The latest casualty in the disconnect betweenpartner reward and firm strategy: AI.

Over the course of our four equity partner surveys spanning a decade, a consistent pattern has emerged in professional partnerships: a gap between the strategic horizon firms ask their partners to think on, and the much shorter horizon their compensation systems actually reward. Our 2026 survey with MHPR Advisors, conducted across more than 140 firms, found this disconnect once again and identified a specific, current test case for it: AI.

Managed revenue remains the key measure
Asked which measures carry the most weight in setting a client-facing partner's profit share, respondents were consistent. Managed and supervised revenues ranked first, by some margin. Originations ranked second. Contribution to firm goals, leadership effectiveness and living the firm's values are gaining ground as considerations but remain secondary in most systems still in use today. Team profit featured as well, though notably less prominently than the top two measures.

This ordering is worth pausing on, because it describes what a compensation system is designed to produce, as distinct from what a firm's strategy documents say it values. Most reward systems continue to measure production with precision, while treating the factors that determine a firm's position several years out (talent development, innovation, collaboration, technology adoption) as a lesser, qualitative afterthought.

A rational response from partners
The behavioural consequence of this weighting is neither surprising nor, in our view, a matter of partner attitude. A partner who declines to invest time in a medium-term or long- term strategic goal is not behaving badly; they are reading the reward system accurately. If the measures that move profit share are overwhelmingly short-term and production-based, then time allocated to the long game is, in a narrow financial sense, time misallocated.

This dynamic has relevance for AI adoption. Our survey found that only 21% of respondents believe their compensation system encourages partners to experiment with AI in how their teams deliver services. Experimentation of this kind carries an unavoidable short-term cost: hours spent testing new tools or reworking a delivery model are hours not spent on billable production. Where the reward system weighs production heavily and innovation lightly, a rational partner will defer the experimentation, whatever the firm's stated ambitions for AI. The constraint is structural, not personal.

We know that the reward system doesn’t alone predict behaviour. We are aware of plenty of examples where partners simply do the right things, in the strategic interests of the firm, because they are the right things to do. But to transform at the pace AI now demands, firms need whole cohorts of partners moving together. The most optimistic thing we can say about the disconnect between reward and strategy is that it leaves a partnership unaligned.

Why the gap persists
Most firm leaders are aware of the mismatch. In fact, since Michael Roch and I co-wrote The Partner Remuneration Handbook, the desire to address “non-financial” metrics has probably been the most popular talking point arising from the book.

Yet awareness of the problem has not translated into redesign, for reasons that are themselves instructive. Financial measures are precise, easily defended in a remuneration committee, and difficult to contest. Measures such as firm-goal contribution, leadership effectiveness or innovation are harder to define, harder to evidence, and more easily set aside when a reward round comes down to a final number.

The result is inertia. Production dominates because it is measurable; measurability drives how reward decisions get made; and reward, in turn, drives behaviour. Firms are often left running a compensation model built for a previous strategic period, one that now sits quietly at odds with the strategy and change they are asking partners to pursue.

Six ideas for change

1. Rebalancing the Contribution Framework
One direct lever available to firms is to ensure the partner contribution framework properly addresses strategic issues. This can be done with a balanced scorecard, aligned to the strategy with supporting metrics (financial and non-financial) aligned to each pillar. Some firms weight contribution areas while others promote holistic assessments. The risk sits at the extremes: where production is weighted at 80% or more and every other area receives only a token allocation, a firm is, in substance, running a financial meritocracy, regardless of how the framework is described internally.

2. Representing long-term investment in the reward formula
‍A second, related adjustment concerns how long-term and short-term measures are combined. The Handbook observes that when longer-term initiatives are tracked on a partner's scorecard but given no distinct weighting in the compensation calculation, the practical effect is that short-term, easily quantified results determine the outcome regardless. If AI experimentation is to compete meaningfully with the pull of billable production, it likely needs its own explicit and protected allocation within the reward formula, rather than an implicit mention folded into a broader category such as leadership.
‍
This weighting can be delivered in different ways: as a distinct, separately scored line item, or as a qualitative modifier applied holistically to the overall outcome. Either approach can work, provided the weighting is specific and visible enough that a partner can see it counted.

3. Using objectives, properly
We know a lot of firms have tried and failed to use objectives as part of their partner contribution and reward process. It could be time to try again, this time with tighter discipline: fewer objectives, agreed collaboratively between partner and partner-leader, and reviewed with the same rigour applied to financial metrics. Partners who set good objectives related to strategy and transformation can be rewarded for their achievement. These objectives will choose relevant metrics (rather than the one-size-fits-all approach to metrics most firms take).
‍
In the example of AI experimentation, rather than assessing AI-related work solely against an immediate revenue outcome, firms can agree specific medium-term objectives for AI adoption and evaluate progress using both a rating and the narrative context behind it. This matters because experimentation, by its nature, includes a reasonable prospect of setbacks; partners are unlikely to take on that risk if any shortfall produces an immediate reduction in compensation. Incorporating qualitative context into the assessment and giving credit for considered experimentation even where outcomes are mixed allows a reward system to recognise the process of innovation.
‍
4. Linking reward to enterprise value
A parallel lever sits outside the annual profit share altogether. Only 17% of firms in our 2026 survey link partner equity to goodwill or enterprise value; most partnerships still rely on year-to-year profit shares alone. A profit share, however carefully weighted, is structurally a one-year instrument, and cannot easily reward a partner for work whose payoff is a firm valuation uplift several years out. Bonus pools tied to multi-year value creation, or equity classes that vest against longer horizons, give partners a mechanism to capture upside from exactly the kind of investment, AI transformation among them, that an annual formula struggles to price.

5. Strengthening RemCom calibration and governance
A well-weighted framework can still be undone at the point of decision if a remuneration committee reverts, under time pressure, to the measure it finds easiest to defend: production. Firms that build explicit moderation into the process, business unit leaders aligning in advance on what good AI-adoption contribution looks like for a given tier, and RemCom members calibrating across partners rather than scoring each in isolation, protect qualitative measures from being quietly discounted at the final reward round. This is a governance fix as much as a design one, and it is often the missing step even in firms that have already rebalanced their framework on paper.
‍
6. Protecting time for strategic investment
Some firms address the short-term cost of experimentation not by changing what is measured, but by removing the trade-off at its source. A fixed number of hours, or a notional budget, is ring-fenced per partner for strategic investment, AI experimentation, business development in a new area, mentoring, and is simply not counted against utilisation or billable targets. This does not require touching the compensation formula; it changes the denominator a partner is implicitly optimising against. Some firms pair this with visible recognition of partners who use the time well, rather than a formal reward linkage, sidestepping the measurement difficulty of qualitative factors altogether.

These adjustments do not require firms to relax financial discipline; production and client relationships remain central to any partnership's economics. What they require is a clear- eyed look at what the current weighting actually rewards, and a willingness to recalibrate it so that the behaviour firms describe as strategically important is also the behaviour their compensation system is designed to support.

‍