Your club has 400,000 followers.
You can name about 200 of them.

Nobody has published a credible number, because almost nobody measures it account by account. On the first production scan CommunityOS ran, 78,181 followers of a Web3 project were processed and 90.96 percent were filtered as bots, farms or inactive shells, leaving 5,806 real accounts. Crypto audiences are the worst case for bot inflation, so a football or basketball club should expect a different figure. The honest answer for any specific club is that the number is unknown until the following is scanned.

Sponsorship valuation platforms such as Zoomph, Blinkfire and Hookit price exposure. They detect logos, count impressions across broadcast and social, and convert that into media value, which answers what a piece of inventory was worth. CommunityOS works one level below that, on the people. It scores every individual account in a following for authenticity and influence and returns a ranked list of real humans worth contacting. The two are complementary: valuation prices the audience, this identifies it.

A fan CRM knows the people who transacted: ticket buyers, members, merchandise customers, newsletter subscribers. That is first-party data and it is valuable, but it covers a small fraction of the audience. A club's social following is usually many times larger and almost entirely anonymous. CommunityOS maps that unmapped part, then hands back names your CRM has never seen.

That is the most common reason a partnerships team runs a scan. Sponsors and their agencies have started asking rights holders to evidence audience authenticity before renewal, and a follower count is not an answer. A scan returns how many accounts survived filtering, why the rest did not, and how the real audience breaks down across four archetypes, which is a defensible answer rather than an assertion.

The method runs on any X account's following, so a league, a club, a rights holder, an individual athlete or a sponsor's own handle all work the same way. The scoring engine is deterministic and rule-based, 60 percent linguistic features and 40 percent normalised vanity metrics, so the same inputs always produce the same outputs regardless of whose audience is being scanned.

Plans start at $499 a month with monthly billing, through $1,499 and $4,999 tiers, with enterprise from $10,000 a month and 17 percent off annual terms. A first scan and ranked queue land within days rather than an enterprise onboarding cycle. A small number of design-partner slots carry pilot pricing locked through 2027 and direct access to the founders.