Selection criteria, size discipline, change management — and the second peer group you have whether you chose it or not.
Almost every downstream number — positioning, target-setting, the pay-for-performance test, the proxy adviser's assessment — inherits from the peer group. Get it wrong and everything built on top is precisely calculated and pointed in the wrong direction.
It is also the decision most likely to be made for the wrong reasons. Peer groups drift upward, because the companies a management team would like to be compared with are usually larger than the ones it should be compared with. Aspirational peers are not a strategy; they are a mechanism for ratcheting.
The question is where you actually compete for this executive — which is often not your industry classification. A CFO is recruited from a broader pool than a Chief Scientific Officer. Where the talent market and the product market diverge, the talent market usually wins for pay purposes.
Revenue, market capitalisation, assets, headcount — whichever measures fit the business, applied as a band around the company rather than a floor beneath it. A peer group whose median revenue is materially above yours will produce pay levels you cannot justify, and proxy advisers screen for exactly that skew.
Too few and a single outlier moves the median. Too many and the group stops describing anything. Most committees land somewhere in the mid-teens to low twenties, and the right number is the one where adding another company would require relaxing a criterion you believe in.
This is the single most useful discipline available. Agree the selection criteria, apply them, and then look at the resulting list — rather than assembling a list and reverse-engineering criteria that produce it. The first approach survives scrutiny. The second is visible to anyone who reads the disclosure carefully.
Year-on-year churn is a flag. Each addition and removal should have a stated reason — acquired, no longer comparable in size, business model diverged — and the disclosure should carry it. A peer group that changes materially in a year when pay outcomes were unfavourable invites the obvious inference.
Proxy advisers construct their own comparison group for the pay-for-performance screen. It is built on their methodology, not the committee's, and where the two diverge materially a company can pass its own test and fail theirs.
The practical response is to know both. Run the analysis against the committee's peer group for decision-making, and against the adviser's likely group for say-on-pay risk assessment. If the second produces an uncomfortable answer, that is information worth having six months before the proxy rather than six weeks after it.
For nonprofits the equivalent question is comparability under IRS §4958, where the standard is appropriate data as to comparability — sector, budget size, geography and scope — rather than a market-cap band.
There is no rule, but stability and relevance pull in opposite directions. Too few and a single outlier moves the median; too many and the group stops describing your market. Most committees settle in the mid-teens to low twenties. A useful test: the right size is the point at which adding another company would require relaxing a selection criterion you actually believe in.
Yes, and it is the single most valuable discipline in the exercise. Agreeing criteria first and applying them produces a group that survives scrutiny. Assembling a list and then reverse-engineering criteria that justify it is visible to any careful reader of the disclosure, and to proxy advisers in particular.
They construct their own comparison group using their own published methodology, primarily for the quantitative pay-for-performance screen. Where their group and the committee's diverge materially, a company can pass its own analysis and fail theirs. Running both analyses, well before the proxy is drafted, is how that surprise gets avoided.
Annually for accuracy — companies get acquired, change size, or shift business model — but with a strong bias against discretionary change. Every addition and removal should carry a stated reason in the disclosure. Material churn in a year of unfavourable pay outcomes invites an obvious and unhelpful inference.
Peer group construction and review, market positioning, and pay-for-performance analysis against both your group and theirs.
Tell us what you’re working through and we’ll follow up directly.