Incumbents Don’t Lose — and How Challengers Replace Them

Executive overview

A general counsel needs to choose a firm for the matter that will define her year, a matter her board will watch closely. Two names are in front of her. One is the firm she has used for a decade, in this job and the one before. The other is sharper and hungrier: stating to the trade press that they are coming for those who traditionally win this sort of work.

From what the GC has in front of her, the second firm is equal on the merits and cheaper.

Still, she surprises no one when she chooses the firm she has always used. She doesn’t stick with the incumbent because she thinks it’s the better firm. She sticks because of what would happen if the sharper, hungrier second firm made a mistake. In that case, the board’s first question would be, “Why did you change firms?” And she has no answer she would care to give.

That decision — made the same way, by nearly every client we modelled, on nearly every important matter — is why incumbent law firms are the Goliath in this story. Goliath here doesn’t mean big; it means established. Goliath is whichever firm the client can already defend choosing, whether that standing comes from a long relationship or a name that raises no questions. One client’s Goliath might be a boutique’s regional rival, or one global firm holding work another global firm wants. Size is relative to the challenger, but the advantage is not.

The headline finding of this research is not that Goliath doesn’t lose. You know that already. The value of this intelligence is the anatomy underneath that knowledge: where the armour is solid and where it isn’t. For a start, there is one exposed point almost no challenger firm aims at.

Five things stand out about Goliath.

1. 1

   The armour is specific to high stakes. Work that might reach the board stays with the incumbent. This is the strongest signal we measured anywhere in the study. Routine work, though, is open. The grip of the incumbent comes from the stakes of the matter, not from the client’s view of the incumbent.

2. 2

   Clients are protecting themselves, not judging the firm. The part you knew from experience now has a number on it. Asked what stops them choosing a challenger for a high-profile matter, clients converged hard, close to nine in 10, on their personal risk: the risk of being the one who switched if it later went wrong. Concerns about the challenger’s capability barely surfaced. (Clients are not talking to challengers they don’t think are capable.) The value isn’t that this is new. It’s that it’s now measured firmly enough to build a pitch around; almost no firm does.

3. 3

   Incumbents lose by stumbling, not to a better pitch. On the rare occasions clients moved high-stakes work, the trigger was the incumbent’s misstep. Incumbents’ mistakes were behind at least two-thirds of the moves in the research. Winning high-stakes work on the strength of a pitch against a qualified incumbent almost never happened.

4. 4

   The proof clients want is the one you’ll resist offering. Across responses to several different questions, clients volunteered the same demand: show me how your firm behaves when a matter goes wrong. All firms will include their shiniest matters, but the desire of clients to know about bumps in the road came up unprompted for more than three-quarters of clients.

5. 5

   Challengers are ruled out on proof and fit — not price. When a challenger was dismissed from consideration early, one reason came up more than any other: a track record that didn’t match the size and complexity of the client’s world, behind more than half of early rejections. Sector and regulatory fit came a close second. Price was almost absent as a reason for ruling challengers out of the process quickly.

Goliath is real. He is also not invulnerable. This briefing is the diagnosis: Goliath’s armour, and why it holds. Part 2 of the findings — David — is the route to the inch that isn’t armoured.

How to read this research

This research uses Asymmetric Strategic Intelligence (ASI), a method that reads how a market thinks. These are Client Proxies: models of the Australian legal buyers who choose law firms, grounded in published research. They are not real clients — and every quotation in these pages is a Client Proxy’s words, not an interview.

The method is built to show how a decision is made, not to count heads. Read every figure as the strength of a lean: a high number tells you a gale is blowing. It doesn’t mean nobody walks into the wind — it tells you which way it is easier to walk.

This is a directional read, sharp enough to act on and worth testing against your own clients before you stake the firm on one.

How we did this, and what the method can’t do →

Steven Lewis ran this research. steven@taleist.agency

Where the giant holds

To map Goliath’s armour, we gave each client the same choice twice. Once for a matter their board would see: the kind of high-profile work a general counsel could end up explaining across a boardroom table. Once for routine work that never leaves the legal team. Same client, same challenger in consideration. The only thing in the scenario that we changed was the stakes.

On the high-stakes matter, the client stayed with the incumbent. This was the strongest signal in the whole study: as close to unanimous as the method produces. When a matter might reach the board, the incumbent is almost certain to keep it.

That is not the same as the incumbent winning everywhere. It is a firm that is almost immovable on the matters that carry the most weight. As the next finding shows, that same firm is far from immovable on the rest of the work.

“If I moved this to a new firm and something went wrong, the first question I’d get is ‘why did you switch firms mid-stream on this one?’ I don’t have a good answer to that.”

Routine work is a different story

Change the stakes and the picture changes with them. On routine, everyday work — the standard matters that fill most of a legal team’s year — the resistance falls away. The same clients who would not put a high-profile matter near a challenger were consistently open to handing the routine work to a new firm in the right circumstances.

This is a critical finding. If the incumbent’s hold on the client came from being the better firm, it would hold on the everyday work too. It doesn’t. The incumbent’s grip is fixed to one thing: whether the person making the decisions for the client is personally exposed if the matter goes wrong.

Why the incumbent keeps the work

The risk clients are protecting is their own

We asked clients directly what would stop them moving a high-profile matter. Their answers converged hard, close to nine in 10, on a single thing; it had nothing to do with the challenger’s ability. It was their personal exposure: the risk of being the person who switched firms on the matter that later went wrong.

Clients seldom led with capability as the reason to keep work with the incumbent: fewer than one in 20. That isn’t indifference. Capability is the price of entry; every firm in the conversation is assumed to be good enough. The reasoning underneath was consistent: capability gets a firm considered, but defensibility gets it appointed. Being good enough is the ticket into the room. What wins the work is a decision the client can defend afterwards.

The worry about personal exposure didn’t belong to a nervous few. It held across clients new to their role and clients a decade in, across large organisations and smaller ones. We went looking for a “cautious client” type: a risk-averse group a challenger could route around. We didn’t find one. That isn’t proof no such group exists. It means there’s none big enough to plan around.

But not everyone carries it equally well. The fear is much the same at every level of seniority; what differs is how much credibility the client has banked to withstand it. More than anything else, that came down to how new they were to the job. Among general counsel in their first couple of years, the worry that a switch gone wrong would land on them personally was the dominant note. For peers with deep roots in the job, the fear barely surfaced. The fear bites hardest on the people with the least cover. It is the one finding here we would most want tested against your own client list before you build around it: tenure is easy to read off a LinkedIn profile, and if the pattern holds in your market it tells you which conversations to have first.

“Some of that credibility comes from being reliable and not creating surprises. If I switch firms on a high-profile matter and anything goes even slightly wrong, I’ve created a problem for myself that didn’t need to exist.”

Incumbents lose by stumbling — challengers almost never win by pitching

When high-stakes work did move, it moved because the incumbent stumbled. The turnaround was slow, a deadline was missed or a partner drifted away from the file and the team churned below. This was the trigger in at least two-thirds of the moves the research surfaced.

The mirror image is as sharp. A challenger winning high-profile work on the strength of its pitch — rather than the incumbent losing it — almost never happened. Clients don’t go looking for a better firm for their most sensitive matters; they get pushed towards new firms by a poor experience when Goliath lowers its guard.

This finding has been tested on real people. A study of 166 chief legal officers at S&P 500 companies asked why they had ended or cut back work with a firm. Quality of service was behind 60% of terminations and 78% of substantial work reductions. Cost on its own accounted for 7%. The study covered exactly the kind of work this research is about: matters of strategic importance to the company. Different method, different market, different decade — same answer.

Coates, DeStefano, Nanda & Wilkins, “Hiring Teams, Firms, and Lawyers: Evidence of the Evolving Relationships in the Corporate Legal Market”, Law & Social Inquiry 36(4), Fall 2011, p. 999. DOI: 10.1111/j.1747-4469.2011.01259.x

“Our incumbent panel firm kept rotating juniors through the matter — a different face every few months, no continuity. The partner was across it nominally, but the day-to-day was a mess.”

The one proof a challenger can’t fake

We wanted to know what proof moves clients, and we came at the question from several directions. What proof would a client need before it would switch to a new firm? What claims would the client press hardest, i.e. where is proof most needed and what proof is sufficient?

We asked different questions and the same answer kept surfacing: show me how you behave when a matter goes wrong. Clients are used to hearing about firms’ big deals, their marquee clients, and the rankings and accolades every lawyer in the field has acquired the way a fridge collects magnets. Clients want to know about the messy matter. When things got complicated, what did you do? More than three-quarters of clients raised this.

This might be the incumbent’s deepest advantage. The client has watched the incumbent handle things going wrong — over years, on real matters. They have never watched the challenger do it. So, if you don’t give the client comfort about what will happen when things are tough, you’re putting yourself on the wrong side of the client’s personal-risk register.

“Show me a matter you ran that went sideways. Not a polished win story. Anyone can win cleanly. I want to see how you behaved under pressure, how you communicated, what you did to fix it.”

The second obstacle: getting past procurement

A client might be open to a challenger but facing an internal obstacle, their company’s procurement process. The procurement process is seen as a cost over and above the cost of onboarding a new firm for seven in 10 clients with a formal procurement process. Clearing procurement is one more thing the client has to justify internally, which means the procurement process is another way their reputation is on the line for backing the unknown name. Procurement is friction and fear in one place.

Ruled out on proof and fit — not on price

When clients talked about dismissing a challenger firm out of hand, the reason was rarely about the work itself. It was about the client believing the firm’s track record didn’t match the size and complexity of the client’s world. That sat behind more than half of the rapid rejections: more than any other reason. A close second was sector and regulatory fit, raised by around half the market.

Being too expensive was almost never why a challenger was ruled out quickly. Fees don’t raise doubts; doubts come from lack of proof.

“If I’m sitting in a panel review and a firm can’t point to comparable transactions, I can’t take the risk of being the person who put them forward and then watched them struggle.”

Price isn’t absent from the process. It can keep a firm off a panel before anyone assesses its work, and it competes hard between firms that have already cleared the proof bar. What price cannot do is answer a doubt about proof. A challenger discounting into a credibility gap is paying to lose more cheaply.

Many challengers never reach the decision room

For a share of clients, the answer was starker still: no challenger had ever pitched them. The work has always gone where it goes, and no other firm has been in the room to argue otherwise. In the research, as many as one in five clients had never seen a challenger pitch at all.

Two things compound it. In regulated corners of the market, some clients expect a big-name firm on the serious work. That’s not because the big firm is believed to be better. Appointing the established name draws no questions. Appointing anyone else invites them. A champion inside the client has to spend real credibility to back an unknown firm.

Every firm knows the contest is hard to start, but not every firm knows where it’s hardest. In the research, the largest and most established in-house teams were the most immovable in the market. They were the least likely to have seen a challenger pitch, and the most likely to rule out a challenger for never having swum in equivalent waters. Test that against your own market before you act on it. It might mean the firms you’d assume are your warmest prospects are the ones to approach last.

“No challenger has even got in front of me.”

“I haven’t built enough internal credibility yet to absorb a failure that gets attributed to an unconventional choice.”

The way through

Here is the giant in full. On work that might reach the board, the incumbent is almost certain to hold. This isn’t because the challenger is weaker, but because the client is protecting themselves. Work moves when incumbents stumble, not when challengers shine. A new firm has to clear the client’s own governance to get in, which requires someone to spend personal capital. And challengers are shut out on proof and fit, almost never on price.

Most challenger firms spend years swinging at the wrong part of the giant. They sharpen their pitch and trim their fees. They would do things differently if they could see where Goliath is exposed. The research is just as clear about where the armour is weak. Five insights:

1. 01

   The one kind of client whose high-stakes work is within reach — and how to spot them early, so you don’t spend a year on the ones that were never available.

2. 02

   The move that puts your name on the shortlist for the matter that matters — before that matter even exists.

3. 03

   Why the right discount has almost nothing to do with being cheapest (and the exact point past which every dollar you cut starts costing you the work).

4. 04

   The one proof that turns “we’ve never heard of them” from a reason to rule you out into a formality on the way to yes.

5. 05

   The reframe that lifts the personal risk off the general counsel’s shoulders entirely (stopping choosing you from being the brave call and making it the obvious one).

This briefing was the diagnosis. Which of the five insights above matters most depends on your firm, your market, and the client you mean to take from an incumbent. That’s why Part 2 is a private briefing specific to your firm, not another document: an hour with the person who ran the research.

Book the hour →

Or write first, and say which of the five you’d want to start with: steven@taleist.agency

Steven Lewis · +61 413 829 909

Goliath isn’t going anywhere while you wait. But there is one unarmoured inch — and one move that reaches it.

About the author

Steven Lewis has seen the decision described in this briefing from three sides. In the early 1990s, he was the client, instructing Clifford Chance, Simmons & Simmons and other law firms across Asia, the United States and the Middle East. Afterwards, he spent three years as commissioning editor of two legal journals for Sweet & Maxwell, publishing the material those firms read. And in the early 2000s, he worked in business development for Clayton Utz, on the other side of the same conversation. He has an MA in Jurisprudence from Oxford, and spent five years as a journalist, writing on technology for the Financial Times and the International Herald Tribune. Today, he is the founder of Taleist, an intelligence agency for professional services.

The research behind this briefing was produced with Asymmetric Strategic Intelligence (ASI), a method Taleist has been developing since February 2021 — before ChatGPT existed. ASI isn’t an “AI tool”, and it didn’t come out of a technology business. The method, and the consultancy around it, came out of thirty years of direct-response copywriting. Those thirty years were spent on two questions: what makes people decide, and how you turn the answer into action. Those are marketing questions with research at their core, not AI or technology questions.

The technology still has to work, of course, so ASI draws on academic work at Stanford and elsewhere testing how accurately language models reproduce the attitudes of real people. That is the academic evidence. The commercial evidence is one sentence long: every organisation that has commissioned an ASI study has commissioned another.

↑ Back to the briefing

Appendix A — What we asked, and how it maps to the findings

A plain guide to the decisions we put to the Client Proxies, and which finding each one informs.

Appendix B — How we did this, and what the method can’t do

The panel, the method behind it, the published work it draws on, and the boundaries of what it can tell you.

The market is Australia: the Client Proxies were built and primed on the Australian corporate legal market.

For this research we built 334 Client Proxies: modelled decision-makers who reason the way specific people in this market reason — general counsel, senior in-house lawyers, legal procurement leads and board-level decision-makers. Conventional research would poll several thousand people and average their answers. This does something different. A survey counts opinions; a Client Proxy reconstructs the reasoning beneath one: how the decision is actually made, not merely which box gets ticked. Three hundred Client Proxies carry more signal about how this market thinks than three thousand ticked boxes, because we are reading the structure of the decision, not taking its temperature. A survey asks the room to raise their hands. This method brings you into the room where the argument happens before the vote.

Two studies have tested the approach behind ASI: one on the quality of the answers it produces, one on how closely it can track a specific person. The first, published in the Journal of Marketing in 2025, put modelled respondents and real people through the same in-depth interview and had independent evaluators rate the answers. The evaluators judged the modelled answers more detailed and more insightful than the human ones. The same study is honest about the method’s limits, and so are we. The modelled answers varied less than real people’s, which is why the figures here describe how strongly a market leans rather than how many people would act. And the method is better at reading a whole market than at telling one group from another. So where a finding turns on a difference between groups, we flag it in the text as something to test with your clients.

The second study, from Stanford, put the same approach to a harder question with 1,052 people: whether a model can predict what one specific person will say. Ask someone the same question two weeks apart and they agree with themselves about 81% of the time. That sets the ceiling on what any method can know about how one particular person thinks. Modelled respondents got most of the way to that ceiling, and how close depended on how richly each respondent was specified.

Both studies point at the same thing. What makes modelled respondents behave like real ones is the depth of context they are given about the world they are reasoning in. That is the part of this method we have invested the most in.

The same method is in use at a global law firm and an international accounting firm, on decisions they can’t afford to guess at.

Where a finding could be checked against published research with human subjects, that’s noted in the text. The full references are in Sources.

More on Asymmetric Strategic Intelligence →

Sources