Direct answer
Because it cannot be promised, and promising it is non-compliant. The Chinese standard verifiable through official channels is T/CAPT 026—2026; it explicitly prohibits providers from promising answer dominance or permanent placement, and AIIA/T 0277—2026 — the corresponding alliance standard — takes the same line. 'Citation rate ≥ 70%' is exactly the kind of quantified guarantee the standard forbids a provider from making. We promise only a verifiable process: what actions were taken, what resulted, and how you re-check it.
Straight answer: it is a prohibited zone
The question clients ask most is 'can you guarantee AI will recommend me'. The honest answer is that nobody can — and anyone guaranteeing it is usually out of line.
Where the rule comes from
The standard verifiable through official Chinese channels is T/CAPT 026—2026. Among its constraints on service providers is an explicit prohibition on promising answer dominance or permanent placement. The corresponding alliance standard is AIIA/T 0277—2026.
In other words: a quantified promise such as 'citation rate ≥ 70%' is precisely what the standard forbids a provider from promising. Hearing that promise is itself a risk signal.
What we do promise
| Not promised | Promised |
|---|---|
| Position in an answer | An action list: what was optimised, what changed |
| A citation percentage | Process evidence: the record and re-check method for each step |
| A guaranteed timeline to results | An observation method: fixed questions, fixed frequency, stable-value rule |
| Answer dominance | Factual accuracy: every published fact has a source and is traceable |
Why this is worth more
A position is an outcome influenced by model versions, algorithm changes and competitors' actions — not within any single party's control. Actions, evidence and factual accuracy are exactly what a delivery party can control and you can accept. That is the asset you actually take ownership of.
One line: tools give you a score, we give you assets and acceptance.
Key facts
| k | v |
| k | v |
| k | v |
| k | v |
Sources
- T/CAPT 026—2026 relevant clauses
- AIIA/T 0277—2026 relevant clauses
- Our client-facing position paper (no quantified outcome promises)
Follow-up questions
Then how do I judge whether it worked?
Look at the controllable items: is the factual base accurate, has the source tier improved, is the process record complete and re-checkable. Those three can be accepted; a ranking cannot.
Others promise it and you don't — is that a capability gap?
It is not that we dare not, it is that we do not. Promising a quantified outcome is non-compliant and shifts uncontrollable variables onto your account — the client ends up bearing the cost.