What to do this week

Check your Wikipedia page. If it exists, verify it's current and accurate. If it doesn't, assess whether your organization meets Wikipedia's notability criteria and begin the process.

Audit your third-party profiles. Search for your brand on TripAdvisor, Google Business, Yelp, and the review platforms relevant to your industry. Are they complete, current, and actively managed?

Search Reddit for your brand name. See what conversations already exist. If your brand isn't part of relevant discussions in your space, that's a citation gap worth closing.

Boost Background

We audited a museum client last fall whose website was exceptional. Beautifully designed, rich with collection information, well-organized. Yet when we ran 50 prompts through ChatGPT and Perplexity, the museum barely appeared. A smaller institution across town with a fraction of the budget showed up three times more often. The difference wasn't content quality. It was citation footprint. The smaller museum had an active Wikipedia page, consistent TripAdvisor reviews, mentions across travel blogs, and a presence on Reddit threads about local attractions. The larger museum had a great website and almost nothing else.

That pattern repeats in nearly every audit we run. What AI says about your brand depends far more on what the rest of the internet says about you than what your own website says.

The numbers behind third-party trust

AirOps research found that 85% of brand mentions in AI answers originate from third-party pages rather than brand-owned domains. Brands are roughly 6.5 times more likely to be cited through external sources than their own websites.

SE Ranking's study of 129,000 domains identified referring domains as the single strongest predictor of whether ChatGPT cites a source. Sites with over 32,000 referring domains are 3.5 times more likely to be cited than those with just a few hundred. Presence on review platforms like Trustpilot, G2, and Yelp correlated with 3 times higher citation rates. And domains with significant Reddit and Quora activity showed roughly 4 times the citation likelihood compared to those with minimal presence.

Yext's analysis of 6.8 million citations added another dimension: 86% came from sources brands can directly manage or influence, split between owned websites (44%) and third-party listings (42%). That second category is where most brands have gaps they don't realize exist.

What a citation footprint actually looks like

Think of your citation footprint as every place on the internet that feeds information about your brand into an AI model. Wikipedia, review profiles, directory listings, press mentions, forum discussions, podcast transcripts, anywhere your brand name appears alongside relevant topic language.

The platforms that carry the most weight vary by model. Gemini pulls over 52% of its citations from brand-owned websites. ChatGPT leans on third-party sources, with Reddit alone accounting for nearly 30% of its U.S. citations. Perplexity draws from industry-specific directories and review platforms more than either competitor.

For premium brands in hospitality and cultural institutions, the highest-leverage sources tend to be Wikipedia, TripAdvisor and Google Business Profile, press coverage from recognized publications, and active review profiles on vertical-relevant platforms.

Where we see the biggest gaps

Three patterns show up repeatedly in our audits. First, Wikipedia. Many established brands either have no Wikipedia page or have one that's outdated, thin, or flagged for issues. Given Wikipedia's dominance in AI citation data, this is one of the highest-return investments a brand can make, and it costs nothing but time and adherence to Wikipedia's editorial standards.

Second, review platforms. Brands often focus on collecting reviews on their own website while neglecting the third-party platforms AI actually reads. A handful of detailed, specific reviews on TripAdvisor or G2 carries more weight in AI answers than hundreds of testimonials on your homepage.

Third, Reddit and forum presence. SE Ranking's data showed that domains with millions of Reddit mentions receive roughly 4 times more ChatGPT citations. You don't need millions. But genuine, helpful participation in relevant communities creates signals that AI systems pick up. The key word is genuine. Promotional posts get downvoted and ignored. Useful contributions in threads where people are asking real questions get remembered by both the community and the models trained on that content.