Where the R&D money goes: tracking development cost by direction in defence tech

«We were sure direction A was our main one and poured resources into it. The first time we split costs by direction, it turned out A cost twice as much as B and was further from a contract. We were betting on the more expensive prototype blind.»

In defence tech a company almost always runs several developments in parallel. A drone, an antenna, a comms system — and all of them eat engineer salaries, components, test flights. The problem is that in the books this is one shared bucket called «R&D», and the owner cannot see which direction costs how much. And without that figure the decision about what to continue and what to close is made on a feeling, not on facts.

Why R&D cost is invisible

Team salaries go out as one sum for everyone. Components are bought now for one prototype, now for another, and in the bank statement it's just «payment to supplier». Test flights, range rental, part printing — all mixed into general expenses. In the end you know development took, say, $90K for the quarter, but you don't know how that $90K split across three directions. And without that you cannot decide which to close and which to fund — you manage the total budget, not the individual bets.

What needs an address by direction

To see the real cost, three things must get an «address» — a specific direction or project.

First, engineer time. Not to the minute, but at least as a share: this designer is 70% of the month on direction A, 30% on B. Labour is the biggest R&D line, and without splitting it any costing is incomplete. Second, components and materials: each purchase is tied to the prototype it was bought for, at the moment of payment. Third, direct test costs: flights, the range, consumables, repairs after failed attempts.

The same $90K budget looks completely different once you can see direction A eating more than half of it. Now the «close or continue» question is asked on numbers, not on a feeling of which prototype is «more promising».

Let's work through an example

Imagine a company with a quarterly R&D budget of $90K and three directions. By the founder's gut, direction A is the main one: it has the most attention and the best team. But once costs were split, A came out at $48K a quarter, B at $24K, C at $18K. Now add a second dimension — closeness to a contract. Direction B has already passed field testing and holds a letter of intent from a customer; direction A is still at the prototype stage with no clear timeline. So the most expensive direction is the furthest from money, and the cheapest, B, is the closest. Without a split by direction you can't see this: in the general report both are simply «R&D spend».

«While R&D is one line in the report, you see how much you spend but not what you're betting on. And the bet is the main decision in defence tech.»

Why this matters beyond curiosity

Direction cost drives three decisions. Prioritisation: limited resource goes where the chance-of-contract to cost ratio is best, not where it habitually goes by inertia. Grant reporting: donors and the state want to see where funds went by direction and stage, not in one line, and if accounting is kept by project from the start, the report assembles from the same data. Due diligence: the investor's first question is how much it costs to bring a prototype to a working state — and a company that answers with a number rather than a «roughly this much» guess looks entirely different and negotiates from a stronger position.

What it looks like in real life

The symptoms are audible in typical phrases inside the team. «We spend a lot on development but I can't say on what exactly.» «Seems like the drone eats the most, but I'm not sure.» «The donor wants a report by stage, and we have it all in one pile — we'll spend a week collecting it by hand.» «The investor asks the cost of a direction, and we answer approximately.» Each of these lines is about costs kept in one bucket, with the company blind to its own bets in numbers.

How to set it up

Splitting by direction does not require a separate finance department. In Finmap you create projects for each development direction and tie costs to them — salaries as a share, component purchases, test spend. The report immediately shows the cost of each direction for the period and the trend: which one gets cheaper with iterations and which only accelerates the burn. When it's time to report to a donor or prepare for a round, the data is already structured — no need to dig through a quarter of statements and recall which purchase went to which prototype.

Related — how to calculate the real unit cost of a drone and what changes in the finances when you move from prototype to series.

A few tips

• Don't chase minute-level precision. Even a rough share of engineer time gives a picture you don't have at all right now.
• Tie a purchase to a direction at the moment of payment, not at quarter-end from memory — otherwise you won't recall half of it.
• Watch direction cost over time: healthy R&D gets cheaper each iteration, sick R&D gets more expensive.
• Put two numbers side by side: direction cost and closeness to a contract. The decision is made at their intersection, not on one figure.
• Keep accounting by direction from project start, not «we'll tidy it up before due diligence» — assembling honest numbers retroactively is nearly impossible.

Several parallel developments is normal for defence tech. What's not normal is not knowing which one costs how much. The moment that number appears, deciding where to invest your last month of runway stops being a blind bet and becomes a calculation.

Money Doesn't Disappear. You Just Don't See It.

Want to see the cost of each R&D direction on its own, not in one bucket? Book a Finmap demo — in 30 minutes we'll show how to set up accounting for your developments.