Benchmarking looks objective because it produces tables, positions and numbers. But the result can be deeply influenced by an earlier decision: who entered the comparison. If that choice is adjusted after the answer appears, the method begins to confirm what someone wanted to demonstrate.
Every comparison begins before collection
Before measuring any company, it is necessary to define the universe in which the reading will make sense. Direct competitors, aspirational references, adjacent companies, digital leaders and geographic cuts answer different questions.
Mixing these categories without criteria produces a list that looks rich but is methodologically weak. Choosing only weaker companies can inflate the observed position. Selecting only global leaders can create an artificially severe distance.
The quality of the benchmark therefore depends on the quality of the question that justified each participant.
Bias appears once we know who won
Once scores and rankings appear, a human temptation emerges: revise the set so the result feels more intuitive, more comfortable or more aligned with the expected narrative.
This is where benchmarking can lose independence. Removing a reference because it “does not seem comparable” or adding another because it “better represents the market” may be justifiable — but if the decision happens after reading the results, it must be treated as a new round, not as an invisible correction.
Without that discipline, the competitive universe becomes a variable that moves until it produces the desired answer.
Freeze before score
That is why Brandile adopts a simple principle: the universe and the evidence set need to be frozen before comparative calculation. Freeze does not turn strategy into pure mathematics; it simply prevents the ruler from being moved after we see the result.
If relevant new evidence emerges or the market definition needs to change, the methodology can evolve. The difference is to record that change and recalculate transparently, preserving the trail that explains why the result changed.
Not every competitor needs to play the same role
A more useful competitive reading recognizes strata. Direct competitors help explain immediate competition. Aspirational references show higher standards. Adjacent companies reveal solutions that can cross categories. Digital leaders help identify expectations users carry from other contexts.
The value is not in pretending they are all equivalent. It is in stating why each group was included and interpreting results according to that role.
The comparison needs to be defensible
A good benchmark is not the one that produces the most impressive ranking. It is the one that makes it possible to answer clearly why those companies were compared, which criteria were applied and when the universe was closed.
When that chain is preserved, comparison stops being report decoration. It becomes a decision basis that can be discussed, reviewed and, above all, repeated in the future without reinventing the ruler.