No, and you should be wary of anyone who does. Nobody controls what an AI model outputs, so a guaranteed citation is a promise about something outside the agency’s control. What we commit to is the work, the measurement, and transparent reporting on the trajectory, including when it is not going well.
The short version
- No agency controls model outputs, so no agency can guarantee a citation.
- A guarantee is usually either meaningless in the small print or a compliance risk.
- What can be committed to: scope, method, measurement, reporting and responsiveness.
- In regulated sectors an unsubstantiated performance guarantee is a problem in its own right.
- The right question is not “do you guarantee” but “what happens if it does not work”.
Why the guarantee cannot be real
Citation depends on the retrieval and selection behaviour of systems operated by OpenAI, Google, Anthropic, Perplexity and Microsoft. Those systems change without notice, they weight sources differently from each other, and they produce different answers to identical questions on different days.
An agency can influence its inputs: how clearly your entity is defined, how extractable your content is, how much independent corroboration exists. It cannot control the output. Anyone guaranteeing a specific citation is guaranteeing somebody else’s behaviour, which is not a thing they can do.
When you read the small print, guarantees usually resolve into one of three things: a promise of activity dressed as a promise of outcome, a metric loose enough to always be satisfiable, or a refund offer priced into the fee so you were funding your own insurance.
In FCA, SRA or CQC-regulated sectors there is a further problem. An unsubstantiated performance claim is a regulatory exposure for the firm making it and, by extension, a risk to the client who repeats it. We price regulated work accordingly and we do not make claims that would fail that test.
What can honestly be committed to
- Scope. What will be done, in what order, and to what standard.
- Method. A published methodology you can inspect before signing, not a black box. Ours is at the Synaptic Authority Engine.
- Measurement. A fixed query set, measured monthly, on the same platforms, reported whether the news is good or not.
- Baseline honesty. A starting position recorded before work begins, so improvement can be demonstrated rather than asserted.
- Responsiveness. If the trajectory is flat after a defined period, changing the approach rather than continuing to invoice.
Better questions to ask any agency
- “What will you show me monthly, and what does a bad month look like in that report?”
- “What is your baseline measurement, and when is it taken?”
- “At what point would you tell me this is not working?”
- “What do you specifically not control?”
- “Can I see the methodology before I sign?”
An agency that answers those directly is telling you more than a guarantee ever would. One that cannot answer the fourth is either inexperienced or not being straight with you.
More on this in red flags when hiring a GEO agency, which includes the ones that apply to us.