Why does PROEMA not guarantee citation rate?
PROEMA does not guarantee citation rate because no one can, LLM behaviour is non-deterministic, citation algorithms change without notice, and a vendor-side change can shift the ground without notice. What PROEMA does guarantee: a measurable methodology, quarterly reporting, and a transparent score grid.
Why does PROEMA not guarantee citation rate, on the ground?
Reason one: non-determinism. The same query asked to the same LLM ten minutes apart can return slightly different answers. Temperature, sampling, freshness ranking, all introduce variance. PROEMA averages across 3 repetitions per question to reduce noise, but a 5-15% citation-rate variance is normal between runs. "Promising a fixed citation rate is promising what the underlying technology cannot guarantee," stresses Lorenzo Eeman, founder of PROEMA.
Reason two: vendor-side algorithm changes. The September 2025 Reddit drop on ChatGPT is the textbook example. SEMrush measured Reddit's share of ChatGPT responses collapse from approximately 60% to approximately 10% in two weeks, a structural shift OpenAI did not announce in advance. Any brand whose citation rate depended on being co-cited with Reddit lost ground overnight. The same risk exists for any concentration of authority signals.
Reason three: market manipulation incentive. If an agency guarantees a citation number, the temptation to manipulate the measurement, cherry-picked queries, paid panels, becomes structural. PROEMA refuses to put itself in that bind. We commit to method (versioned panel, multi-LLM, reproducible protocol) and to reporting transparency (quarterly score grid published to the client) rather than to a number.
What PROEMA does commit to
Four concrete commitments: (1) a versioned panel of vertical-specific test questions, frozen at audit and re-tested quarterly; (2) multi-engine coverage (ChatGPT, Perplexity, Gemini, Claude, Copilot) so a vendor shift in one engine doesn't blind the measurement; (3) a transparent score grid (the PROEMA 10-axis grid published in the methodology pages) so the client understands every movement; (4) operational improvements at agreed cadence, robots.txt, Schema.org, llms.txt, Wikidata, FAQ silos, press push when relevant. The aggregate of these commitments produces a measurable upward trajectory in 9 cases out of 10 across the GEO Rocket portfolio. We don't promise the slope. We promise the discipline. Sources: SEMrush 2025-2026 Reddit citation study.
| Item | Guaranteed |
|---|---|
| Citation rate target number | Never |
| Versioned test panel | Yes |
| Multi-LLM coverage | Yes |
| Quarterly reporting | Yes |
| Score grid transparency | Yes |
| Operational improvements at agreed cadence | Yes |