Local Citations

Citation value is unevenly distributed: a handful of authoritative, trade-relevant sources matter far more than hundreds of empty web directories. Which handful depends on the trade — our research records exactly which sites assistants cite per market, and legal's list (Avvo, Justia, Super Lawyers) looks nothing like roofing's (BBB, Angi, Reddit threads).

The measured split is sharper than most citation advice implies. Across the 2,840 sources we recorded, directories accounted for about 21% of citations on average — but that average hides a range from roughly 8% in med spa answers to over 40% in some legal markets. In trades at the low end, assistants were reading businesses' own sites and discussion threads instead, which means a directory-first strategy would be optimising a channel the assistant barely consults.

The strategic implication is to stop thinking in counts. "We built 200 citations" describes effort, not evidence. The useful inventory is short: the two or three general aggregators that feed everything else, the licensing or association body for your trade, the vertical directories your specific market's assistants actually cite, and the local institutions with real pages. Beyond that, additional listings mostly create more places for your data to drift out of date.

Maintenance is where the real risk sits, because citations decay silently. Businesses move, phone systems change, a directory imports stale data from an aggregator, and suddenly three sources disagree about where you are. Every one of those disagreements is a small reduction in how confidently a machine can resolve you.

The word overlaps confusingly with AI citations — the sources an assistant links in an answer. The connection is real: strong local citations often become AI citations, which is precisely why they are worth keeping accurate rather than merely numerous.