knowledge Walks Out the Door

Emboldened by the efficiency gains, the business goes further.

The veteran scheduler gets replaced by an optimisation algorithm. The experienced estimator by a pricing model. The senior teacher whose adaptations drove results by a standardised AI curriculum.

This is where the real damage happens - and it's invisible on every dashboard.

Institutional knowledge - the "I've seen this before and here's what it means" - leaves with the people who held it. Nobody's capturing it because the assumption is the AI will be better anyway.

It will be better at the process. It has no understanding of the rationale.

And rationale is what you need when the process breaks.

Year 3-6: The False Signal Phase

This is the most dangerous stage because the business appears to be performing.

Margins up. Efficiency improving.

What the systems aren't detecting: customer satisfaction drifting before it shows in churn. Regulatory changes creating compliance risk the automated audit trail isn't flagging. Market shifts that require strategic pivots.

The humans who would have noticed these signals are gone, or have learned their judgment gets overridden anyway.

Either way, the signal is lost.

Year 5-10: Ossification

The business is now very good at doing something that may no longer be exactly what the market needs.

Processes are optimised. Costs are controlled. And it has almost no capacity to change, because the adaptive intelligence that would drive change has been systematically removed.

When disruption hits - and it will - the response is slow, expensive, and usually too late.

That's the whole game.

The 5-Question Audit (Run This on Your Business Today)

I use these five questions when evaluating any acquisition. They tell me more about a business's real value than 12 months of financials.

Can your team explain why the AI is making its three most consequential decisions?

If the answer is "we trust the model" - you have a structural accountability problem.

When did your last meaningful service or process change come from an employee?

Not from a client complaint. Not from a competitor move. From someone inside who noticed something.

If you can't remember, your internal signal detection is failing.

What % of your client interactions involve a human who could catch dissatisfaction before it becomes churn?

If that number is shrinking quarter on quarter, you're building brittleness into your revenue base.

Can you name the five most important heuristics your senior practitioners carry that aren't documented anywhere?

If not, that knowledge is one resignation away from being permanently gone.

If your AI systems were wrong about something important today, who would notice - and how?

If the honest answer is unclear, your feedback loop is broken.

EBITDA measures what the business is producing today. These five questions measure whether it will still be producing it in five years.

Businesses that score well here are worth materially more than ones that don't - regardless of current margin.

If You Score Badly, Here's Where to Start

Don't try to fix everything at once. That's how nothing gets done.

Start with question four.

Sit down with your two or three most senior practitioners this week - the people whose departure would genuinely hurt - and ask them to walk you through the five calls, decisions, or patterns they rely on that they've never written down.

Record it. Structure it.

That's the first step toward capturing institutional knowledge before it walks out the door with its owner.

It's not glamorous. It's not a tech project. But it is the single most valuable thing you can do before your next wave of AI deployment - because everything you build on top of undocumented knowledge is building on sand.

The Stakes Are Higher Than Most People Realise

Here's something that doesn't get said enough, especially in Australia where the government released its Guidance for AI Adoption in October 2025 and stood up a new AI Safety Institute shortly after.

Mandatory guardrails were proposed, debated, and the direction of travel is clear even where specific legislation hasn't landed yet.

As AI-driven errors compound across industries, insurers and regulators are starting to ask the same question: can this business demonstrate a qualified human was in the loop when something went wrong?

A business that has automated away its human judgment layer may find itself in a position where professional indemnity premiums become prohibitive. Where an operating licence becomes conditional on human oversight structures it has already eliminated. Where a contract requires accountability it can no longer provide.

The "optimised" business isn't just brittle. It can become uninsurable.

Not because of anything it did wrong, but because the regulatory environment is being shaped by AI errors across the entire industry.

Building resilience today is, in concrete terms, keeping yourself licensable and insurable tomorrow. That's not a soft argument. That's a balance sheet argument.

The Uncomfortable Truth for Tech Companies

Tech companies are already seeing this play out - entry-level hiring down since 2022, senior leaders publicly questioning whether AI productivity gains were worth destroying the junior pipeline. The pattern is consistent across industries.

What Actually Works

The businesses winning with AI aren't the ones automating the most. They're the ones automating the right things while protecting everything that makes the business worth automating in the first place.

The trade business owner who's spent 25 years building licensed practitioners, running informal apprenticeships, and maintaining client relationships competitors can't replicate - that person has more AI-ready infrastructure than they probably know.

The question is whether they deploy AI to amplify what they've built, or allow the wrong advice to automate it away.

I work in technology delivery leadership for major companies while actively evaluating businesses across completely different industries. The pattern is consistent: organisational foundations matter more than technical sophistication.

The businesses I'd actually pay for aren't the ones with the shiniest tech stack. They're the ones where the owner can name their five most important heuristics, where junior staff still exist and are learning, where humans still catch the early warning signals before they become dashboard problems.

That's what separates a business worth acquiring from a zombie corporation that just hasn't realised it yet.

Run those five questions on your own business this week.

Which one makes you most uncomfortable - and what are you actually doing about it?