1,070 words · 4 min read
Glacier CapitalEssays
Home/Essays/DEAL NOTES/NO. 5
DEAL NOTES · REPEAT BUSINESS · NO. 62 · THE ONLY PROOF

Repeat Business Is a Report Card You Cannot Fake

The third stage of the rocket cannot be manufactured: whether an institution will look at your next mandate.

Zhang Jiakang (JK) · Founding Partner, Glacier Capitalapprox. 714 words · 3 min readArchived 2026-08-16

After one closing, an institutional investor asked us: how do you tell whether a firm in your line of work is reliable?

Our answer was short: do not listen to what it says. Look at whether its clients come back.

Our position first. We built ourselves a three-stage rocket: whether a mandate closes, how many tables we can turn in a year, and how often institutions come back to invest. The first two stages can be worked on — pick your mandates and the first goes up, grind harder and the second goes up. Only the third cannot be manufactured. Why? Because we are not the ones who fill it in.

That third rate is simply whether an institution, once we have served it, will look at the next mandate. There is a plainer name for it: repeat business, or repeat clients. Praise can be politeness and a referral can be a favour. Only repeat business cannot be faked. Where an institution has been burned, it does not go back a second time. No mandate escapes this one.

Do the arithmetic first. We take twenty to thirty mandates a year, half new and half existing clients coming back for the next round — which means half of a year's capacity is bought by last year's delivery. The institutions we have closed with most often are almost all repeat clients. That half of the table is our denominator. When the denominator caves in, the first year does not hurt. The second year does.

Investors are not downstream of our pipeline. They are upstream of our thinking

Repeat business comes from ordinary days, not from the moment of a deal. Call investors only when you have a mandate in hand and the relationship stays stuck at the transaction: what they remember is a project, not a person whose judgement holds. Hand over a read on the market when you want nothing, and they will take your call next time.

This work has one problem: it carries no immediate return, so it is the first thing a busy week squeezes out. When things get busy, everyone wants to clear the mandate in hand first. Is it still worth doing? Yes. From 2018 to this year is eight years, and across those eight years it was exactly these people's most exacting doubts that forced us to make every mandate solid. The time saved gets paid back later at double.

Repeat business is an organisational matter. One person cannot do it

A project manager watches only the five or six institutions currently in hand, and the rest of the relationships are wasted by default. Carrying an investor from the last project to the next cannot be done on one person's memory. First, the standard of the materials has to be uniform, so what he receives is the same grade every time. Second, feedback has to be filed: who frowned at what price, so nobody has to ask again. Third, someone has to remember who owes whom a call back.

All of it is dumb work. But compounding grows on dumb work.

Eight years serving and accompanying more than 80 portfolio companies; the public register lists 58, and our own capital is co-invested in 19 of them — one in three on that register has us sitting at the same table. Without your own money in, you look at a deal differently. With it in, the short view can no longer hold you down. That is why we dare to talk long with investors.

Does anyone not come back? Of course. Some houses change strategy, some contacts leave, some are simply out of tempo with us. When the window has not come, everybody sits on the bench. That holds, but it is not the whole story — repeat business measures method, not luck. As things stand, it is still the least dilutable ruler we hold.

Some hundred-per-cent figures and rankings about us have circulated online. None of them came from us. We do not accept numbers like that. What we accept is this: three months after a mandate closes, the colleague who ran it calls that institution once, with no new project to sell, and asks one question about how the last one turned out.

That is it. Year after year, it adds up.