FURNITURE CUSTOMER SEGMENTATION WORKSHEET
Published by Furn | Revised October 21, 2026
Guide: https://furncmo.com/blog/furniture-customer-segmentation

Use fictional identifiers here. Keep personal profiles and private service details
in your approved platform. This worksheet does not create audiences or send messages.

1. PURPOSE
Campaign and owner:
Buyer question:
Useful difference in message, offer or next step:
Approved product facts and actual offer terms:
Channel, market and expected frequency:

2. ELIGIBILITY (required before interest rules)
Permission source and scope for this marketing purpose/channel:
Subscription status checked at:
Applicable suppression and deliverability checks:
Market/offer eligibility:
Unknown or conflicting status handling:
Final eligibility recheck before send:
Operational/service messages handled separately by:

3. SEGMENT RULE
Required signals (AND), with source and date:
Alternative signals (OR), explicitly grouped within required eligibility:
Entry criteria:
Exclusions and business-policy pauses:
Exit conditions (changed preference, return, unsubscribe, other):
Signal-review date and expiry:
Overlapping segment precedence:
Deduplication and frequency coordination:
Unknown preference fallback (only within permitted purpose):
Rule owner and review date:

4. FICTIONAL DINING-LAUNCH QA EXAMPLE
Require campaign eligibility AND current declared dining preference AND no
promotional exclusions. Among those profiles, verified trade receives the trade
version; everyone else receives the consumer version. One version per profile.
The open-service-case pause and trade precedence are illustrative business choices.

A: Eligible, dining preference, no exclusions -> Dining launch.
B: Unsubscribed from this marketing audience, dining preference -> Exclude.
C: Eligible, dining preference, verified trade, no exclusions -> Trade version only.
D: Eligible, dining preference, open service case -> Pause promotion.
E: General-newsletter eligible, dining preference unknown -> No dining launch.
   A separate general newsletter is possible only if its purpose fits the subscription.
F: Marketing permission unknown, dining preference -> Hold for verification.

Add cases for your real rules using fictional identifiers:
Input / expected route / actual route / reason:
Eligibility fails despite trade or interest:
Preference changes after initial inclusion:
Relevant purchase returned:
Suppression arrives before send:
Multiple matching segments:
All relevant signals unknown:

5. PREVIEW AND APPROVE
Included count and comparison to source:
Representative included AND excluded records reviewed in approved system:
Unexpected count changes investigated:
Integration freshness and field mapping checked:
Exit and preference-change rehearsal:
Message, destination and product facts reviewed:
Reviewer and approval date:

6. AI DRAFTING BOUNDARIES
Use anonymized rule descriptions and fictional records.
Do not infer wealth, household status or trade eligibility from names/appearance/domain.
Do not put private profiles or service details into general drafting prompts.
Human checks for invented facts, missed exclusions, unsupported offers and false labels:

7. MEASUREMENT
Eligible profiles / delivered messages:
Qualified action definition, count and denominator:
Orders and net revenue; tracking coverage and refund treatment:
Complaints and unsubscribes with explicit denominators:
Attribution rules and observation window:
Comparable test within the same eligible group, allocation and decision rule:
Counts and uncertainty (open rate alone is not proof of intent):
Decision and next review: