Whoever runs the engagement polls, town halls, and suggestion programs sets rules that filter out serious dissent, so senior leaders see only positive or manageable views and take them as validation.
An example
Quarterly engagement results at a hotel group showed scores up four points and a top theme of pride in service. Its regional vice president of operations for the West had different numbers: housekeeping attrition at 61 percent, and a guest-review site full of complaints about room readiness. She asked HR for the open-text comments. They were confidential, she was told, and had been summarized; the summary grouped them under 'areas for growth' and called the most frequent one 'workload perceptions'. Town hall questions had been submitted in advance and screened by the HR lead whose staffing model the comments criticized. At the next executive meeting the chief executive cited the scores as validation of that model. The western region lost two general managers that summer.
Why it works
The curator rarely lies about the data; they own the rules by which data becomes the data. Forced-choice questions, screened town halls, and executive summaries make filtering cheap, and the polished result looks like good news, so leaders accept it as validation. Raw dissent gets labeled venting or a vocal minority, which strips its value, and anyone who carries it upward around the channel looks disloyal and risks retaliation. Capture holds when the channel is the only formal route, senior leaders have no direct contact with the front line, and the curator is rewarded for scores.
The thinking behind it
Erving Goffman Erving Goffman's The Presentation of Self in Everyday Life (1956) describes performances staged in a front region before an audience, with contradicting facts kept to a back region the audience does not see. A captured feedback channel is a front region: the polished scores and themes are the show, and the raw comments stay backstage. Asking for raw data and outside comparisons is a way of looking into the back region.
George A. Akerlof George Akerlof's The Market for "Lemons": Quality Uncertainty and the Market Mechanism (1970) showed how a party that knows what the other cannot check sets the terms of the exchange. The curator of a feedback program sees every raw comment, while senior leaders see only the summary. Parallel channels, third-party administration, and raw results sent to the board close that gap, which is why they defeat the capture.
Warning signs
Engagement scores rise while attrition, absenteeism, and outside review sites move the other way. Town hall questions are submitted and screened in advance. Open-text comments are summarized by the people whose areas they criticize, or withheld as confidential. Dissent is described as venting or a vocal minority. The same group designs the questions, runs the channel, and reports the results.
How to defend yourself
Compare the channel's story with numbers it does not control.
- Put attrition, absenteeism, exit feedback, and outside reviews beside the engagement scores, and send the comparison to the decision-maker.
- Ask for the raw comments, anonymized, and for the method: who screened what, by which criteria.
- Build your own routes to the people affected: skip-levels, site visits, ordinary conversations.
- If you carry raw dissent upward, present it as data with a source, and copy the curator so it cannot be framed as going around them.
When the gap is material, take it to internal audit or the audit committee. Either can demand the raw data with authority you lack.
How a leader shuts it down
Never rely on a single pipeline to a population, since its curator then edits your view of reality whatever their integrity. Use third-party administration with raw results going straight to the board, keep skip-level access open, and run an anonymous external line to the audit committee. Pay curators for accuracy, and stop paying for high scores.