A quiet assumption
Comparable-company analysis carries an assumption that rarely gets stated: that the businesses in the set face broadly the same cost structure, the same demand drivers, and the same competitive dynamics. When that holds, comps are an efficient shortcut. When it stops holding, they are a confident-looking way to be wrong.
In several sectors the set has drifted. Two companies with similar revenue and similar margins can now sit on entirely different cost curves depending on how much of their delivery is automated, where their inputs are sourced, and how exposed they are to a single distribution channel.
Go down a level
The correction is not to abandon comps but to stop using them at the company level. Segment economics, contribution margin by customer cohort, cost to serve, retention by acquisition channel, tend to be more stable across a cycle than blended multiples, and they surface the divergence that a top-line comparison hides.
This is more work. It also produces a diligence output that a buyer can act on after close, which a multiple range does not.
A multiple range is not something a buyer can act on the Monday after close. Segment economics are.
Raef Khan, Chief Executive Officer
From valuation to durability
When the peer group stops behaving like a peer group, the interesting question changes. It is less “what is this worth relative to others” and more “which parts of this business will still work in three years, and what would have to be true for them to stop working.”
That is a harder question to answer in six weeks. It is also the one the investment committee is actually asking.
