Business / Digital

Five Checks That Earn Propane Dealers AI Search Citations

Business / Digital

Five Checks That Earn Propane Dealers AI Search Citations

AI Overviews appear on about 68% of local business queries, per Whitespark, and Local Falcon's whitepaper testing found no distance ranking beneath them. Five checks on a dealer's own site decide who gets cited.

The Situation

A propane buyer three miles from your yard can now ask Google a question and get an answer that never mentions you. AI Overviews now appear on roughly 68% of local business-type queries, according to a Whitespark study that ran hundreds of manual searches across six local industries and multiple markets; separately, Local Falcon's whitepaper testing found they do not rank results by distance the way the map pack does. That breaks the oldest rule in local marketing for a propane company. The citation goes to whichever business the engine can read, understand, and place on a map of its own. Most independents have not moved on any of this yet, so the advantage is still open to the first yard in a market that does.

The Facts

What the click data actually shows. In a study published in July 2025, Pew Research Center analyzed roughly 68,000 real queries from March 2025 and found users clicked through to a result 8% of the time when an AI summary was present, against 15% when it was not. Google disputes Pew's methodology, so don't hang your hat on the exact spread. The direction holds across other measurements. Seer Interactive put organic click-through on AI Overview queries at 0.61%, down from 1.76%. The name inside the summary is what a buyer sees first now.

The map pack's rule does not carry over. For two decades, local visibility rewarded one thing above all: being close. Google's local pack weighs proximity heavily, so a dealer with a yard in town surfaced for buyers in town. Local Falcon's testing found AI Overviews don't work that way — no distance-based ranking underneath them. Nearness stopped being the deciding input. What decides it is whether an engine can crawl the site, tell what's sold there, and confirm the delivery footprint.

The answer names fewer companies. A Sterling Sky analysis across 322 local markets found AI local results surface only about 32% as many unique businesses as traditional map packs. Fewer slots. For the operator who moves first, that scarcity works in his favor — very few independents have touched any of this yet.

The file that decides whether you exist. Type your domain followed by /robots.txt into a browser. Read every line. A single disallow aimed at GPTBot, ClaudeBot, OAI-SearchBot, PerplexityBot, or Google-Extended removes the site from that engine entirely. Plenty of dealer sites inherited a blocking line from a web builder's default template years ago and nobody has looked since.

Say plainly who you are. Organization or LocalBusiness JSON-LD, with the company name, street address, and phone matching the Google Business Profile character for character. Add sameAs links to the state propane association listing, the chamber page, the Facebook page. Engines cross-reference those to confirm the delivery footprint and match the company to local queries. A phone number that differs by a formatting quirk between the site and the profile weakens the match.

Draw the radius you actually run. Service-area schema takes a GeoCircle. Put the honest number in it: the 35 miles you actually run from the yard rather than a statewide claim. Over-claimed coverage gets contradicted by everything else the engine reads, and a contradiction is worse than a modest, verifiable footprint. A dealer running two offices publishes two circles.

Answer the question inside one passage. AI systems lift self-contained blocks. Answer "how long does a 500-gallon tank last in winter" completely in one place and the passage is quotable; scatter it across three sections and the engine moves on. Same for service-area pages: real detail about the townships served beats a city name swapped into a template.

Mentions carry more weight than links. Ahrefs analyzed 75,000 brands in August 2025 and found branded web mentions correlate 0.664 with AI Overview inclusion, against 0.218 for backlinks — roughly three times stronger a signal for earning a mention inside the answer. The citation-factor research published separately by Whitespark points the same direction, with "best of" list placements ranking as a top citation factor. State association member directories, local trade coverage, and community sponsorship pages all count as mentions.

Business Impact

BrightLocal's 2025 SMB marketing research puts small-business SEO spend around $497 a month for the roughly 74% of firms investing at all — near $6,000 a year. The five checks above cost an afternoon of someone's time and no new line item. Second-order effect: as AI answers name fewer businesses per query, each citation carries more of the call volume it used to split with nine other listings. A UPrinting survey found 25% of small businesses say they have already seen clients go to AI-visible competitors. Fix robots.txt, schema, and radius accuracy this month, and the citations start pointing at the delivery radius your bobtails already run.

Key Takeaways

  • Whitespark's study of real local searches found AI Overviews fronting about two of every three local business queries, and Local Falcon's separate whitepaper testing found no distance ranking beneath them. A buyer sees the name inside the summary before any map pin, and site legibility decides whose name that is, wherever the nearest yard sits.
  • AI local results name only about 32% as many businesses as map packs across 322 markets (Sterling Sky), which makes each earned citation worth more traffic than a map-pack slot.
  • One disallow line for GPTBot, PerplexityBot, or Google-Extended removes a site from that engine completely, so a template robots.txt can erase visibility nobody knew was lost.
  • Branded mentions correlate 0.664 with AI Overview inclusion versus 0.218 for backlinks (Ahrefs, 75,000 brands), meaning association directories and local press outrank link-building for this purpose.
  • Pew measured 8% click-through with an AI summary present versus 15% without, on roughly 68,000 queries — Google contests the method, but being named in the summary matters more than ranking beneath it.

Action Steps

  1. 1.Open your domain's /robots.txt in a browser today and remove any disallow line targeting GPTBot, ClaudeBot, OAI-SearchBot, PerplexityBot, or Google-Extended.
  2. 2.Compare the company name, street address, and phone on your site's LocalBusiness schema against your Google Business Profile, and fix every character that differs.
  3. 3.Set the service-area GeoCircle to the real delivery radius you run from each office — one circle per yard — and delete any statewide coverage claim.
  4. 4.Rewrite your single highest-traffic page first so it answers one customer question — the how-long-does-a-500-gallon-tank-last kind — completely within one passage, then take the next two as time allows.
  5. 5.Claim or correct your listing in the state propane association directory and the local chamber, then ask your two nearest trade contacts to list you.

Competitive Advantage

Most independent dealers have not audited robots.txt or service-area schema even once. Pull those five checks into a single afternoon this month, start with the robots.txt file, and you claim citations in your own delivery radius before the question of who owns them is settled locally.

Have you read your own robots.txt file this year, and does it let AI crawlers reach your service-area pages?

Get the next one in your inbox.

Free propane industry news, twice a week. 30,000+ operators already read it.

No spam, unsubscribe anytime. We'll email you a link to confirm — that's how we keep the list real. See our privacy policy.

Published by PropaneInsider.com

AI-powered propane industry news for 30,000+ professionals