Claim
85% of companies create buyer personas, but 77% never use them. The failure mode is not personas as a concept — it is the demographic scaffolding ("Mary Marketer, 28, has a cat") that produces documents nobody can act on. Personas built on Jobs to Be Done segmentation rarely require more than 3-5 personas (vs. 10+ demographic ones) and every line in them maps to a business decision.
Mechanism
Demographics describe who the buyer is; JTBD describes what the buyer is hiring the product to do. Content, product, support, and sales decisions all branch on the latter, not the former. A persona document that lists hobbies cannot tell you what blog post to write; a persona organized by job ("eliminate the manual reconciliation step in month-end close") tells you exactly what content topic, what product feature, what sales script. Barnes's discipline: if a piece of persona information cannot be traced to a specific business action, it does not belong in the persona.
Conditions
Holds when:
- The team has access to actual customers for qualitative interviews (warm-introduced, not cold).
- Marketing/product/sales actually intend to consume the persona, not check a box.
Fails when:
- The "persona" exercise is a brand-hygiene deliverable for an agency pitch — JTBD rigor is wasted there.
- Pure-consumer brand work where demographic and identity signals genuinely drive purchase.
Evidence
"85% of companies create buyer personas, but 77% never use them."
"Using JTBD segmentation, Barnes has never seen a company require more than three to five personas, compared to the ten or more that demographic approaches typically produce."
— Adrienne Barnes, Best Buyer Persona methodology
Signals
- Persona document opens with the job-to-be-done, not a stock photo + age.
- Each section in the persona names which downstream artifact it informs (content topic, sales script, support macro, product brief).
- Persona count per company stays in the 3-5 range across every segment.
Counter-evidence
B2C brand and DTC marketers (Sarah Levinger, behavioral-design school) argue identity-and-emotion signals predict conversion better than functional jobs. JTBD also under-serves industries where the buyer is not the user — channel-led enterprise software needs a buying-committee map that JTBD alone does not provide.
Cross-references
- (none in current corpus)