Anchoring Fluff
Convert vague generics and hypotheticals into concrete past behavior
Anchoring Fluff is a technique for converting three types of useless conversational data into actionable insights: generic claims ('I usually,' 'I always,' 'I never'), future-tense promises ('I would,' 'I will'), and hypothetical maybes ('I might,' 'I could'). Each of these is a form of fluff that sounds informative but carries no reliable signal about actual behavior.
The technique works by pivoting from the vague statement to a specific past event. When someone says they 'always' do something, you ask when they last did it. When they say they 'would definitely buy that,' you ask what they have already tried to solve the problem. The goal is to get from their idealized self-image to their actual behavior, because people describe themselves as who they want to be, not who they actually are.
The most dangerous form of fluff is 'I would definitely buy that,' because it sounds like a concrete commitment. In reality, people are wildly optimistic about their future behavior. One startup lost approximately ten million dollars by mistaking fluffy future promises for real commitment. The antidote is always to anchor back to what they have actually done, not what they claim they would do.
- People describe themselves as who they want to be, not who they actually are
- The world's most deadly fluff is: I would definitely buy that
- Generic claims need to be tested against specific past behavior
- If someone hasn't looked for ways of solving a problem already, they are not going to buy yours
- Fluff-inducing questions are not toxic to ask, but the responses are useless as data
- Detect the fluffListen for generic claims, future-tense promises, and hypothetical statements. Key trigger words include 'usually,' 'always,' 'never,' 'would,' 'will,' 'might,' and 'could.' Recognize that these responses, while they feel informative, carry no reliable signal.
- Anchor to a specific past eventAsk when the situation last happened. For example, if someone says 'I always struggle with X,' respond with 'when is the last time that happened?' This transitions from an abstract claim to a concrete data point.
- Walk through the full storyOnce they have identified a specific instance, ask them to talk you through exactly what happened. How did they try to solve it? What tools did they use? How long did it take? What did it cost? This is where the real insights emerge.
- Test for edge casesIf they maintain their generic claim, probe for situations where it did not hold true. Ask about times things fell apart or exceptions to the rule. The edge cases often reveal the most important insights about real behavior versus ideal behavior.
Someone claims to be an Inbox Zero zealot whose inbox is always under control. Rather than accepting this, the interviewer asks what their inbox looks like right now (ten emails) and when it last fell apart (three weeks ago while traveling, took ten days to recover). The generic claim of perfect inbox management hid a significant pain point around email recovery after disruptions.
Fitzpatrick identified fluff as one of three types of bad data (alongside compliments and ideas) that systematically mislead founders during customer conversations. He witnessed a startup lose ten million dollars from trusting fluffy future promises, and developed the anchoring technique as a practical countermeasure that any founder can apply immediately.