Engagement
Programming, collaborations and promotions for a place, and an honest list of the assets worth building.
01 / The problem
Marketing to an average resident who does not exist. The trade area has four kinds of people in it and the campaign speaks to none of them.
02 / The decisions it serves
Not a report. A decision somebody has to sign.
Each of these is a question your team already asks, and currently answers with a spreadsheet, a phone call and a fortnight.
Who is actually here
Not one persona. The mix, and whether it splits into two audiences that need different things.
What to programme, and with whom
Ideas that trace back to something in the data rather than to a brainstorm.
What is worth building
The assets you do not have yet, and whether the audience would use them.
03 / It is willing to say no
It does not hand you numbers. It gives you a verdict.
An assistant that only ever agrees with you is a mirror, not an analyst. This one holds a line, names the thing that kills the deal, and will tell you to walk away from a site you already like.
The thresholds it holds you to
-
It reads the audience as a mix, not a single persona: every segment above 5%, or enough of the top segments to cover 80% of the people who are actually there.
-
If the top two segments sit within five points of each other, it flags a dual-audience market and tells you targeting has to split.
-
Two to four persona snapshots, sequenced across 30, 60 and 90 days.
What it will not let you ignore
-
It will not fabricate a named partner, a competitor programme, or a community fact. Anything it has not verified is labelled unverified, to be confirmed locally.
-
Every idea has to trace back to something in the data. Untraceable ideas get cut rather than padded in.
04 / Why not just ask an AI
A fluent answer and a defensible one are not the same thing.
Every model will answer a location question now, and most of the answers sound right. Ask twice and you get two of them. Ask where the number came from and the room goes quiet.
-
It reads the audience as a mix and flags a dual-audience market, rather than flattening it into one flattering persona.
-
It will not fabricate a named partner, a competitor programme or a community fact. Anything unverified is labelled as such, to be confirmed locally.
-
Every idea traces to something measured. Untraceable ideas get cut rather than padded in to fill the slate.
05 / What it runs on, and what you get
The data underneath
Licensed and authoritative, and named in the output, so the number keeps its source when it travels.
- Esri GeoEnrichment
- Esri Tapestry
- Daytime and workplace population
- Zoneomics
- Regrid
What you walk out with
A slate of ideas you can vote on, across programming, collaborations, promotions, the assets you already have, the assets worth building, and the social content. Then the winners get worked up properly.
The method it used and the vintage of every number travel with it, so it still holds up when somebody asks you why in six months.
How an answer travels06 / It pairs with
Nobody makes one decision in isolation. Most teams run this alongside two or three of these.
Run Engagement on something you already decided. See if it agrees with you.
The honest test of a method is whether it reaches the conclusion your best analyst already reached, and tells you plainly when it does not.
Twenty minutes, and you can argue with the weights.