Endurance brand voice

You extract a voice profile for an endurance-sport brand with the specific lexical patterns and credibility markers that distinguish genuine endurance writing from generic sports-lifestyle marketing.

This is an extension of the brand-voice-extraction skill. Run the standard extraction first, then this extension. The result is a voice profile that captures the structural patterns (sentence length, cadence, openers) plus the endurance-specific lexicon and credibility tells.

Inputs to gather first

Beyond the standard brand-voice inputs:

1. Sport(s) the brand serves — cycling road / gravel / mountain, running road / trail / track, swimming, triathlon, mountaineering, skiing, climbing, multi-sport. Sub-profiles per sport if the brand spans multiple.
2. Category context corpus — 8–12 pieces from credible endurance writing the brand sees as exemplary canon (race reports from respected publications, training journals from credible athletes, equipment essays from publishers with category trust). These anchor the credibility calibration.
3. Audience tier — what level does the brand’s actual customer race at? Elite (1% of audience), competitive (top 25% of finishers at amateur races), participation-focused (general field). Voice should match the customer, not the most extreme tier.

The pipeline

Run after the standard brand-voice-extraction skill produces .lens/voice-profile.json. This skill extends that file with endurance-specific fields.

Phase 1 — Endurance lexicon extraction

SYSTEM: You analyse endurance-sport brand writing for category-
specific language patterns. You distinguish credible endurance voice
(writing by people who have done the work) from generic sports-
lifestyle voice (writing about people doing the work). You return
verbatim phrases from the corpus. You do not invent.

USER:
Brand corpus: {BRAND_CORPUS}
Category context corpus: {CATEGORY_CONTEXT_CORPUS}
Sport(s): {SPORTS_LIST}

Extract for the brand, per sport if multi-discipline:

{
  "terrain_and_conditions": [
    {"phrase": "<verbatim>", "frequency": "<common|occasional|rare>"}
  ],
  "time_of_day_patterns": [...],
  "distance_duration_framing": [...],
  "discomfort_vocabulary": [
    {
      "phrase": "<verbatim>",
      "frequency": "<common|occasional|rare>",
      "tone": "<honest|euphemistic|absent>"
    }
  ],
  "quietness_signals": [
    {"phrase": "<verbatim>", "context": "<one sentence>"}
  ],
  "category_credibility_score": <0-10>,
  "category_credibility_evidence": [
    "<observable patterns supporting the score, verbatim>"
  ]
}

Calibration:
- Compare brand corpus phrases against category context phrases.
- "Common" means in 30%+ of brand pieces.
- "Honest" tone uses discomfort vocabulary directly; "euphemistic"
  softens or motivational-isms over it; "absent" means the brand
  doesn't reference discomfort.
- Credibility 10 = indistinguishable from canon, 5 = competent but
  generic, 0 = clearly outside-looking-in.

Rules:
- Verbatim only.
- Empty fields where the corpus genuinely lacks them.
- Evidence must point to observable patterns, not impressions.

Phase 2 — Inauthenticity tells

Generate the endurance-specific “never does” list:

• Hyperbolic outcome framing (“crush your goals”, “unleash your potential”)
• Generic motivational closers (“you’ve got this”, “earn the day”)
• Mis-used technical terms (5k called a “marathon”, aerobic ride called “intense”)
• Comparison of recreational athletes to elite times
• Gear or technique references the brand can’t credibly speak to

Phase 3 — Channel-mapped prompt extension

Extend the prompt library with endurance-aware versions of the recurring outputs. Examples:

• Race-week email subject — includes the specific race name and one tactical observation, never generic “Good luck this weekend!”
• Long-form blog opener — uses the brand’s typical opener pattern plus an endurance-specific anchoring detail (terrain, weather, time-of-day)
• Athlete-introduction social post — names the athlete’s actual race results, names the courses, doesn’t inflate

The eval harness

Standard voice rubric plus three endurance-specific gates:

Endurance E1 — Credibility floor. Drafts score against the category-credibility rubric (terrain accuracy, duration plausibility, discomfort-vocabulary tone match, gear-reference accuracy). Below 3/4 = doesn’t ship.

Endurance E2 — Hyperbole regex. Hard block on the inauthenticity phrases list. Auto-suggest a verbatim replacement from the brand’s actual canon.

Endurance E3 — Quietness presence. Across a campaign or month, ≥20% of pieces include a quietness signal (training alone, the early start, the unwitnessed PR, the work no-one sees). Endurance brands that only celebrate the win sound shallow.

Failure modes to watch

• Category context corpus dilutes the brand voice. If brand outputs start sounding like generic race-report writing, the category corpus is over-weighted — reduce its share.
• Founders who don’t ride / run / swim — the pipeline can produce passing copy but the founders will eventually be exposed. Push back. Recommend they do the work.
• Over-claiming on a single founder’s credibility — separate the “founder voice” from the “brand institutional voice.” Different pieces use different ones.
• Wrong sport’s vocabulary — cycling-language has poor overlap with running-language. Multi-discipline brands need sub-profiles per sport.
• Performative struggle — if the brand’s actual customers are recreational 5k runners, don’t write like the brand serves Ironman finishers. Match the voice to the audience’s actual practice.

Capability boundary

Reliable: lexical pattern extraction, structural analysis, drafting against the extended profile.

Unreliable: distinguishing genuine endurance voice from confidently- written tourist voice without anchor examples. The category context corpus is the anchor. Don’t run this skill without it.

The pipeline does not give the brand credibility it hasn’t earned. A founder who hasn’t trained will eventually be exposed regardless of how voice-correct the copy is. See the capabilities reference for the wider context.

Hand-off

Extended voice profile feeds all downstream drafting skills (eval- gated-drafting, social-content-factory, lifecycle-journey-builder, earned-media-pitch, race-result-content-engine). They all read .lens/voice-profile.json for the structural pattern and .lens/endurance-voice-extension.json for the lexicon.