Back to all articlesAudience segmentation

Turn an audience brief into explicit eligibility rules

Alactic Editorial10 min read

Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.

“Contact suitable customers for replenishment” sounds specific until staff start selecting records. One person chooses anyone who bought before. Another looks for an old purchase date. A third includes people who clicked a product link. Each has interpreted suitability differently.

To define audience eligibility rules, turn the request into a test with separate parts: mandatory conditions joined by AND, acceptable alternatives grouped with OR, evidence requirements for each condition, and exclusions that override an otherwise positive result. Keep optional signals outside the membership decision.

The result should let two people inspect the same evidence and reach the same answer. The fictional worked example below builds that test for a replenishment audience, including what to do when the evidence is missing or contradictory.

1. Define the membership decision before choosing fields

Start with the question the audience must answer, not with the fields that happen to be easy to filter. In this hypothetical example, the decision is:

Which existing customers have a confirmed purchase, an identified product compatible with the refill, and evidence that replenishment may be due, with no applicable exclusion?

This wording separates three different claims: the person bought something, the refill fits their product, and there is a reason to consider replenishment now. None proves the others.

A prior purchase does not establish compatibility. Compatibility does not establish a replenishment need. An expression of interest does not establish a prior purchase. If those distinctions disappear into a label such as “good prospect,” staff must supply their own interpretation.

Choose a unit of membership as well. For this example, evaluate a customer–product pair, then include the customer if at least one pair passes. This prevents a purchase date for one product from being combined with compatibility information for another.

Keep the output narrow: eligible, ineligible, or held for missing evidence. Audience membership is not authorisation to send a message, and it is not a prediction that the customer will buy.

2. Separate requirements, alternatives, signals, and exclusions

Before writing the final expression, classify each part of the brief. This exposes conditions that otherwise get mixed together.

  • Required conditions: every qualifying record must satisfy them. Here, prior purchase and product compatibility are required.
  • Alternative evidence: any accepted route can satisfy a particular requirement. Here, either a stated replenishment need or an elapsed usage interval can establish timing.
  • Optional signals: useful for ordering an eligible audience, but not enough to admit someone. A product-page visit might belong here.
  • Exclusions: conditions that remove a record even when its inclusion tests pass.

Write the hypothetical rule with visible parentheses:

Base inclusion =
  confirmed prior purchase
  AND compatible product
  AND (current stated replenishment need OR elapsed usage interval)

Final audience =
  base inclusion
  AND no applicable exclusion

The parentheses matter. Without them, a reader may interpret an elapsed interval as a standalone route into the audience. That could admit someone whose purchase or product compatibility has never been established.

Optional signals should not rescue a failed requirement. A customer who viewed the refill page repeatedly still fails this audience if their product is incompatible. Conversely, a customer who satisfies the rule should not fail merely because no browsing activity is recorded.

3. Filled hypothetical eligibility worksheet

Fictional worked example: the products, dates, intervals, records, and thresholds below are hypothetical. They illustrate a proposed operating design, not an Alactic feature or a product-use recommendation.

Scope and evaluation point

  • Audience: replenishment candidates for the fictional R refill used with the fictional M device.
  • Evaluation date: 30 June 2026.
  • Evaluation unit: customer–device pair, linked to the purchase and refill evidence for that device.
  • Customer-level result: include a customer once if at least one pair passes and no customer-wide exclusion applies.
  • Rule owner: the designated audience operations owner approves changes to evidence definitions and thresholds.

Required purchase and compatibility evidence

  • Prior purchase required: a completed order linked to the customer and the relevant device or refill. A quote, abandoned basket, or cancelled order does not satisfy this condition.
  • Compatible product required: an identified device model listed as compatible with R in the business’s maintained product reference. A broad category such as “device owner” is insufficient.
  • Unknown model: hold for verification rather than assume compatibility.
  • Conflicting model evidence: hold until the discrepancy is resolved; do not choose whichever value makes the customer qualify.

Timing evidence: either route is sufficient

  • Route A—stated need: a dated customer statement requesting more R refill, recorded within the hypothetical previous 30 days and not superseded by a later fulfilment, cancellation, or statement that no refill is needed.
  • Route B—elapsed interval: at least a hypothetical 60 days since the latest completed R supply purchase for that device, with the standard single-pack quantity confirmed.
  • Interval boundary: a purchase dated 1 May 2026 satisfies the 60-day threshold on 30 June 2026; one dated 2 May does not. Use calendar-date differences consistently for this worksheet.
  • Nonstandard quantity: do not use Route B if the last purchase contained multiple packs or its quantity is unknown. Route A can still qualify the record.
  • Latest-event rule: a newer completed R supply purchase replaces an older purchase as the interval starting point.

Explicit exclusions applied last

  • Customer-wide: an active outreach suppression or opt-out removes the customer from this outreach audience.
  • Pair-specific: the customer reports that they no longer own or use the identified device.
  • Pair-specific: an open replenishment order already covers the identified need.
  • Pair-specific: a later customer statement says replenishment is not currently needed.
  • Unresolved exclusion status: hold the affected record when a required exclusion check cannot be completed.

Applying exclusions last means giving them final precedence in the decision. It does not mean staff must first prepare outreach for people already known to be excluded.

Output and retained explanation

  • Eligible: all mandatory conditions pass, at least one timing route passes, and all required exclusion checks are clear.
  • Ineligible: a mandatory condition fails, both timing routes fail, or an exclusion applies.
  • Held: unresolved evidence prevents a reliable decision.
  • Retain with the result: evaluation date, rule version, qualifying pair, evidence references, timing route, and decisive exclusion or hold reason.

This worksheet deliberately avoids treating the elapsed interval as proof that the customer has run out. It is an agreed membership proxy. The stated-need route is more direct, but depends on a current, product-specific statement.

4. Test contrasting records, not just obvious matches

The following fictional records use the hypothetical worksheet above. Each tests a different part of the logic.

  • Asha—eligible through stated need. A completed purchase and compatible M device are confirmed. Her latest single-pack R purchase was 10 June, but she requested more R on 25 June. No exclusion applies. The interval fails; the stated-need route passes.
  • Ben—eligible through elapsed interval. A completed purchase and compatible device are confirmed. His latest single-pack R purchase was 1 May. No stated need is recorded and no exclusion applies. The interval passes exactly at the boundary.
  • Cleo—ineligible because neither timing route passes. Purchase and compatibility are confirmed. Her latest single-pack purchase was 2 May, and no current need is stated. Being close to the threshold is not the same as meeting it.
  • Dara—held for compatibility evidence. A completed purchase and current request for R exist, but the device model is unknown. A stated need cannot substitute for the required compatibility check.
  • Eli—ineligible because an exclusion overrides inclusion. Purchase, compatibility, and elapsed interval all pass. An active outreach opt-out removes Eli from the outreach audience.
  • Farah—ineligible because prior purchase fails. A compatible device and current refill request are recorded, but there is only an uncompleted quote. This audience specifically requires a prior completed purchase.
  • Gus—ineligible after a newer purchase resets the interval. An older purchase is beyond the threshold, but a completed single-pack purchase on 20 June supersedes it. No current stated need remains.
  • Hana—held because quantity is unknown. Purchase and compatibility are confirmed, and the purchase date is old enough, but its quantity is missing. There is no stated need. The interval route cannot be evaluated as defined.

For customers with multiple devices, test the pair linkage explicitly. A compatible model on one record and an old refill purchase for a different model must not combine into a passing result.

5. Decide what counts as evidence—and what stays unknown

A field name is not an evidence standard. “Customer,” “compatible,” and “needs refill” can each conceal assumptions. Specify which record supports each claim, how it links to the evaluated pair, and whether later evidence supersedes it.

Use three states for individual tests: true, false, and unknown. Unknown means the evidence cannot establish either answer; it should not silently become false or true.

That distinction matters when evaluating OR conditions. If a current stated need is confirmed, an unknown interval does not prevent the timing requirement from passing. If no current need is recorded and the interval cannot be calculated, the timing requirement remains unresolved.

Likewise, a known exclusion can settle the result even when another field is missing. There is little value in resolving compatibility merely to admit a customer who is already excluded from outreach.

Avoid collecting evidence simply because it is available. For this audience, a phone number’s apparent validity cannot establish purchase, compatibility, or replenishment need. Add evidence only when it supports a defined membership test or resolves a specific ambiguity.

6. Choose the trade-off openly and recover without exceptions

A strict evidence policy produces a smaller, more explainable audience, but may hold customers who genuinely need a refill. A loose policy admits more people while increasing the chance that staff contact someone for an unsuitable or already fulfilled need.

The worksheet’s single-pack restriction illustrates that trade-off. A date-only rule is easier to apply, but ignores how much was purchased. Holding nonstandard quantities avoids pretending the same interval fits every order. If too many useful records are held, improve quantity evidence or define a separately reviewed quantity-based rule; do not quietly waive the requirement.

Route stale or contradictory records to a correction queue with a specific missing item, such as “verify model” or “resolve latest purchase.” After correction, reevaluate the full rule, including exclusions. Resolving one gap does not automatically establish eligibility.

Do not fix a disputed result by adding someone manually without an explanation. Determine whether the evidence is wrong, the rule is wrong, or the result is correct but commercially inconvenient. Evidence corrections affect a record; rule changes require reviewing other affected records as well.

Keep membership separate from communication readiness. A qualifying replenishment record still needs the appropriate channel checks before any send. For WhatsApp, recipient permission and opt-outs matter; consult the current WhatsApp Business Messaging Policy for precise requirements. Membership alone is not permission to contact.

7. Validate agreement before using the audience

Run a small, deliberate review before relying on the rule. Include boundary dates, missing quantities, multiple devices, recent purchases, and exclusions—not only clean qualifying records.

  1. Have two reviewers independently classify the same test records using only the worksheet and available evidence.
  2. Compare both the final status and the decisive reason. Matching statuses can conceal different interpretations.
  3. Resolve each disagreement by clarifying the rule or correcting the evidence, then rerun the affected cases.
  4. Retain the tested rule version with the audience result so later changes can be explained.

Measure reviewer agreement, the share of records held, and the reasons for holds or exclusions. Agreement measures consistency, not whether the timing proxy predicts a real need. A high held share can reveal a data gap; it is not automatically a reason to lower the standard.

As a hypothetical next exercise, select 20 records, including deliberately difficult cases, and ask two staff members to classify them independently. Record every disagreement and its cause. The practical completion test is simple: both reviewers can explain each result from the same written conditions, without relying on an unwritten exception.

Audience segmentationEligibility rulesCRM data qualityReplenishment

Put the next conversation to work.

Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.

Explore Communications