Audience groups

An audience is rarely one kind of person. stage splits it into a small number of mutually exclusive groups, by the behaviors that most separate them, and describes each group well enough to write to it. The group card follows the standard criteria for useful segments (Wedel and Kamakura, 2000); see research note 4.

A run

1. Frame it, or don’t. At the start of a run you may say what is offered, what people should do, and what to optimize for. Skip it and you get a pure discovery read of the audience.
2. The framing is fixed. It is recorded on the run, so a re-run on the same audience reproduces the same groups.
3. The engine splits. The audience is split into roughly three to eight mutually exclusive groups by what most separates them.
4. You shape the result. Rename a group, park it (parked groups leave the package), re-cut, and print a brief.

What a group card shows

Field	Meaning
Size and share	How many people, and what share of the audience
Lead behavior	The behavior that most sets the group apart, with its multiple against the audience
Sharpness	How distinct the group is: Sharp, Clear, Soft or Mixed
Angle	The message direction the evidence supports
Objection	What the group needs answered before it acts
Evidence	The departures from the audience the group was built on

Rules the engine follows

• A missing measure is left out, never faked. If a measure cannot be read for a group, the card omits it.
• Buying language needs evidence. A card uses in-market language only when a live in-market measure actually separates the groups.
• Demographics are context, not evidence. Age band or income can describe a group, but never count as the reason it fits.
• A tag needs two kinds of difference. A behavior is called out only when it clears both a ratio and an absolute gap against the audience, so a large ratio on a tiny base does not become a headline (Wainer, 2007).
• Your words win. Text an operator types outranks the engine’s own.
• Group-level, not person-level. Explanations describe the group. They are never claims about one member.

Why a fixed framing matters

A group read is only useful if it holds still. Fixing the framing on the run means the brief your creative team wrote last week still describes the same people this week.