The Demo Is Not the Due Diligence

The Most Expensive Advice You Never Gave

There is a call you haven't gotten yet.

Your client discovered a tool that promises to automate their payroll, chase their invoices, forecast their cash position, and manage their workflows. They are excited. The demo was compelling. The vendor made it look simple.

They are calling to tell you they signed up.

Or they are not calling at all.

That second scenario is the one worth thinking about.

The Numbers Behind the Excitement

Before your client commits, consider what Bain & Company found when they surveyed CFOs — financial leaders with dedicated teams, implementation budgets, and professional staff — about their AI investments.

Only 31% rated their AI outcomes in finance as strongly positive.

These are professionals whose entire function is financial oversight, with every resource advantage your client does not have. Strongly positive outcomes less than a third of the time.

Bain noted they are pressing forward anyway — not because early returns have been spectacular, but because the gap between organizations that have scaled AI and those that haven't is becoming too large to ignore.

That is not a success story. That is fear of being left behind dressed as strategic urgency.

Your client is feeling the same pressure with considerably less margin for error.

What the Data Is Already Telling You

You already have the most important input to whether this purchase makes sense.

Not a new diagnostic. Not a formal process. What you already know from working their books.

The client who cannot give you a clean explanation of how revenue becomes cash — where it slows down, what holds it up, what the timing looks like week to week — has a process problem that no automation tool resolves. You have been navigating the downstream symptoms of that problem every engagement cycle. Inconsistent AR aging. Revenue that does not convert on any predictable timeline. Margin that moves without a clear explanation.

That is not a bookkeeping problem in isolation. It is a process health signal. And it is telling you something important about whether this client's financials are stable enough to become reliable automation inputs.

The same is true for workforce complexity. A client running hourly employees, salaried staff, 1099 contractors, and W-2 workers — potentially across citizenship and residency classifications — without a clear understanding of how each category needs to be handled for payroll and onboarding is not ready for payroll automation. You already see this in the recurring manual corrections. The exceptions that come up every cycle. The categories that do not reconcile cleanly.

The vendor demo did not show any of that. You already know it exists.

What the Data Does Not Cover

Financial data is one lane. It is your lane, and you know it with authority.

But workflow automation tools frequently touch HR, inventory, operations, and distribution as well. Those domains have their own data integrity requirements and their own process health signals — and they require their own domain expertise to read.

A 1-2 hour consultation across the relevant domain advisors costs a fraction of the annual subscription commitment your client is about to make. That calculation gets starker when implementation time, configuration effort, data migration, and the remediation work required to reach the value the demo promised are added in. Tools that go in before the data is ready require additional investment to become functional — in cleanup, configuration, and ongoing maintenance. That investment is easier to justify when the foundation was sound from the beginning. It is harder to justify when the foundation is what needed attention first.

The consultation is not a delay. It is the cheapest decision your client will make this year.

The Signals You Are Already Seeing

You have seen the indicators before. You just have not always had a name for what they were pointing at beyond the immediate bookkeeping problem.

Categorization that shifts from period to period without a business reason. Margin that moves in ways the revenue does not explain. AR aging that has no discernible pattern — not because the clients are unpredictable, but because nobody ever defined what the collection process was supposed to look like. Payroll that needs manual correction most cycles. Expense categories that overlap or migrate between periods. Revenue that gets recognized on a timeline that changes depending on who touched it last.

You have been working around these. Every engagement cycle. Quietly adjusting, cleaning, correcting.

What the automation conversation does is give those signals a second question to sit alongside the first.

The first question was always: what does this mean for the books?

The second question is: if a tool were running this process automatically, what would it be doing with this?

The financial outputs are one layer of the answer. HR, inventory, operations, and distribution carry their own. What you see is what you see. What sits underneath it may need someone else in the room.

What Automation Actually Does to a Broken Process

There is a line from process improvement work that has outlasted every generation of business technology it has been applied to.

If you automate before you improve, you are doing the same thing Ex-Lax does.

You produce shit faster.

Enterprise software. Cloud migration. Digital transformation. AI. The tool changes. The sequence problem does not.

Whether the tool touches the financials or the operations or both, it executes whatever process it is given. Faster. At higher volume. With greater consistency. It does not know the categorization is inconsistent. It does not know the payroll logic has exceptions nobody documented. It does not know the AR aging has no policy behind it.

It just runs. Confidently. At scale.

The Sequence — And Where You Fit In It

Three steps. No shortcuts.

Assess. Improve. Mechanize.

Assess is a discovery conversation, not a solo diagnosis. The financial data you hold is one input. HR, operations, inventory, distribution — each domain has its own signals and its own expertise required to read them. The contribution here is bringing what the numbers show and being clear about where the numbers stop talking. What sits in the other domains needs the right people in the room. The purchase decision waits until that conversation has happened — with people who are not invested in its outcome.

Improve is where the conversation gets difficult.

A resource-strapped client does not want to hear that the path forward requires additional time, effort, and outside help before they spend a dollar on the tool they were hoping would solve the problem. That resistance is real.

But the frame they already understand is the argument.

Prototyping the improved process with tools the client already owns is a one-time capital effort. Maintaining a tool that was deployed before the foundation was ready is a recurring operational cost — in cleanup, corrections, configuration revisions, and the downstream advisory time required to make sense of outputs the underlying process was never stable enough to produce reliably.

One-time CapEx to prove the concept. Recurring OpEx to manage the consequences of skipping it.

That math is already in their language.

If the prototype works — concept proven, no new spend required, client knows exactly what any future tool needs to do. If it does not work manually or semi-automatically, it would not have worked automated either. That answer cost a fraction of an annual contract to find.

And occasionally — more often than vendors would like — the problem gets solved entirely with what the client already has.

Mechanize is what comes after the foundation is stable and the process is proven. That is when the tool delivers what the demo showed. Because for the first time, the foundation actually supports it.

 The Advisor's Diligence

At some point your client is going to ask you what you think.

Maybe before they sign. Maybe after the first quarter of corrections that were not supposed to happen. Maybe when the forecast did not match reality for the third month running and the tool that was going to simplify everything has become its own category of problem to manage.

The conversation is available before the contract. That is the easier version of it.

There is one question that cuts through the vendor pitch, the solution engineer's cleanup estimate, and the fear of being left behind by competitors who may or may not actually be further ahead.

Are your processes, as they are right now, something that would benefit your business if they were done faster?

For a client already managing one operational or cash crisis after another, the answer is already in the data you are holding.

So is the conversation worth having before they sign?