Clients stay because switching puts the decision-maker’s own career on the line, and staying needs no business case. In Taleist’s study of 334 Client Proxies modelled on Australian corporate legal clients (July 2026), close to nine in 10 named their own exposure as what stops them moving high-stakes work to a capable challenger.
What does the decision-maker risk by switching?
A general counsel who moves a board-level matter to a new firm owns that choice. If the matter goes well, nobody asks how the firm was picked. If it goes badly, she is the one asked why she switched. Keeping the incumbent carries no such question, because the board accepted that firm last time.
That is the reasoning the Client Proxies gave. Asked what stops them giving important work to a challenger that is just as capable, close to nine in 10 of the 334 pointed to their own exposure. It was the strongest theme in the answers to that question, and it held across every type of client the study modelled. One Client Proxy put it this way:
“If anything wobbles, I’m the one answering ‘why did you switch firms on this one?’”
Does all work stick to the incumbent the same way?
No. The study put the same choice to the same Client Proxies twice: once for a matter the board would see, and once for routine work that stays inside the legal team. On the board matter, none of the 334 chose the challenger. On routine work, more than nine in 10 were open to one. Nothing changed between the two questions except the stakes.
Read that split as a hypothesis. It compares two scenarios answered by one synthetic panel, and a synthetic panel can make a contrast sharper than real clients would. A zero means none registered; at this sample size, a very small share and none read the same. You can test it against your own records. List the matters that moved to a new firm in the past three years, mark the ones that reached the board, and see whether the moves cluster in routine work.
What makes a general counsel switch firms?
The incumbent slips. When the Client Proxies described what had moved high-stakes work in the past, the incumbent’s own failure was the trigger in at least two-thirds of the accounts: deadlines missed again and again, problems acknowledged and never fixed, a relationship that broke down.
Research with real clients points the same way. Harvard Law School’s Corporate Purchasing Project surveyed and interviewed 166 chief legal officers of S&P 500 companies in 2006 and 2007, about the years 2003 to 2006. Of those who had dropped a firm, 60% named the quality of its service as the reason. Ethics problems, particularly conflicts of interest, came next, and the authors note that cost played a smaller part than the usual story assumes. The same study found that dropping a preferred firm was rare: about a fifth of companies had done it more than once or twice in three years, and over 30% had not done it at all. When hiring, 97% said their most important source was personal knowledge of the lawyers or the firm.
The surveys tell you why clients leave. The Client Proxy work adds why they stay when nothing has gone wrong: the person making the choice is protecting their own position, and a capable challenger does not change that on its own.
How does a challenger win work from an incumbent?
Every firm is the incumbent on the work it holds and the challenger on the work it is chasing, often in the same week, so this answer cuts both ways.
On high-stakes work, the opening comes after the incumbent slips. And when clients consider a firm they haven’t used, the proof they ask for is uncomfortable to give. More than three-quarters of the 334 Client Proxies raised how a firm behaves when a matter goes wrong, across four questions that never mentioned failure. A credentials section full of successful matters doesn’t answer that.
For the firm holding the work, the same finding reads as a warning. The client is watching how you handle the matter that goes wrong, because that is when the work becomes movable.
The full diagnosis is in Incumbents Don’t Lose, Taleist’s briefing on the study.
What can’t this research tell you?
It is not a count of real general counsel. The figures are what 334 Client Proxies reported, so read them as how hard the market leans, not as shares of real clients. The reasoning and the language are the dependable part, because an argument doesn’t need the panel to match the market head for head.
It can’t predict who will move. Synthetic research is far weaker at predicting behaviour than attitudes. What it gives you is what the decision-maker values, and that is what a pitch has to answer.
The stakes split may be sharper than reality. Your own matter records can test it in an afternoon, as described above.
How do you find out what your own clients care about?
A client listening programme asks your existing clients for feedback, usually after a matter or once a year. It deepens the relationship while it gathers the information. It only hears from clients who stayed, and clients soften what they tell a firm they plan to keep using.
Independent client interviews put a third party between you and the client, which gets franker answers. They are the best way to settle a specific question, and they are limited by how many senior clients will give up an hour.
Synthetic research models the people who decide, including the clients you lost and the ones who never chose you, and puts questions to them no real client would sit through. Taleist’s ASI is one provider. It shows how clients reason, not what a given client will do, so it works best before the interviews, to find the questions worth asking.
Sources: Taleist, Incumbents Don’t Lose. Part 1: Goliath, 334 Client Proxies modelled on Australian corporate legal clients, fielded July 2026. Harvard Law School Program on the Legal Profession, Corporate Purchasing Project: How S&P 500 Companies Evaluate Outside Counsel, 2011 (survey and interviews 2006–2007).