Mention Frequency
Mention frequency is how often your brand is named across web sources and AI answers in a period. Learn how to measure it and where the number misleads.
TL;DR
Mention frequency is how often your brand is named in a given period, counted across articles, reviews, forums, social posts and AI answers. It is a volume measure, so it rises when the market talks more about you and falls when attention moves elsewhere. Read it beside sentiment, because volume alone cannot tell a launch from a crisis.
Why mention frequency matters
Frequency is the base layer under most brand measurement. Share of voice, sentiment and citation analysis are all calculated from a mention count, so the count has to be right first.
- Campaign readout: A launch either produced mentions or it did not, and frequency shows that within days rather than quarters.
- AI eligibility: Engines assemble answers from sources that name brands, so how often you are named across the web shapes how often they can name you.
- Anomaly detection: A sudden spike usually means something happened, and finding out what is the point of watching the number.
- Baseline setting: Every relative metric needs a denominator and a starting level, both of which come from frequency.
How to measure mention frequency
Define the trigger set
Brand name, product names, domain, common misspellings and key executives. Whatever you include, apply the same rule every period or the trend is meaningless.
Split web from AI
Published web mentions and mentions inside generated AI answers behave differently and move independently. Counting them together hides both.
Deduplicate carefully
Syndicated articles, cross-posts and a single answer naming you three times can all inflate a raw count. Decide upfront what counts as one mention.
Segment by surface
Track reviews, news, social, forums and AI answers separately so a spike traces to its origin without further digging.
Weight by reach
A mention on a widely read page is not equivalent to one in a dead thread. Reach weighting turns a count into an attention estimate.
Normalise the window
Compare equal periods. Monthly counts against a five-week month produce false growth that survives into slide decks.
Mention frequency vs. share of voice
Mention frequency: An absolute count of how often you were named. It rises and falls with total market activity, so a strong quarter for the whole category lifts everyone.
Share of voice: Your mentions as a percentage of all brand mentions in the category. It controls for market activity, so it shows whether you gained ground rather than whether the category got louder.
Read frequency first to see whether anything happened, then share of voice to see whether it mattered competitively.
Web monitoring counts what got published. It cannot count how often engines choose to repeat your name, which is what buyers actually encounter. Findrix measures how often each of seven AI engines names you across a market-sized prompt set, with the confidence range under every rate and named competitors measured the same way. Every gap comes with the fix already written: technical, content and off-site. The audit is free, takes about a minute, and requires no signup.
Metrics to read alongside mention frequency
- Share of voice: The same count expressed against category totals, which is the competitive version of the number.
- Sentiment split: The positive, neutral and negative breakdown, since frequency counts mentions without reading them.
- Reach-weighted volume: Mentions scored by audience size rather than treated as equal units.
- Linked ratio: How many web mentions carry a link, which is your outreach backlog.
- AI mention rate: How often engines name you across a fixed prompt set, tracked apart from web volume.
Where mention frequency misleads
The most common failure is treating a spike as a win. A product recall, an outage or a critical thread all produce more mentions than a good quarter of marketing, and a frequency chart cannot tell the difference. Sentiment has to sit beside the count for it to mean anything.
Volume also flatters the wrong sources. A thousand mentions in low-traffic forums register identically to fifty in publications engines actually cite, which is why reach weighting and source quality matter more than the headline total.
Version your trigger list and date every edit. Adding one product name can move volume fifteen percent and quietly break every historical comparison.
Frequently asked questions
What is a good mention frequency?
There is no benchmark, because the number depends entirely on your category size, brand age and what you counted. Your own trend on a stable trigger set is the only meaningful reference, with share of voice added when you need a competitive read.
How is mention frequency different from mention rate?
Frequency is an absolute count over a period. Rate is usually a percentage against some denominator, most often the share of tracked AI prompts where you are named. Vendors use both terms loosely, so check what the denominator is before comparing numbers between tools.
Does mention frequency affect AI visibility?
Indirectly and substantially. Engines learn which brands belong to a category from how often and where they are named across third-party sources, so sustained mention volume in sources engines trust makes it likelier they name you. Volume in sources they never read changes nothing.