Brand Mentions: See Who Is Talking About You

Somebody wrote about you last week. Maybe a trade publication covered your product launch, maybe a forum thread recommended you, maybe a partner republished your press release. Unless you already pay for a media monitoring tool, you probably found out late, if at all.

Brand mentions are now part of the DataForSEO connection in AI Data Stream, on every plan. Tick Brand mentions under Tools the AI can use, add your brand terms, and ask your AI what was said about you, where, and how it compares with the months before.

What You Can Ask

The mentions come from DataForSEO’s Content Analysis index of pages across the web. Your AI gets two tools to read them:

• Mentions: the pages that mention your brand in a date range, from news sites, blogs and forums by default. Each one comes with the source, a snippet, the date DataForSEO found the page and a rough sentiment
• Trends: how many mentions per day, week or month, with the sites mentioning you most and the positive, negative and neutral split

The useful questions mix them with the rest of your data. “Who wrote about us this month, and did any of it send visitors?” is one question in chat: the AI lists the coverage, then checks GA4 for referral sessions from those sites. “Did the launch get picked up?” compares the weeks around a date. Add up to five competitors in the settings and “how do we compare?” works without naming anyone.

The new Brand Mentions Digest skill turns the latest coverage into a report: press and news, industry blogs, forum discussion, republished copies and noise, each in its own section. On paid plans you can schedule it for a weekly email.

One Story, Counted Once

Press releases get republished. A single announcement can appear on a dozen news network sites within a day, word for word. Count those as twelve mentions and a quiet week looks like a media blitz.

So copies of the same text are grouped. You see the story once, with how many sites carried it and a few examples. When trends are counted from stored mentions, they report both numbers: pages, and unique stories.

You Pay for a Page Once

Every DataForSEO call is billed to your own account, at around $0.03, or about $0.06 for a full page of 1,000 results. Brand mentions are historical: a page found last Tuesday is still the same page next month. Paying for it every time someone asks would be a waste.

So mentions of your brand terms and competitors are stored. The first question about a period fetches it. Asking again costs nothing, and later questions only pay for pages DataForSEO found since. Each mentions lookup makes at most three metered calls, and the AI tells you what a lookup cost, or that it came from storage for free. If a long period is not fully stored yet, the AI says the data is partial and the next question carries on where it stopped.

Turn on Sync brand mentions daily and new mentions of your brand terms are fetched once a day, so the AI reads them from storage instead of paying for them in chat. It is off by default. Competitors are only fetched when the AI uses them, so the cost stays predictable.

Edit the connection to see Stored mentions: the dates stored for each term, the last sync and what it has cost so far. Remove stored data deletes a term’s mentions when you no longer need them.

Every Brand Has Its Own Noise

Raw mention data is noisy, and the noise depends on the brand. A software company finds its name on directory and comparison sites. A retailer finds it on marketplace listings. A broker finds it in the footer of every partner site that shares its listings. DataForSEO’s page types help, but they are a rough guide: plenty of listing and directory pages are tagged as blogs.

So the connection has two filters you set for your own brand: Exclude mentions containing, for phrases that mark a page as noise, and Exclude domains, for your own sites and anything else that should never count. Because mentions are stored, changing a filter costs nothing and applies to everything already stored, straight away. The one exception is a very large brand stored as a sample: its sample is fetched again the next time the AI looks it up.

You do not have to work them out by hand. When the same sites or phrases keep showing up without being real coverage, the AI points them out and suggests what to exclude. If you agree, it updates the settings for you and says what it changed. It never changes them without asking, and a scheduled report only ever suggests.

Honest About the Data

A few things the AI is told to say plainly, because they matter when you read the results:

• Found, not published: Dates are when DataForSEO found a page. Many pages have no publish date, so an old article can turn up as a new mention
• Sentiment is rough: It is calculated automatically. The AI reads the snippet before calling anything negative
• Small brands, thin data: If there is little coverage, the AI says the data is thin rather than claiming nobody mentions you
• Big brands are sampled: A household name can collect tens of thousands of mentions a week. Storing them all would cost a lot and help nobody, so very large brands are stored as a sample of the most prominent news, blog and forum stories each week, and the AI is told never to present the sample as everything

Getting Started

Open a property’s Connections tab and edit your DataForSEO connection, or add one. Under Tools the AI can use, tick Brand mentions, enter your brand terms (the full name: a partial name can be a street or a common phrase), and optionally your competitors and exclusions. Then ask your AI what was said about you this month.

The setup guide is in the documentation.