The capex approval process: why your champion cannot say yes

The pilot worked. Your champion is genuinely enthusiastic, the kind of enthusiastic where they start sentences with "once we have this". You agreed on next steps. Then the deal went somewhere you cannot follow, and the updates got shorter and vaguer until they stopped.

Where it went is a document. A capital request, written by your champion, reviewed by people who have never spoken to you, measured against projects you have never heard of. Most deeptech founders lose industrial deals in that document without ever knowing it existed. This is how it works, and how to stop losing there.

Your champion cannot spend the money

The engineer who ran your pilot, the plant manager who wants your machine, the R&D lead who keeps introducing you to colleagues. None of them can say yes. They can say "I want this", which is a different sentence. Converting "I want this" into a purchase order is a separate process with its own owners, and it does not care how the demo went.

In most manufacturers, capital spending is approved in tiers, and the tiers are set by the amount. Under roughly 25k, a plant or department manager can usually sign. Up to around 100k, a site director with finance countersigning. Above that, divisional or group management, then a capital committee, then the board. The numbers vary by company. The structure does not.

Every tier adds two things: weeks, and a person who has never met you. Your champion's enthusiasm travels up that chain only as far as it is written down, and it is written down by someone who has never done it before.

The business case gets written at night

Here is the scene nobody tells you about. Your champion, an engineer with a day job, at the kitchen table at ten in the evening, filling in a corporate template that asks for things like "risk-adjusted internal rate of return" and "impact on working capital". They have an engineering degree. They have never been trained on this form. Nobody has.

A weak case is the most common reason a technically won deal dies, and it is the most fixable one. The suppliers who convert are the ones who draft the case themselves and hand it over for editing. Not because champions are lazy. Because you have written forty of these and they have written none.

Finance does not price innovation

What does the reviewer on tier three actually do with the document? This is one of the few questions in business with a genuinely good answer, because John Graham and Campbell Harvey at Duke asked 392 CFOs directly. Their survey, The theory and practice of corporate finance, published in the Journal of Financial Economics, found that about three quarters of CFOs always or almost always use net present value or internal rate of return when evaluating projects. More than half still use the payback period as well, and the smaller the company, the more likely payback is the main tool.

Note what is not on that list: novelty. Strategic importance gets a nod. How advanced your technology is does not appear anywhere. Finance prices cash flows and the confidence attached to them. Which means your pilot data matters only insofar as it becomes a number in their units: scrap percentage, yield points, downtime hours, energy per unit, labour hours. "Efficiency gains" is not a number. "Eleven hours of unplanned downtime per month on line two, which you measured during our pilot" is.

The counter-argument: the gatekeeping is rational

It is tempting to experience all of this as bureaucracy defending itself. Resist that. Capital equipment is the one purchase a manufacturer cannot easily undo, and the process exists because the mistakes are permanent.

There is older evidence that the bar is set high deliberately. James Poterba and Lawrence Summers surveyed the CEOs of the Fortune 1,000 for MIT Sloan Management Review (A CEO Survey of U.S. Companies' Time Horizons and Hurdle Rates, 1995) and found hurdle rates far above what the cost of capital would justify: an average of 12.2 percent applied to constant-dollar cash flows. Companies were not miscalculating. They were rationing. More projects clear a sensible return than there is money, attention and management capacity to execute, so the threshold floats up until the pile fits the budget.

Your project is not competing against doing nothing. It is competing against every other request in the same pile: the roof repair, the forklift fleet, the ERP upgrade someone has been promised for three years. Arguing absolute merit loses to that. Arguing relative payback, in numbers from the plant's own floor, does not.

Ask the threshold question

One habit is worth more than everything else in this article. Early on, ask your champion: what can you sign yourself, and where are the approval lines above you?

I call it the threshold question, and it is amazing how rarely it gets asked. The answer changes your pricing more than any competitor does. If the site director signs up to 100k and the next tier sits at group level with quarterly meetings, then a 95k scope and a 105k scope are not five percent apart. They are three months apart, and one of them meets a committee that has never heard your name.

This is not an argument to underprice. It is an argument to structure: a first phase under the line, explicitly scoped as the first phase, with the expansion priced separately and approved separately. You get a purchase order, an installed reference inside the building, and a champion who now has evidence instead of enthusiasm.

The calendar runs the deal

Most manufacturers plan next year's capital during the second half of this one. Plants submit their lists in late summer or autumn, approvals land before year end, and money is released from January. (In the Netherlands this is near-universal, because almost every company runs a calendar fiscal year. Ask any Dutch plant manager when the investeringsbegroting is due and they will give you a date, usually somewhere in October.)

Two consequences. First, timing beats persuasion. Being in front of your champion in August, while they draft the list, is worth more than any meeting in March. Ask in the first conversation when the plant submits its capital plan, and work backwards from the date.

Second, missing the round does not kill the deal. It postpones it a year, which for a startup is the same thing wearing a coat.

The unbudgeted exception

There is a door next to the calendar. Companies can approve spending outside the annual plan, but the bar is higher and the approver is higher, so it needs a trigger that makes waiting expensive: a compliance deadline, a failing asset, a customer audit finding, a documented safety issue.

If your technology attaches to one of those triggers, say so explicitly and say it early. "This keeps you ahead of the regulation that bites in eighteen months" is an unbudgeted request. "This improves efficiency" is a queue position.

Write the case for them

Everything above converges on one job: make the document strong before your champion opens the template. The checklist is short and none of it is optional.

• Quantify the benefit in their units. Scrap, yield, downtime, energy per unit, labour hours, warranty claims. Not "efficiency".
• Use their baseline. Numbers from their own reporting during the pilot are unarguable. Numbers from your brochure are not.
• Be conservative and say so. A case built on the low end of measured results survives scrutiny. An optimistic one invites a discount and then fails the hurdle.
• Include the full cost. Installation, integration, training, spares, downtime during changeover. One omitted cost makes every other number suspect.
• Show the do-nothing option. What the current situation costs over three years is often the strongest line in the document.
• Name the risks and the mitigations. Staged payments, acceptance criteria, a factory acceptance test, performance guarantees.
• Attach the evidence. Pilot report, test data, reference installation.

Common questions

What is the capex approval process?

The internal sequence a company uses to authorise capital spending: a written business case, finance review of the numbers, approval at tiers set by the amount, then requisition and purchase order. Your champion runs it; you never see it.

How long does capex approval take?

Typically three to twelve months, dominated by the budget cycle and the number of approval tiers rather than by the analysis itself. A request that misses the annual round usually waits for the next one.

What payback period do manufacturers require?

Commonly under two years for cost-saving investments, sometimes twelve months for non-strategic spend. The Graham and Harvey CFO survey shows payback remains a primary tool especially in smaller companies, alongside NPV and IRR.

Who writes the business case for a capital request?

The internal champion, on a corporate template, usually without training. Suppliers who draft the case for them, using the customer's own measured data, convert noticeably more often.

What is an unbudgeted capital request?

Spending outside the approved annual plan. It is possible but needs a stronger case, a higher approver and usually an urgent trigger such as a compliance deadline or a failing asset.

Can a purchase avoid the capex process?

Sometimes. Priced as a service, rental or consumable, the same money can fall under operating budgets with far lighter approval, which is one reason first engagements are often structured as paid pilots. It shortens the path, at the cost of leaving the capital conversion for later.

What this means for you

Industrial deals are not won in the demo. They are won in a document written at night by someone who does not do this for a living, reviewed against a pile of competing requests by people pricing cash flows. Selling to manufacturers means arming that person properly, asking the threshold question early, and knowing which month the calendar turns.

Related reading: how industrial buyers compare total cost of ownership, how a new supplier gets on the approved vendor list, and why a letter of intent is not a purchase order.

If your technology is proven and the buying process is what stalls, that is exactly the gap we work in. Tell us what you have built.