Tuesday, the usual bar in Southeast, floor sticky in the way that stopped being a complaint years ago and became a feature. I sang Total Eclipse of the Heart with enormous commitment and no accuracy whatsoever. The host, who calls everyone baby, told me I had really gone for it, which is the most honest review I have ever received.
Then I watched him work the book for an hour and forgot to drink my beer.
A guy put in a slow ballad at 9:40. The host moved him back forty minutes and put a bachelorette party's Shania in front of him. Later he pulled a wedding-anniversary couple up early, before the room got loud enough to bury them. He never announced a single one of these decisions and nobody in the bar noticed any of them happen. The night just worked.
He outperforms most segmentation research for one boring reason: he is reading what the room is doing right now, in a loop about four minutes long, and he pays for every wrong call within the hour.
Research asks people to describe themselves. That is the original sin and every good strategist knows it and we all do it anyway, because a described self fits in a deck and an observed one does not. The man with the book never asks anyone what kind of night they want. He watches who is standing, who is holding a coat, who has been waiting long enough to get self-conscious, and he sequences accordingly.
He also segments by state instead of identity. The people in that bar at nine are not a different demographic from the people there at eleven. They are the same people in a different condition, and they want completely different things, and no persona document I have ever seen accounts for that. Your brand's audience is the same. The person browsing on a Sunday morning and the person buying at 11pm on a Wednesday share every attribute in your segmentation and almost nothing that matters.
And he has consequences. If he misreads the room the energy dies in front of him, immediately, in a way he cannot spin. Most research has no such loop. It gets presented, admired, filed, and never once contradicted by a room.
Watch the line outside Salt & Straw sometime. Watch what people actually do when they get to the front, how many of them abandon the flavor they came for after one sample. Then read a stated-preference study about flavor loyalty and tell me which one you would bet money on.
Research asks people to describe themselves.
Got home at half past midnight. Truffles was on the record player, Barnaby was in the sink again, and neither of them asked how the night went. Correctly. They were watching the door.