You can begin each of the four checks below this week. Verification, technical changes and genuine review collection may take longer, or need the account owner or a developer. Each step cites a specific source, but the sources measure different things, so the order follows effort and customer consequence rather than a proven ranking of effects.
1. Claim and fix relevant directory listings
In Yext's API-derived sample of 6.8 million citations, OpenAI model output drew 48.73% from controllable third-party listings.1 That makes relevant directory accuracy a quick, low-risk audit target. Verify the sources shown in repeated ChatGPT checks.
Go to each platform, search the business, and check three things: is the listing claimed, is the category correct, and does the profile have enough content — photos, a real description, a handful of reviews — for a crawler to have something worth citing. An unclaimed listing with a generic category is close to useless even if the business itself is excellent. Claiming and verification can take longer than one sitting, and some platforms require the account owner.
2. Build a current, representative review stream
In SOCi's 2026 Local Visibility Index, recommended locations averaged 4.3 stars on ChatGPT, 4.2 on Perplexity and 3.9 on Gemini.2 These are averages, not thresholds. Compare Google and relevant directory ratings with the actual local businesses customers see, then improve service, recency, response quality, and review coverage together.
If the business trails its visible competitors or has gone quiet, start a steady review request cadence with recent customers and respond to recent feedback. Do not offer incentives or pressure only happy customers. BrightLocal's 2026 survey found 74% of consumers trust only reviews from the last three months and 32% only the last two weeks.3
3. Check robots.txt for AI-search crawler blocks
Visit yourdomain.com/robots.txt and check the effective policy for these three named search crawlers: OAI-SearchBot, Claude-SearchBot, and PerplexityBot. If one is explicitly blocked, that search crawler cannot retrieve the affected pages. If robots.txt is missing, crawling is allowed by default. GPTBot and ClaudeBot are separate training crawlers, so those privacy choices are assessed on their own rather than misreported as search visibility.
Broad rules can block search crawlers by accident — don't remove them blindly
An accidental block can come from a broad User-agent: * / Disallow: / rule added by a site builder or security plugin before AI crawlers were a consideration. Check with whoever manages the site before changing it: a blanket rule may be deliberately blocking other bots too, or sitting alongside an intentional training-crawler choice elsewhere. If it's truly accidental, prefer adding explicit allow rules for the three search crawlers above over deleting the broader block outright, and retest a specific public page afterwards.
4. Rewrite your homepage's top paragraph
Read the first paragraph of the homepage as a stranger would. Count adjectives ("best," "trusted," "leading," "premier") against concrete, checkable facts (years in business, a specific service, a specific location, a specific number). The 2024 GEO (Generative Engine Optimization) paper reports gains of up to 40% more visibility on its own benchmark, varying by domain.4 "Trusted by the community" gives an AI model nothing to cite. "Family-owned since 2003, specializing in same-day emergency repairs" gives it a fact it can quote.
This doesn't require a copywriter — it requires cutting every adjective that isn't attached to a fact and replacing it with a fact that's already true about the business. An afternoon is enough for the homepage and the two or three most important service pages.
What to skip this week
Schema markup and an llms.txt file both get recommended constantly in AEO advice, and both sound like they should matter. The studies below found small or no measured effects, so treat both as lower priority this week.
Schema and llms.txt: measured, and measured small
Ahrefs tested schema markup on 1,885 pages that AI already cited heavily (each had 100+ AI Overview citations before schema was added), matched against 4,000 control pages, and found a statistically significant 4.6% decline in Google AI Overview citations, alongside small, not-statistically-significant changes for Google AI Mode (+2.4%) and ChatGPT (+2.2%).5 It does not cover pages AI is not citing yet, so treat schema as an accuracy layer, not a quick visibility win. In a separate study of 137,210 domains, 97% of published llms.txt files received no requests during May 2026.6 Schema is still useful for machine-readable accuracy; llms.txt can remain an optional, low-cost artifact. Neither should be presented as a proven ranking lever, and schema specifically shows no measured citation benefit for any AI product tested.
None of this means schema and llms.txt are permanently pointless — the AI ecosystem changes fast, and a null result today isn't a null result forever. It means that in a week with limited hours, the four accuracy checks above are a better use of time, and schema/llms.txt can wait for a week with more time.
"Spend this week on four accuracy checks: directory listings, reviews, crawler access and your homepage's opening claim. Schema and llms.txt can wait."
Notes and sources
1 Yext, "AI Citations, User Locations, & Query Context," Yext Research, Oct 9, 2025. 6.8 million citations from ~1.6 million questions across OpenAI, Gemini and Perplexity APIs, July-August 2025, four industries. In its API-derived sample, OpenAI model output drew 48.73% from controllable third-party listings and Gemini drew 52.15% from first-party websites. API citations can differ from direct user applications.
2 SOCi, "2026 Local Visibility Index," January 2026. Dataset: 2,751 multi-location brands, about 350,000 US locations. Benchmarks for ChatGPT, Perplexity and Gemini: share of locations recommended 1.2%, 7.4% and 11.0% (Google 3-Pack: 35.9%); average rating of recommended locations 4.3, 4.2 and 3.9 stars; business-profile accuracy 68.3%, 68.0% and 100.0%, measured against Google Maps data.
3 BrightLocal, "Local Consumer Review Survey" 2026. Sample: n=1,002 US consumers. Findings cited: 74% of consumers trust only reviews from the last 3 months; 32% trust only the last 2 weeks.
4 Aggarwal, Murahari, Rajpurohit, Kalyan, Narasimhan, Deshpande, "GEO: Generative Engine Optimization," submitted November 2023, published KDD 2024. Finding: "GEO can boost visibility by up to 40% in generative engine responses" on the paper's GEO-bench benchmark; the effect varies by domain.
5 Ahrefs, "schema markup and AI citations study," Louise Linehan, May 11, 2026. Sample: 1,885 pages that each had 100+ AI Overview citations in February 2025, before schema was added, matched against 4,000 control pages; difference-in-differences methodology. Finding cited: Google AI Overview citations fell 4.6% (statistically significant); Google AI Mode changed +2.4% and ChatGPT +2.2% (both statistically indistinguishable from zero). All JSON-LD types were pooled rather than isolated by type.
6 Ahrefs llms.txt study, 2026. Dataset: 137,210 domains. Finding cited: 97% of published llms.txt files received no requests during May 2026. ahrefs.com
RaveHQ Insights share our views, based on our research and the sources cited. They are general information, not legal, financial or professional advice; check what applies to your business before acting.