Cited Methodology

We do not audit prompts. We investigate recommendations.

Cited investigates recommendations through defined travel buying situations, not isolated prompts.

The four questions every Cited engagement answers

In this order. Always.

If a page or engagement starts at question four, the thesis is lost. Cited always starts at question one.

What happens after the experience?

What customer, client, traveller or market evidence is created after the booking, stay, flight, tour or programme? Reviews, advocacy, complaints, ratings, case studies, media, awards what exists publicly?

Why can that matter before the next purchase?

How can that evidence become relevant to AI supported discovery and recommendation? Post-purchase reputation is becoming part of pre purchase discovery. That is the market shift Cited is built to serve.

Where does the Recommendation Gap emerge?

What does Cited observe between the evidence that exists and the AI recommendation gaps that exist today? Where is evidence strong but not reflected? Where is it genuinely absent?

What does Cited actually do about it?

Define the buying situation. Establish the baseline. Trace the evidence. Strengthen what is legitimate. Rerun the same journeys and measure what changed against the baseline.

The Cited method

Define. Baseline. Trace. Strengthen. Rerun.

Five steps. Applied to one defined buying situation at a time. Analyst led and never automated.

Define

Agree the specific buying situation: the recommendation to compete for, the traveller type and competitor set.

Baseline

Establish where the brand stands today across five AI channels. Document what AI currently recommends and with what associated evidence.

Trace

Investigate the customer, market and authority evidence behind the recommendation gap.

Strengthen

Improve genuine customer, brand and authority evidence where legitimate gaps exist. Grounded in what customers and the market already prove.

Rerun

Run the same defined buying journeys again. Measure what changed against the baseline. Every finding traced to evidence.

Recommendation Evidence States

What Cited finds and what it means.

After tracing the evidence associated with a recommendation pattern, Cited classifies what exists and what the appropriate response is. These are not scores. They are diagnostic states.

Own

Evidence is strong and aligned

Customer word of mouth reaches AI recommendation. The priority is maintaining and extending this alignment as the evidence environment changes.

Activate

Evidence exists but is not reflected

Strong customer evidence exists but is fragmented, unstructured or not sufficiently present in AI-accessible sources. The fix is activation, not creation.

Protect

Evidence is present but under threat

Existing evidence reaches AI recommendation presence, but negative signals or competitor momentum create risk. Requires active evidence management.

Fix

Evidence and experience are misaligned

Where public customer evidence does not support the recommendation the brand wants to win, no content strategy resolves this. The underlying experience requires attention first.

What Cited does not do

Precise about our own limits.

We cannot tell you why AI produced a specific answer.

AI systems do not disclose their reasoning. We find the gap between what customers say and what AI recommends — and measure what that gap costs. We do not claim causal access to the model.

We do not guarantee specific AI recommendation outcomes.

AI outputs vary by engine, date, query phrasing and context. We document the baseline, strengthen legitimate evidence, rerun the same journeys and measure what changed.

We do not manufacture evidence.

Every Cited intervention is grounded in what customers and the market already prove. We activate existing evidence. We do not create claims the underlying experience does not support.

We do not decide whether you are trustworthy.

We find the gap between what your customers and market prove and what AI actually recommends. The customer is the arbiter of the evidence.

We do not audit prompts. We investigate recommendations.

In this order. Always.

Define. Baseline. Trace. Strengthen. Rerun.

What Cited finds and what it means.

Precise about our own limits.

Want to understand how this applies to your brand?