Citation gap outreach

Use when the sources feeding AI answers in a category are known and the brand needs to get into them. Produces a ranked get-listed plan with the right ask per source and a re-measure loop that proves each placement landed.

Rank by answers fed, not authority

Order target sources by how many category answers each one fed. Domain authority, traffic, and prestige all mislead here: a modest directory feeding nine answers outranks a famous publication feeding one. The citation count is the value of the placement; everything else is habit carried over from link building.

Match the ask to the source

Each source type gets updated a different way, and the ask must match it. A directory or listing site wants a submission or a claimed profile. A comparison or review page wants inclusion with evidence — real differentiators, pricing, proof — not a pitch. A community thread is entered honestly, with a real answer from a real account, or left alone. A publication wants the data or story its page is missing. Write the ask so the person who maintains that page can act on it in one sitting.

Run it like pipeline

Track each target as a row: source, page, answers it fed, whether it mentions the brand today, ask type, the human contact, status. Work the top of the list first; small placements land fastest and compound.

Verify the placement landed in answers

After a placement goes live, re-ask the questions that source was feeding and compare. If the answers do not move within a re-measure cycle, either the source mattered less than the count suggested, or the mention lacks the substance engines quote — specifics, numbers, and comparisons get cited; adjectives do not. Fix the mention before adding new targets.

What good looks like

The best operators notice the shape of what engines quote from each source — a comparison table, a price, a spec — and ask for that placement, not a logo on a partner wall. The mediocre version is generic link-building outreach re-sent with new words in the subject line, ordered by domain authority, never re-measured. Good output attaches evidence to every target — the answers that source fed — so each outreach message writes itself, and every closed row has a before-and-after answer pair proving it mattered.

Rules

• MUST rank targets by answers fed.
• MUST re-measure the affected questions after each placement.
• NEVER astroturf community threads.
• NEVER present paid placements as organic mentions.