Docs

Persona Voice Deep-Dive

Question it answers: For a chosen stakeholder layer — C-suite, Director/operator, champion/EB, or end-user — what are they actually saying in their own words (concerns, objections, language, success metrics), and which of those concerns correlate with losing? Cadence: Monthly (run per persona layer) GTM owner: PMM

Data sources

Internal: interactions filtered by persona layer — role_level (seniority), is_champion / is_economic_buyer / champion_score (stakeholder role on deal_stakeholders), and speaker_type = 'external' (buyer voice, not rep). Raw utterances attributed to the persona. interaction_enrichment topic, sentiment, and psychographics tags. Each theme is joined to deal OUTCOME (deals Closed Won / Closed Lost) and deal_qualification (binding_constraint, outcome_score) so a theme's prevalence can be correlated with how deals resolve. cluster_search with target: 'interactions' for the thematic backbone.

External: None — unfiltered first-party voice is the entire value. Optionally external_search { action: enrich_company } on the speaking accounts to segment who's speaking by firmographics.

Living Document

A monthly deep-dive into one persona layer's voice. The persona layer is the parameter — run it for C-suite this month, operators next, champions/EBs the month after. This is the general engine; #3 Champion / EB Voice Digest is the champion/EB preset of this same pattern. What this doc adds over #3 is (a) any-persona scope, not just champion/EB, and (b) a win/loss correlation on every theme.

Sections:

  • Who we're hearing from — breakdown of the persona this run: roles, seniority levels, industries, number of companies and utterances represented
  • Top themes — 6–10 strategic themes for this persona, each with 2–3 direct quotes (with interaction_id) and a brief synthesis of what's being signaled
  • Objection & fear patterns — the concerns and hesitations specific to this persona (the buying-decision blockers)
  • Language they use — recurring words and phrases; the vocabulary that resonates, for messaging and persona cards
  • Win/loss signal — for each major theme, how much MORE it appears in lost deals than won deals (the delta) — flags which concerns predict losses (e.g. "team-displacement concern +36% in lost; budget/ROI +86% in lost")
  • Metrics they name — what success looks like in this persona's own words, ranked by frequency (operators: pipeline > conversion > meetings > ROI)
  • What's new vs. last run — themes or language that emerged or shifted since the prior run for this persona

Monthly cadence means this becomes the ground truth for persona cards, messaging decks, and sales enablement — per persona layer. Versioned; PMM reviews and promotes.

Page

A live persona-voice view. The persona selector is the parameter — pick the stakeholder layer and the whole page re-scopes.

Persona selector: C-suite / Director-operator / champion-EB / end-user (filters on role_level + stakeholder flags).

Stats row:

  • Companies represented (this persona, this period)
  • Utterances mined
  • Avg outcome_score of deals where this persona is active

Bar chart: Theme frequency for this persona — top 15, sorted descending.

Bar chart — Win/loss delta per theme: Diverging bars showing each theme's prevalence in lost minus won deals — the "which concerns predict losses" view.

Radar chart: Theme coverage by sub-persona within the layer (e.g. CEO vs. CRO vs. CFO) — shows where their concerns differ.

Table — Voice quotes: One row per representative utterance — quote, company, role_level, theme, deal stage, outcome (won / lost / open), interaction_id.

Word-frequency view (table with count): Top 20 phrases/terms used by this persona, ranked by frequency.