Glossary
Organic prompt
Mirrors how buyers actually type, not how marketers wish they typed.
Best mixed with controlled templates for stable week-to-week scoring.
Definition
An organic prompt is a question phrased the way a shopper or practitioner would naturally ask it—messy synonyms, local language, missing brand names, sometimes wrong category terms. It contrasts with synthetic prompts, which brands craft to optimize mentions. Organic prompts reduce blind spots but add variance, so you need enough volume to see signal.
Divergence example
| Prompt type | Mention rate | Correctness |
|---|---|---|
| Synthetic | 62% | 81% |
| Organic | 37% | 64% |
Interpretation: brand performs well in curated phrasing but underperforms in real-world language. Priority should be vocabulary and FAQ rewrites, not just more landing pages.
Add 5-10 fresh prompts from support logs each month to keep this block representative of live buyer phrasing.
How it's computed
Getllmspy blends catalog organic prompts with brand-safe templates so scores stay comparable. When you add custom organics, they pass safety filters, then join the prompt pack rotation. Fanout queries extend organics with controlled variations.
Representation check
Organic coverage ratio = organic prompts / all prompts
Target range for mature packs: 40% to 70%
How it works in practice
Organic prompt sourcing channels
- Search query reports (non-branded and long-tail)
- Sales call transcripts
- Support ticket headlines
- Community forums and review sites
These sources reveal language patterns teams often miss in synthetic copy.
How to read it
If organic and synthetic packs disagree, trust the organics for positioning copy and the synthetic pack for executive KPIs—document both. Divergence often reveals jargon your customers never use.
Track this divergence explicitly in monthly reviews; it is usually your best indicator of message-market language mismatch.
Organic prompt program (2 weeks)
- Export top 200 non-branded queries.
- Remove duplicates and obvious spam.
- Pick 30 prompts across intent clusters.
- Run with current synthetic pack.
- Track divergence and patch language on source pages.
- Review newly surfaced language monthly with support and sales teams.
Organic quality bands
| Signal | Weak | Healthy |
|---|---|---|
| Organic vs synthetic score gap | >15 points | <8 points |
| New vocabulary captured monthly | <10 phrases | 30+ phrases |
| Prompt drift monitoring | ad-hoc | monthly refresh cycle |
Lower gap means your published language matches buyer language.
Keep a changelog of newly discovered buyer phrases so copy teams can close vocabulary gaps systematically.
When to use
- Localization teams validating non-English phrasing.
- Category discovery before building a monitoring grid.
- Training internal stakeholders on real buyer language.