What leaders usually conclude
The obvious read is internal rivalry — territorial behaviour between groups that see each other as competitors for credit rather than colleagues serving the same firm. Rivalry of this kind does exist in some organisations, but assuming it here skips past a much more mundane and more fixable explanation.
What the evidence suggests
Each group has its own revenue target, and there is no mechanism for splitting credit on work that one group refers to the other. Referring work reduces the referrer’s own number without adding anything to it. Steven Kerr’s research on reward systems, examined across a manufacturing company, an insurance firm, and settings ranging from politics to medicine, describes exactly this pattern: organisations often reward one behaviour while stating a hope for a different one, and people respond to the reward, not the stated hope. As Kerr put it, “most organisms seek information concerning what activities are rewarded, and then seek to do… those things, often to the virtual exclusion of activities not rewarded.” A firm that says it values cross-referral while measuring and rewarding only individual group revenue has, in effect, built a system that makes referral irrational for the people inside it.
The knowledge management dimension follows directly. As a firm grows and its practice groups specialise, referral and cross-group knowledge sharing become more valuable to the client but structurally harder to sustain, because the compensation architecture that once didn’t matter — in a smaller firm with fewer groups and less specialisation — now actively discourages the behaviour the firm most needs.
Jay Galbraith’s research on organisation design offers a complementary lens on why this is so difficult to correct informally. Galbraith describes “lateral relations” — deliberate mechanisms for connecting otherwise separate parts of an organisation — as a necessary structural response once an organisation grows past the point where informal, proximity-based coordination can carry cross-group knowledge on its own. A small firm’s practice groups sit close enough, physically and socially, that useful precedents and techniques travel between them almost by accident, through casual conversation and shared social occasions. A larger, more specialised firm has usually lost that proximity without building any deliberate replacement for it, leaving cross-group knowledge sharing dependent entirely on the compensation incentive Kerr’s research describes — which, as already established, actively discourages it.
There is a further, less visible cost worth naming: knowledge itself becomes siloed alongside the revenue. A specialist insight developed in one practice group — a useful precedent, a client-handling technique, an efficient way of approaching a recurring problem — has no natural channel for reaching a second group that could use it, precisely because the two groups have limited reason to interact at all. The firm loses not only the specific referrals that never happen, but the compounding value of cross-pollinated expertise that a genuinely integrated firm would generate almost as a byproduct of routine collaboration.
Competing explanations
A firm might look at the same data and conclude the groups simply have poor working relationships, and invest in team-building or joint social events to fix it. This can help morale, but if the compensation structure is unchanged, the underlying arithmetic that makes referral costly to the referrer remains exactly as it was — and the behaviour is likely to persist regardless of how well the two groups get along personally.
It is also worth noting what this pattern is not: it is not evidence that the individuals involved are uncooperative by disposition. Kerr’s own examples span manufacturing, insurance, politics, medicine and academia precisely because the mechanism is general — it appears wherever a stated goal and a measured goal diverge, regardless of the people or the profession involved. A firm that personalises the problem, attributing it to particular individuals’ attitudes, is likely to see it recur with different individuals in the same roles, because the roles — not the people occupying them — are where the incentive is built.
The pattern often survives multiple attempts to address it culturally — firm values statements encouraging collaboration, leadership exhortations to “work as one firm” — precisely because these interventions do not touch the compensation arithmetic that is actually driving the behaviour. A practice group asked to collaborate more, while still measured purely on its own individual number, faces the same rational calculation as before; the request has changed, but the underlying incentive has not.
Where this shows up
This perspective is written with professional services firms in mind. It sits within the Professional and Knowledge-Based grouping, and the questions it raises recur in other sectors too.
Questions worth considering
- What does the firm’s compensation structure make irrational, even when everyone involved has good intentions?
- Is a lack of cross-referral a relationship problem, a measurement problem, or both?
- As the firm adds practice groups, how many new places has it created where individual and collective incentives quietly diverge?
- If a referral fee or shared-credit mechanism existed between these two groups, would the pattern of client service change even without anyone’s attitude changing?
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.