One Account, Four Personas: Build an Outbound Message Matrix
TL;DR: Keep one verified account thesis, then adapt the job, evidence, risk, message angle, and CTA for each buying role. Do not send four versions of the same email. Coordinate the touches at account level so one person is not contradicted by another message, and use placeholders until the actual role and concern are verified.
How do you personalize outbound for different personas at the same company?
Separate account facts from persona assumptions. The account layer contains verified events, initiatives, systems, and constraints. The persona layer states what a role may need to decide, what proof would help, and which question is appropriate. The message is created only where the two layers intersect.
The four roles below are adaptable buying jobs, not claims about every committee. Replace each bracketed variable with research that is verified for the target account. If one person covers two jobs, combine the jobs and send one coherent message.
| Role or job | Likely concern to validate | Relevant trigger | Proof needed | Message angle | CTA | What to avoid |
|---|---|---|---|---|---|---|
| Economic approver | Whether [INITIATIVE] merits budget and attention | A verified strategic change, public commitment, or operating constraint | Business-case logic, downside, and decision criteria | Frame the decision and tradeoff, not a feature tour | "Worth comparing the decision criteria for [OUTCOME]?" | Invented ROI, false urgency, or assuming final authority |
| Operational owner | Whether the workflow can run reliably with current people and systems | A verified process change, hiring pattern, or execution bottleneck | Workflow map, ownership, controls, and time-to-operate assumptions | Show how the work would change day to day | "Open to mapping where [WORKFLOW] currently breaks?" | Implying the team is failing or bypassing the owner |
| Technical or data evaluator | Whether data, integration, security, and recovery requirements can be met | A verified stack, migration, integration, or governance event | Architecture, data flow, permissions, failure handling, and test plan | Lead with interfaces and risk boundaries | "Should I send a field-level integration checklist?" | Claiming compatibility or security that has not been validated |
| Practitioner or champion | Whether the change removes work and improves execution quality | A verified role need, manual step, or active initiative | Concrete task change, usability, guardrails, and escalation path | Use the person's workflow language and a small next step | "Would a sample [PLAY OR WORKFLOW] be useful to pressure-test?" | Making the person responsible for an executive decision |
Build the account thesis first
- Verified account fact: [SOURCE-CHECKED EVENT OR CURRENT STATE].
- Operating implication: This may affect [WORKFLOW OR DECISION]. Label this as a hypothesis.
- Relevant roles: [ROLE A] owns the decision, [ROLE B] operates it, [ROLE C] validates it, and [ROLE D] uses it. Confirm rather than assume.
- Proof gap: The team still needs evidence about [INTEGRATION, BUSINESS CASE, PROCESS, OR USER WORK].
- Shared question: "How is [ACCOUNT] deciding [DECISION]?" This anchors every persona-specific message.
Translate the thesis into four different openings
Use the same account fact, but ask a question that fits the role. These examples deliberately use variables so they remain templates rather than invented account claims.
- Economic approver: "[VERIFIED EVENT] may change the economics of [WORKFLOW]. How are you deciding whether to invest now or keep the current model?"
- Operational owner: "With [VERIFIED EVENT], teams often revisit who owns [STEP]. Is the current handoff something your group is changing?"
- Technical evaluator: "If [VERIFIED EVENT] changes [SYSTEM OR DATA FLOW], the key questions are usually field ownership and recovery. Is that in your evaluation?"
- Practitioner: "I am trying to understand how [STEP] works today. Is [MANUAL TASK OR DECISION] part of your role, or does another team own it?"
Coordinate touches so the account hears one story
| Moment | Primary role | Purpose | Other-role action | Shared context to record | Stop or change rule |
|---|---|---|---|---|---|
| Initial hypothesis test | Most likely operational owner | Validate the workflow and ownership | Research other roles, no simultaneous duplicate pitch | Account fact, hypothesis, question, owner, and source | Change role map if ownership is incorrect |
| Proof request | Technical evaluator or practitioner | Test feasibility or day-to-day relevance | Tell the account owner before sending | Requested proof, open technical questions, and objections | Stop feature claims that lack verified support |
| Decision framing | Economic approver | Discuss criteria and consequences after relevance is established | Reference validated workflow, not private replies | Confirmed problem, decision criteria, and participants | Do not escalate around an active owner without context |
| Multi-thread follow-up | Role with the clearest next action | Advance one agreed task | Other contacts receive no parallel generic chase | Latest response, next step, and accountable sender | One human owner controls cadence |
| Reply, meeting, or opt-out | Conversation owner | Handle the live motion or compliance state | Suppress conflicting automation across all personas | Outcome, suppression scope, and next step | The most restrictive valid state wins |
Use a message card for every persona
A message card prevents copy from drifting away from the account thesis. Complete it before writing the email:
- Role or buying job: [ROLE].
- Verified account fact and source: [CLAIM ID].
- Hypothesis, explicitly labeled: [POSSIBLE IMPLICATION].
- Role-specific decision: [DECISION THIS PERSON MAY INFLUENCE].
- Proof offered: [WORKFLOW, CHECKLIST, ARCHITECTURE, CASE, OR DEMO].
- One question: [QUESTION THE PERSON CAN ANSWER].
- One CTA: [LOW-FRICTION NEXT STEP].
- What the message must not claim: [UNVERIFIED PAIN, AUTHORITY, ROI, OR TIMING].
Resolve contradictions before sending
If the economic message promises speed while the technical message minimizes implementation risk, both must be true under the same proposed approach. If the practitioner receives a workflow-level CTA while an executive receives a generic demo request, the account hears two different sales motions. Review the set of messages together, not one at a time.
- Use one account owner and one shared activity timeline.
- Make the account thesis visible in every message card.
- Record which role received which proof and question.
- Prevent two automated sequences from owning the same account motion.
- Route every reply to the owner before another persona is contacted.
- Retire a persona hypothesis as soon as a contact corrects it.
Evidence and privacy guardrails
The Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile includes privacy and confabulation among the risks that need controls. For multi-persona research, keep only business-relevant public evidence, mark inference, and avoid private or sensitive details. A role title is not evidence of budget, authority, or personal priorities.
How Unify fits
Unify Agents and B2B Company and Contact Data can support account and contact research, while Multi-channel Sequencing can coordinate role-specific touches. The account thesis, role map, evidence rules, and human ownership still need to be explicit.
Start using Unify to coordinate account research, persona context, and role-specific sequencing in one workflow.
Frequently asked questions
Do all accounts need four personas?
No. Four is a planning model. Use only the buying jobs that are relevant and verified for the specific decision.
Should every persona receive a different value proposition?
They should receive a role-relevant angle on the same account decision, not unrelated value propositions.
Can two people receive outreach on the same day?
Only when the account plan, ownership, and message roles are coordinated. Avoid simultaneous generic outreach that makes the team appear unaware of its own activity.
What if one person covers two roles?
Combine the relevant decisions and proof into one message. Do not send two persona templates to the same person.
How should replies change the matrix?
Treat replies as higher-quality evidence. Update ownership, concerns, proof needs, and next actions before contacting another role.
What should stay constant across all messages?
The verified account facts, core decision, source trail, and current owner should remain consistent.
Sources
- Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile, National Institute of Standards and Technology, accessed September 2026.
- B2B Company and Contact Data, Unify, accessed September 2026.
- Agents, Unify, accessed September 2026.
- Multi-channel Sequencing, Unify, accessed September 2026.
