Clubs collect enormous quantities of supporter opinion and increasingly ignore much of it. The reason is a consistent gap between what people say they want and what their behaviour shows.
Surveys measure intention, not action
A questionnaire asks someone to predict their own behaviour in a hypothetical situation. People answer with the version of themselves they would prefer to be.
Supporters reliably overstate how much they would use statistical content, tactical analysis and archive footage. Usage logs show attention concentrated on results, team news and highlights.
Neither answer is dishonest, because the respondent genuinely values the thing they will not actually open.
Who answers is not who attends
Survey response requires time, willingness and an existing relationship with the club. Those conditions select for the most engaged part of the supporter base.
The occasional attender, the overseas follower and the person who watches only the biggest matches are systematically underrepresented. They are also the groups with the most growth attached to them.
A club that designs around survey respondents is designing around the supporters it already has secured.
Question order shapes the answers
Asking about ticket prices immediately after asking about matchday experience produces different responses than asking in the reverse order. The earlier question primes what the later one is compared against.
Rating scales compound this, since respondents anchor on the first item they see and grade subsequent ones relative to it.
These effects are well understood and still routinely ignored in surveys commissioned under time pressure.
Behavioural data has its own blind spots
Logs record what was available to be done rather than what someone wanted. A feature nobody found is indistinguishable from a feature nobody wanted, so absence of use is ambiguous in a way that a survey answer is not.
Attention metrics also reward whatever is placed most prominently, which makes the data partly a record of past design decisions.
Treating usage as pure preference therefore entrenches existing choices rather than testing them.
The useful approach combines both
Surveys are strong at explaining why something happened and weak at predicting what will. Behavioural data is the reverse, so each covers the other's blind spot almost exactly.
Clubs that use them together tend to test small changes in the product and use survey work to interpret the resulting shifts.
That sequence puts the question after the observation, which is the opposite of how most supporter research is commissioned and most of the reason it disappoints.