What Eighteen Months on the Buy Side Taught Me About Services Businesses
What eighteen months advising private equity buyers taught a former operator about how services businesses are really valued.
I spent the better part of two decades building and running corporate services and fund administration businesses. Country roles, then regions, eventually a third of a global firm. I thought I knew what a good services business looked like from the inside.
Then I crossed the table. For the last eighteen months I have been advising private equity firms as they evaluate businesses in this sector. Reading the materials, testing the numbers, forming the view a buyer forms. I will not talk about any specific process or client, and I never will. But I can talk about what the view from that side of the table does to your thinking, because it changed mine.
Much of what we report internally as commercial success, a buyer discounts within the first hour. The things a buyer prices most highly are often, though not always, the things operators track least. And that "not always" usually comes down to leadership experience.
Logo retention is gravity
Every services business I have ever seen presents its client retention rate with pride. Ninety-five percent, ninety-seven percent, near zero churn. On the buy side you learn to acknowledge the number and then look past it. In fund administration and corporate services, switching providers mid-cycle is painful, disruptive, expensive and rare. Clients stay because leaving is hard. High retention is the structural condition of the industry. It is gravity. Claiming credit for it is like an airport hotel claiming credit for guests sleeping overnight.
That is not to diminish the real work that goes into keeping clients happy. A great deal of genuine effort happens every day. But from the buyer's seat the retention statistic itself tells you less than it appears to. The number that actually reveals whether clients are happy is share of wallet. Are they giving you more of their business over time, across more services and more jurisdictions? A client can be quietly dissatisfied for years while their retention statistic stays perfect. They simply give all their new work to someone else. Wallet share is NPS with money attached.
When I ran regions we tracked satisfaction surveys with a mixture of fear and pride. Share of wallet mattered too, but never as much as it should have. Looking back, I had the priority the wrong way round. The survey is an early warning system; the verdict is in the wallet.
Revenue per person is the tell
Revenue per person never appears in a client newsletter, and it dominates every buy-side conversation. In this sector it ranges roughly from $50,000 at labour-heavy offshore operators, to somewhere around $120,000 at the established mid-tier, to north of $250,000 at the most technology-enabled platforms.
That spread tells a buyer how the work is actually done, how exposed the model is to wage inflation and AI disruption, and how much transformation upside or risk they are underwriting. Two businesses with identical revenue and identical margins can sit at opposite ends of that range, and they are not remotely the same business.
If you do not know where your business sits on that curve and which direction it is moving, someone evaluating you will work it out in an afternoon, and they will price it whether you have thought about it or not.
The meeting that happens versus the meeting that produces
Most sizeable services firms now run some version of a key account programme. Named sponsors, client directors, quarterly business reviews. The architecture is usually sound; the execution is where the value leaks.
There is a world of difference between a quarterly review that happens and one that produces. The first is a satisfaction ritual: service metrics presented, pleasantries exchanged, actions noted. The second is a commercial event: whitespace mapped against everything the firm could be doing for that client, gaps discussed openly, expansion opportunities worked as a pipeline with the same discipline as new-logo sales.
From the buy side you can tell within a few management meetings which version a firm runs. Ask what percentage of new business comes from existing clients, then ask to see the account plans. If most growth comes from the base but the plans are thin, the growth is happening by accident. Accidental growth is real revenue, but a buyer will not pay for the machinery behind it, because there is none.
The delivery question has changed
A buyer does not stop at the commercial questions or the polished sell-side packs. Sooner or later they pull apart how the work actually gets done. For twenty years the delivery model conversation in this sector was about labour arbitrage and process optimisation. Where can the work be done cheaper and more efficiently. I built delivery centres myself and I would defend every one of them. But the question a sophisticated buyer asks today is different: is your offshore scale a moat or a liability?
If the work moved offshore is fundamentally rules-based processing, it is exactly the work most exposed to AI over the next few years. A cost advantage built purely on labour arbitrage is a melting asset. The delivery models that will command premiums are the ones where technology does the processing and people do the judgment, the relationships and the exceptions. That transition is expensive and awkward, and businesses priced on legacy margins have every incentive to delay it. Buyers know this, and it is now one of the first things they test.
Run it as if a buyer were reading it
You do not need to be selling a business for any of this to matter. The disciplines that survive diligence are the same disciplines that serve clients: knowing your share of each client's wallet and working it deliberately, knowing your revenue per person and improving it structurally, running account reviews that produce rather than perform, and building a delivery model whose advantage compounds rather than melts.
The best operators I have met run their businesses as if a knowledgeable buyer were reading the numbers every quarter. Not because they plan to sell, but because the buyer's questions are simply the right questions, asked without sentiment.
It took me eighteen months on the other side of the table to see that. I would have been a better operator if I had crossed it sooner.
Originally published on LinkedIn, 25 August 2026.
