--- name: sales-personality-selling description: "Personality-based selling strategy — reading a buyer's DISC type (and Big Five/OCEAN) to adapt outreach, discovery, objection-handling, and email tone to how they decide, across tools like Crystal, Humantic AI, Humanlinker, and Happysales. Use when you want to tailor a cold email to a specific prospect's personality, adapt your pitch to a Dominant/Influential/Steady/Conscientious buyer, decide whether personality prediction is accurate enough to trust, choose a personality-intelligence tool, handle a buying committee with mixed communication styles, or stay compliant (consent, legitimate interest) when profiling people from public data. Do NOT use for a specific tool's setup or API (use /sales-crystal) or for generating fictional marketing personas/audience segments (use /sales-idea-validation)." argument-hint: "[what you're trying to personalize — e.g., 'adapt this cold email to a D-type CFO' or 'pick a DISC prediction tool']" license: MIT version: 1.0.0 tags: [sales, personality, disc, personalization, strategy] --- # Personality-Based Selling Adapt *how* you sell — tone, pace, proof, ask — to *how a specific buyer decides*, using a behavioral model (usually **DISC**, sometimes **Big Five/OCEAN**). This is tool-agnostic method; the platforms (Crystal, Humantic AI, Humanlinker, Happysales) just *predict* the type — the selling adaptation is yours. This skill is about a *real, named individual*, not a fictional marketing persona (that's `/sales-idea-validation`). ## Step 1 — Gather context 1. **What are you adapting?** - a) A cold email / opener to one prospect - b) A live discovery call or pitch - c) Objection handling / negotiation - d) A buying committee with mixed styles - e) Choosing a personality-prediction tool 2. **Do you already have a predicted type**, or do you need a tool to get one? 3. **How confident is the prediction?** From a rich source (their own writing, an assessment) or a thin one (a sparse LinkedIn profile)? This decides how hard you lean on it. Skip-ahead rule: if the prompt already answers these, go to Step 2. ## Step 2 — The DISC lens (the working model) Most tools output **DISC**. Read a buyer on two axes: **pace** (fast/assertive vs measured) and **priority** (task/logic vs people/emotion). That yields four primary types: | Type | Reads as | They want | Sell by | Avoid | |---|---|---|---|---| | **D — Dominance** | Direct, fast, results-driven, impatient | Control, ROI, the bottom line | Leading with the outcome/number; giving them the decision; being brief | Small talk, long build-up, hand-holding | | **I — Influence** | Outgoing, enthusiastic, relationship-first | Recognition, vision, social proof | Energy, stories, big-picture vision, testimonials | Dense data dumps, cold formality | | **S — Steadiness** | Warm, steady, risk-averse, loyal | Safety, support, no surprises | Reassurance, references, a gradual low-risk path, patience | Pressure, sudden change, hard closes | | **C — Conscientiousness** | Analytical, precise, skeptical, reserved | Accuracy, proof, logic | Data, documentation, specifics, letting them verify | Hype, vagueness, rushing, unsupported claims | Most people are a **blend** (e.g. `Di`, `Sc`). Adapt to the dominant letter, soften for the second. **Big Five/OCEAN** is the more research-validated model (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism); some buyers/tools prefer it. It's a spectrum, not four boxes — useful when you want nuance over a quick archetype. ## Step 3 — Apply it to the moment **Cold email** - D: subject = the outcome; 2–3 sentences; one clear ask. I: warm hook + a name/logo they'd know. S: low-pressure, "no rush," offer a small next step. C: a specific claim + a link to proof. **Discovery / pitch** - Match their pace and depth. D wants you to skip to impact; C wants you to slow down and substantiate. Mirror their words. Ask, don't assume — the prediction is a hypothesis to *test* on the call. **Objections / negotiation** - D: give options and control. I: preserve the relationship and their image. S: de-risk and reassure. C: answer with evidence and let them audit it. **Buying committee (mixed styles)** - A deck for a C-type economic buyer and an I-type champion needs both a data appendix and a vision slide. Map each stakeholder's style (see `/sales-account-map`) and tailor the artifact each receives. ## Step 4 — Guidance and guardrails - **Treat every predicted type as a hypothesis, not a fact.** Predictions from public data (especially a thin LinkedIn profile) are low-confidence and often wrong. Check the tool's confidence score, prefer predictions built from the person's **own writing** over name-only lookups, and **update your read from observed behavior on the call** — never let a label override what the buyer actually does. - **Personalize the delivery, not the substance.** Adapt tone, pace, order, and proof to the type; don't fabricate different facts for different people. - **Compliance is real: personality prediction is profiling.** Profiling people from public data is automated processing of personal data under GDPR/CCPA. You (not the tool) are the data controller — rely on a lawful basis (usually legitimate interest), disclose it in your privacy notice, and honor access/deletion requests. Don't use these predictions for hiring/credit decisions. - **Choosing a tool:** pick on prediction accuracy for *your* market, whether it predicts from writing vs just a name, native CRM/calendar sync, and whether you need an API/MCP (see the comparison below). Accuracy claims are vendor-reported — pilot on 20–30 known contacts before trusting it at scale. - **It's insight, not outreach.** None of these tools send the email or run the sequence — pair them with a cadence tool (`/sales-cadence`). If you discover a tactic or correction not captured here, append it to `references/learnings.md`. ## Tools that predict buyer personality - **Crystal (Crystal Knows)** — the category standard. Predicts DISC (+ OCEAN/Enneagram/16Personalities) from LinkedIn/email/text; Chrome extension, HubSpot/Salesforce sync, a Personality API, and an MCP server for Claude/Cursor. Strong per-profile communication tips. Deepest coverage → **`/sales-crystal`**. - **Humantic AI** — DISC + OCEAN, positions on higher behavioral accuracy; native calendar/email integrations and a bulk personality dashboard. Common Crystal alternative. - **Humanlinker** — pairs personality analysis with AI outreach personalization (LinkedIn + email) and 360° enrichment — more of an "insight → message" tool than pure prediction. - **Happysales** — AI sales-research + personality insight aimed at outbound personalization. - **Adjacent:** conversation-intelligence tools (e.g. Gong/Chorus) infer communication style from calls rather than public data — different input, overlapping goal (see `/sales-coaching`). Accuracy is vendor-reported and depends on input richness — validate on your own contacts first. ## Gotchas > *Best-effort from research (2026-07) — validate vendor accuracy claims and plan gates against current sources.* - **Predictions ≠ ground truth.** Low-confidence guesses from sparse profiles are common; treat the type as a testable hypothesis and defer to observed behavior. - **DISC is a communication heuristic, not validated psychometrics.** It's useful for adapting delivery; don't over-index on it or use it to judge a person's worth. Big Five is the more research-backed model if you need rigor. - **Profiling triggers privacy law.** Predicting personality from public data is GDPR/CCPA profiling — you are the controller; document a lawful basis and honor data-subject rights. Not for hiring/credit decisions. - **Vendor accuracy numbers are marketing.** "X% accurate" is self-reported. Pilot on 20–30 people you know before rolling out. - **These tools don't send anything.** They're insight only — you still need a sequence/CRM to act on the read. - **Blends and context matter.** People flex style by situation (a D at work can be an S at home). One label per person under-describes them. ## Related skills - `/sales-crystal` — Crystal (Crystal Knows) platform help: profiles, Chrome extension, API, MCP server - `/sales-enrich` — Contact/company enrichment strategy (personality is one enrichment attribute among many) - `/sales-cadence` — Build the outbound sequence that acts on the personality read - `/sales-account-map` — Map a buying committee so you can tailor per-stakeholder - `/sales-coaching` — Conversation intelligence that reads communication style from calls - `/sales-idea-validation` — Generating fictional buyer/audience *personas* (different job — segments, not a named individual) - `/sales-do` — Not sure which skill to use? The router matches any sales objective to the right skill. Install: `npx skills add sales-skills/sales --skill sales-do -a claude-code` ## Examples **Example 1**: "Rewrite this cold email for a CFO whose Crystal profile says 'D'." → Rebuilds it D-style: outcome-first subject, 2–3 sentences, a hard number, one clear ask, no build-up — and notes to verify the read from any reply. **Example 2**: "My champion is an 'I' but the economic buyer is a 'C'. How do I run this deal?" → Tailors per stakeholder: vision/energy and social proof for the I champion, a data-and-documentation appendix and room to verify for the C buyer, and one artifact that serves both. **Example 3**: "Is DISC prediction accurate enough to base my outreach on, and is scraping LinkedIn for it legal?" → Explains prediction is a low-confidence hypothesis to test (richer input = better), that accuracy claims are vendor-reported, and that profiling from public data is GDPR-regulated with you as controller (legitimate interest + privacy-notice + data-subject rights). ## Troubleshooting **The predicted type doesn't match how the buyer actually behaves** - **Cause**: The prediction came from thin public data (a sparse LinkedIn profile) → low confidence, or the person is a blend/flexes by context. - **Fix**: Check the tool's confidence score; prefer a prediction built from the person's own writing; and update your read live from what they say and do. The label is a starting hypothesis, not a verdict. **Personalizing by personality isn't lifting reply rates** - **Cause**: You changed the *facts* per type instead of the *delivery*, or you're over-personalizing thin predictions at scale. - **Fix**: Keep the offer constant; vary tone/pace/proof/order only. Pilot on a small segment of known contacts, measure, and only scale the adaptations that move replies. **Legal/compliance flagged your personality-profiling workflow** - **Cause**: Profiling prospects from public data without a documented lawful basis or privacy-notice disclosure. - **Fix**: Establish and document legitimate interest (or another basis), disclose profiling in your privacy notice, provide access/deletion on request, and never use the predictions for hiring/credit decisions. Confirm specifics with counsel.