The 14-day method: measure, build, place, re-measure

"Get cited by AI" sounds like a vibe. It isn't — it's a sequence of specific, checkable steps on a fixed clock. Here's the actual cycle we run on every topic, and where the honesty gates sit.

Day 1 — lock the battery and measure

We start by agreeing the prompt battery: the 20 real questions your buyers type into an AI engine. Then we run the baseline — 20 prompts × 5 repetitions × 4 surfaces (ChatGPT, Perplexity, Google AI Overviews, Claude) = 400 real AI answers, each one stored with its timestamp. That's the world as it is today, on the record, before we touch anything.

Day 1 — the go/no-go gate

Right after the baseline, we run forensics on it. If the data says the topic isn't winnable — too few prompts have a claimable or pitchable citation surface — we tell you before any fulfilment work begins. No work happens before a recorded "go." We'd rather be honest about a difficult topic than take the work.

Days 2–4 — on-page rewrite

We rebuild your target page to the citation spec: answer-first structure, the claims an engine can lift cleanly, the evidence and sourcing that make it quotable. This is the groundwork — necessary, rarely sufficient on its own.

Days 8–10 — entity + placement

Two levers in parallel:

• Entity build. A Wikidata entity and sameAs hygiene, so engines resolve your brand as a thing they know, not an ambiguous string.
• Placement. A listicle placement on a DR50+ publisher. This is the part most "AEO" advice skips and the part that actually moves citations — answer engines lean heavily on third-party "best X" lists when they decide who to name.

Day 14 — endline and the receipts

We re-run the exact same battery — same prompts, same reps, same surfaces — and read the movement per surface. You get an evidence pack: baseline versus endline, the dated transcript behind every number, and the citation rate per engine.

Why the symmetry matters

The whole method rests on one discipline: the endline is the same measurement as the baseline. Same battery, same method, no moving the goalposts. That's the only way a "+36 points on ChatGPT" claim means anything — and the only kind of claim we're willing to make.

Want to see the first step on your own URL? Start free and run a real scan today.