Three Tonnes of Supplies, Thirty Kilometres a Day

More than one founder has asked: why are you always counting headroom?

Our view first. Headroom is not caution, it is the only way to turn "probably works" into "certainly arrives". Why? Because what goes wrong is never the plan itself.

A little over a century ago, two teams set out for the same South Pole. Amundsen had fewer men and three tonnes of supplies, and covered thirty kilometres a day in good weather and bad; he reached the pole and returned safely. Scott had more men and one tonne of supplies, a third of the other, enough on paper; his team was lost to the last man. The difference was not headcount. It was headroom. Supplies are not stocked for the days that go well.

So what counts as headroom? Only what can actually be consumed.

Three times over is the ruler for headroom

An example, with illustrative figures. A company wants to raise a billion. The discipline we set ourselves is to negotiate towards three times that in term sheet coverage. It sounds excessive. But between a term sheet and money in the bank sit due diligence, the investment committee and terms negotiation, and at every gate people drop out. Most of them do not arrive, and that is precisely why headroom means something. Three times over is the insurance that turns "signed" into "in the bank". With two extra alternatives in hand, a founder can stand straight when negotiating terms.

The other way round, not all piling-on is redundancy. One more chip on the board, one more version of the deck, one more batch of unrelated investors — that is usually not redundancy, it is insufficient optimisation. Real headroom is like-for-like, substitutable and able to step in. Everything else is luggage.

Pacing is harder than sprinting

Thirty kilometres a day is a pace. Good weather, thirty. Bad weather, thirty.

Someone will say the window is open, so sprint — the skill in a bad market is getting done what cannot be done, and the skill in a good one is confirming the lead in the shortest possible time and wringing the window dry. That holds. But what gets wrung dry is that company's window, not our pace. The sprint belongs to the project. The pace belongs to the firm. A deal that is eight or nine tenths done and standing in front of goal needs an undivided block of a partner's time, and that block has to have been kept empty long in advance. It cannot be grabbed at short notice.

So the number of mandates is strictly limited. Glacier takes twenty or thirty a year, half new, half existing clients coming back for the next round. One more mandate takes headroom away from the companies already in hand. A limited capacity is not a pose, it is arithmetic: judgement, research, coordination and a partner's attention are all non-renewable. Spread across a hundred projects, it is a hundred broken promises.

Do we miss things because of it? Yes.

Then we miss them. Every year good companies finish their rounds with someone else. But missing is the cost of this approach, not a hole in it. As we see it, the first duty is safety, not odds. A company's win rate can be talked up to ten out of ten and the odds may be down to half — count the redundancy first, the return second. On this one we are fairly confident.

Arriving is not the end. Getting back to camp is. A closing is not the end either. The company still standing in the next cycle — only then is the journey finished.

(On why service capacity is limited, see the entry of that name in the Archive.)