Audience

An audience is who a piece of evidence comes from. Glare uses four types: the Project Team, Stakeholders, Customers and Participants. Each one is narrowed to a segment and a few attributes that matter.

For example, the SecondBank analytics export reflects Customers: new cardholders in their first 30 days who haven't redeemed points. The concept test reflects Participants: 10 people recruited to match that profile. Knowing which is which tells leaders how much weight each piece of evidence deserves.

Why it matters

When audience isn't tracked, every opinion counts the same. A comment from an executive in a hallway ends up carrying as much weight as the behavior of thousands of customers, and decisions drift toward what insiders believe.

Naming the audience makes the weight of evidence visible. Leaders can see whether a decision rests on what customers actually do, what test participants showed, or what the team thinks, and ask for stronger evidence where it's thin.

Say who the evidence comes from, narrow it to a segment with a few clear attributes, and weigh it accordingly.

How we write one

Every audience is written as one line that names the type, the segment, and the attributes that define it:

An audience type**:** a segment with attributes.

Customers: new cardholders with fewer than 30 days on the card and no points redeemed.

Naming the type first shows at a glance how much weight the evidence carries, and the attributes keep the group narrow enough to design for.

How it differs from similar ideas

  • A user need is what people need. An audience is who has that need, or who the evidence about it comes from.

  • A situation describes what an audience is doing. The audience is the group itself.

  • An input is the material evidence arrives in. An audience is the people whose voices or behavior it reflects.

  • A stakeholder's opinion is evidence from one audience type. It doesn't speak for customers, even when it's about them.

A good audience

  • Names one of the four types

  • Narrows to a specific segment

  • Uses three to five attributes that matter

  • Is small enough to design for

  • Is linked to the evidence it produced

  • Has its signal weight clear

What happens to it

  1. Identified. Ray finds it in an input, or the team names it in a brief or ask.

  2. Defined. The segment and attributes are written down and agreed.

  3. Active. It's the group the team is studying and designing for, and evidence from it is linked.

  4. Refined. New evidence shows the segment should be split or narrowed, and a new audience is linked.

  5. Retired. The team stops designing for it, and it's kept for history.


For Ray

This part tells Ray how to identify the audience behind evidence in an input, such as a survey file, a test transcript, an analytics export, a stakeholder interview or a sales call summary, and how to store it.

How to find one

  1. Identify whose voice or behavior it is. For every piece of evidence, ask who produced it.

  2. Assign one of the four types. Project Team (the people doing the work), Stakeholders (internal leaders and partners with a say), Customers (people using the product in real life) or Participants (people recruited for a test or study).

  3. Find the segment. Name the narrowest group the input supports, such as new cardholders or habitual users.

  4. Pick the attributes. Choose three to five that matter: behavioral, motivational, contextual, lifecycle or demographic. Leave out attributes that don't change the design.

  5. Check for proxies. If someone speaks for customers, such as a sales rep or support lead, the audience is the speaker's type, not Customers.

  6. Note signal weight. Record whether the evidence is observed behavior or stated opinion.

  7. Write the line. Use the format: [Audience type]: [segment] with [attributes].

  8. Fill the record, score it, link it and store it using the rules below.

Cues to look for

  • Recruitment criteria and screeners in test plans

  • Filters and segments in analytics exports, such as "first 30 days" or "no redemptions"

  • Role and title of the speaker in interviews and meetings

  • Group language: "new users", "power users", "our customers", "leadership"

What doesn't count

<table xmlns="http://www.w3.org/1999/xhtml" style="min-width: 50px;"><colgroup><col style="min-width: 25px;"><col style="min-width: 25px;"></colgroup><tbody><tr><th colspan="1" rowspan="1"><p>If the statement...</p></th><th colspan="1" rowspan="1"><p>Store it as</p></th></tr><tr><td colspan="1" rowspan="1"><p>Describes what a group is doing</p></td><td colspan="1" rowspan="1"><p>Situation, linked to this audience</p></td></tr><tr><td colspan="1" rowspan="1"><p>Describes what a group needs</p></td><td colspan="1" rowspan="1"><p>User Need, linked to this audience</p></td></tr><tr><td colspan="1" rowspan="1"><p>Is a sales or support lead speaking for customers</p></td><td colspan="1" rowspan="1"><p>Stakeholders, with a note that it's a proxy</p></td></tr><tr><td colspan="1" rowspan="1"><p>Is "everyone" or "all users"</p></td><td colspan="1" rowspan="1"><p>Nothing yet; ask the team to name a segment</p></td></tr><tr><td colspan="1" rowspan="1"><p>Is a single named person</p></td><td colspan="1" rowspan="1"><p>An attribute of the input, not an audience</p></td></tr></tbody></table>

The record

<table xmlns="http://www.w3.org/1999/xhtml" style="min-width: 75px;"><colgroup><col style="min-width: 25px;"><col style="min-width: 25px;"><col style="min-width: 25px;"></colgroup><tbody><tr><th colspan="1" rowspan="1"><p>Field</p></th><th colspan="1" rowspan="1"><p>What to fill in</p></th><th colspan="1" rowspan="1"><p>Values</p></th></tr><tr><td colspan="1" rowspan="1"><p>Statement</p></td><td colspan="1" rowspan="1"><p>The one-line audience</p></td><td colspan="1" rowspan="1"><p>[Audience type]: [segment] with [attributes].</p></td></tr><tr><td colspan="1" rowspan="1"><p>Type</p></td><td colspan="1" rowspan="1"><p>Which of the four types</p></td><td colspan="1" rowspan="1"><p>Project Team, Stakeholders, Customers, Participants</p></td></tr><tr><td colspan="1" rowspan="1"><p>Segment</p></td><td colspan="1" rowspan="1"><p>The specific group</p></td><td colspan="1" rowspan="1"><p>Short text, such as a lifecycle stage</p></td></tr><tr><td colspan="1" rowspan="1"><p>Attributes</p></td><td colspan="1" rowspan="1"><p>Three to five that matter</p></td><td colspan="1" rowspan="1"><p>Behavioral, Motivational, Contextual, Lifecycle, Demographic</p></td></tr><tr><td colspan="1" rowspan="1"><p>Signal weight</p></td><td colspan="1" rowspan="1"><p>How much the evidence counts</p></td><td colspan="1" rowspan="1"><p>Observed behavior, Stated opinion, Proxy</p></td></tr><tr><td colspan="1" rowspan="1"><p>Size</p></td><td colspan="1" rowspan="1"><p>How many people the evidence covers</p></td><td colspan="1" rowspan="1"><p>Number, or blank</p></td></tr><tr><td colspan="1" rowspan="1"><p>Inputs</p></td><td colspan="1" rowspan="1"><p>Inputs that reflect this audience</p></td><td colspan="1" rowspan="1"><p>Links to Input</p></td></tr><tr><td colspan="1" rowspan="1"><p>Linked nouns</p></td><td colspan="1" rowspan="1"><p>Situations, needs and tests that involve it</p></td><td colspan="1" rowspan="1"><p>Links to Situation, User Need, Test</p></td></tr><tr><td colspan="1" rowspan="1"><p>Status</p></td><td colspan="1" rowspan="1"><p>Where it stands</p></td><td colspan="1" rowspan="1"><p>Identified, Defined, Active, Refined, Retired</p></td></tr><tr><td colspan="1" rowspan="1"><p>Confidence</p></td><td colspan="1" rowspan="1"><p>How clearly the group is defined</p></td><td colspan="1" rowspan="1"><p>High, medium, low</p></td></tr></tbody></table>

Customers and Participants showing behavior carry the most weight. Project Team and Stakeholder opinions are recorded and counted, but never as customer evidence.

Scoring confidence

Confidence measures how clearly the audience is defined in the source, not how important the group is.

<table xmlns="http://www.w3.org/1999/xhtml" style="min-width: 50px;"><colgroup><col style="min-width: 25px;"><col style="min-width: 25px;"></colgroup><tbody><tr><th colspan="1" rowspan="1"><p>Level</p></th><th colspan="1" rowspan="1"><p>When to use it</p></th></tr><tr><td colspan="1" rowspan="1"><p>High</p></td><td colspan="1" rowspan="1"><p>Type, segment and attributes are stated, such as in a screener or an analytics filter</p></td></tr><tr><td colspan="1" rowspan="1"><p>Medium</p></td><td colspan="1" rowspan="1"><p>Type and segment are clear, but attributes are missing or vague</p></td></tr><tr><td colspan="1" rowspan="1"><p>Low</p></td><td colspan="1" rowspan="1"><p>Only the type can be inferred, or the group is described as "users" or "everyone"</p></td></tr></tbody></table>

If any part of the record can't be filled from the source, leave it blank and lower confidence by one level.

Judging the signal weight

Signal weight measures how much evidence from this audience should count toward a decision.

<table xmlns="http://www.w3.org/1999/xhtml" style="min-width: 50px;"><colgroup><col style="min-width: 25px;"><col style="min-width: 25px;"></colgroup><tbody><tr><th colspan="1" rowspan="1"><p>Level</p></th><th colspan="1" rowspan="1"><p>When to use it</p></th></tr><tr><td colspan="1" rowspan="1"><p>Strong</p></td><td colspan="1" rowspan="1"><p>Observed behavior from Customers, or from Participants in a structured test</p></td></tr><tr><td colspan="1" rowspan="1"><p>Moderate</p></td><td colspan="1" rowspan="1"><p>Stated opinion from Customers or Participants, such as survey answers or interview quotes</p></td></tr><tr><td colspan="1" rowspan="1"><p>Weak</p></td><td colspan="1" rowspan="1"><p>Opinion from the Project Team or Stakeholders, or a proxy speaking for customers</p></td></tr></tbody></table>

Weak signals still count. They help form hunches and asks, but a decision supported only by weak signals is flagged.

Linking

  • Check for duplicates first. If an existing audience has the same type, segment and attributes, add the new input to it instead of creating another.

  • Link to every input that reflects this audience.

  • Link to the situations and user needs it's part of.

  • Link to the tests it was recruited for, if it's a Participants audience.

  • Link related audiences. Link Participants to the Customers segment they were recruited to represent, and link refined audiences to the one they replaced.

When to store and when to ask

  • Store automatically when confidence is high or medium.

  • Add to the review digest when confidence is low, when the audience is "everyone", or when a proxy is being treated as customer evidence.

When to flag

  • A decision is supported only by Project Team or Stakeholder voices

  • A Participants audience doesn't match the Customers segment it represents

  • An audience is defined as "everyone" or "all users"

  • A situation or user need has no audience linked

  • A proxy is counted as Customers

Worked example

Input: a concept test plan from the research lead.

"Recruit 10 participants. Criteria: SecondBank cardholder for under 30 days, has earned points but not redeemed any, uses the mobile app at least weekly. These should mirror the new cardholders in the October analytics."

What Ray finds:

  • "Recruit 10 participants" makes the type Participants, size 10.

  • "Cardholder for under 30 days" is the segment: new cardholders.

  • "Earned points but not redeemed" and "uses the mobile app weekly" are the attributes.

  • "Mirror the new cardholders in the October analytics" links this audience to the matching Customers audience.

  • The test will capture behavior, so signal weight is Observed behavior.

Stored record:

<table xmlns="http://www.w3.org/1999/xhtml" style="min-width: 50px;"><colgroup><col style="min-width: 25px;"><col style="min-width: 25px;"></colgroup><tbody><tr><th colspan="1" rowspan="1"><p>Field</p></th><th colspan="1" rowspan="1"><p>Value</p></th></tr><tr><td colspan="1" rowspan="1"><p>Statement</p></td><td colspan="1" rowspan="1"><p>Participants: new cardholders with fewer than 30 days on the card, points earned but none redeemed, and weekly app use.</p></td></tr><tr><td colspan="1" rowspan="1"><p>Type</p></td><td colspan="1" rowspan="1"><p>Participants</p></td></tr><tr><td colspan="1" rowspan="1"><p>Segment</p></td><td colspan="1" rowspan="1"><p>New cardholders</p></td></tr><tr><td colspan="1" rowspan="1"><p>Attributes</p></td><td colspan="1" rowspan="1"><p>Under 30 days (Lifecycle); points earned, none redeemed (Behavioral); weekly mobile app use (Contextual)</p></td></tr><tr><td colspan="1" rowspan="1"><p>Signal weight</p></td><td colspan="1" rowspan="1"><p>Observed behavior</p></td></tr><tr><td colspan="1" rowspan="1"><p>Size</p></td><td colspan="1" rowspan="1"><p>10</p></td></tr><tr><td colspan="1" rowspan="1"><p>Inputs</p></td><td colspan="1" rowspan="1"><p>Concept test plan, October 2026</p></td></tr><tr><td colspan="1" rowspan="1"><p>Linked nouns</p></td><td colspan="1" rowspan="1"><p>Test: abstract-points hunch; represents Customers: new cardholders, first 30 days</p></td></tr><tr><td colspan="1" rowspan="1"><p>Status</p></td><td colspan="1" rowspan="1"><p>Defined</p></td></tr><tr><td colspan="1" rowspan="1"><p>Confidence</p></td><td colspan="1" rowspan="1"><p>High: type, segment and attributes stated in the screener</p></td></tr></tbody></table>

Next step: stored automatically and linked to the matching Customers audience and the concept test.

Edge cases

  • Proxies. A support lead saying "customers hate the rewards tab" is Stakeholders evidence. Store it that way and note it as a proxy.

  • One person, two roles. An employee who is also a cardholder counts as Project Team or Stakeholders when speaking at work.

  • Too broad. "All cardholders" is too wide to design for. Store it at low confidence and ask which segment matters most.

  • Mismatched recruits. If participants don't match the customer segment they stand for, flag the test before its findings are used.

Identify where decision quality breaks down

The Glare Design Assessment helps teams spot weak validation, stakeholder friction, alignment gaps, and assumptions that scale without measurable learning—so you have a clearer starting point for improvement.

About 5 minutes · Team-based · Diagnostic snapshot you can act on

Take the Design Assessment
Audience | Glare