← Newsroom

Do followers predict NIL roster pay? Results from 112 athletes

We studied 112 athletes with published contract values to test whether social media audiences help explain what schools and collectives pay. After accounting for on-field production and program scale, we could not detect a relationship between audience size and roster pay.

Social media advice often focuses on brand deals

College athletes are often advised to grow their following and engagement to attract sponsors. NIL Club, for example, describes follower thresholds for brand deals:

“Many brand deals are only available to athletes who meet certain follower counts or engagement benchmarks, so growing your audience can open doors that were not there before.”

NIL Club, How to Grow Your Social Media as a College Athlete

Follower counts can help athletes qualify for brand deals, where sponsors pay to reach an audience. Our study examines a different source of income: payments from schools and collectives.

School and collective payments are a separate market

An athlete’s earnings now come from two very different places. Commercial money is what brands pay for endorsements. Roster pay is what the school and its collective pay to have the athlete on the team, and since revenue sharing began it has become the dominant number for football and basketball athletes at this level.

Our roster‑pay benchmarks come from ESPN’s survey of more than 20 college general managers and agents (College football 2025: How much does each position cost?). They run from $200K–$400K for a tight end to $1M–$2M for a quarterback. The athletes in this study carry school and collective contracts between $500K and $6.5M. For context on the earlier commercial market, in the three years before revenue sharing existed, a top‑25 football earner did roughly 22 paid activities a year (Opendorse, NIL at 3, covering July 2021 to June 2024). That figure is cited only for the market it describes — the one that existed before schools began writing checks directly.

Our own research on college sports money puts the institutional side in context (NIL Collectives Are 5.5% of the Money): across 54 programs whose funders we could verify, 51 booster foundations raised $1.69B against 12 NIL collectives’ $98.6M — a 17‑to‑1 gap, with a single athletic association out‑raising every identified collective combined. Those filings describe organizational revenue. They do not report individual athlete compensation.

Brands value the audience an athlete can reach. Schools and collectives also consider what that athlete contributes to the team. A factor that predicts endorsement income may have a different relationship with roster pay.

We tested whether follower count added information about roster pay after accounting for position, on-field production, and program revenue.

Production and program scale were associated with higher pay

The estimates below use a common scale, roughly the difference between an average and an above-average value for each factor. The production estimate was about three times the follower-count estimate, and its range excluded zero.

Treat endorsement income and roster pay separately

These results give athletes little reason to expect that growing an Instagram following alone will increase school or collective pay. Performance had a clearer relationship with roster pay in this sample.

For athletes and advisers, follower growth remains relevant to brand deals. Roster pay should be assessed separately, using the athlete's position, production, and program.

For valuations, the results support including production and program scale. They provide less support for placing heavy weight on followers when estimating roster pay.

The data

112 athletes, each contributing four measurements.

What they’re paid

On3’s published NIL valuations. In July 2026 On3 changed these from estimating an athlete’s marketing potential to tracking what schools and collectives actually pay, which is what makes the figure usable here. Values run from $500K to $6.5M, median $2.0M.

Their audience

Instagram follower count, plus an engagement rate across roughly 30 recent posts. Engagement is a median, not an average: one post announcing a transfer can outperform an athlete’s normal output tenfold, and an average lets that single moment set their whole figure. Followers run from 78 to 750,336, median 26,126.

Their production

Where an athlete ranks against everyone else playing their position, from ESPN’s published season statistics. A quarterback is ranked against quarterbacks, a receiver against receivers.

Their program’s scale

How much money the school’s athletics department brings in, as a percentile against other programs, from the U.S. Department of Education’s Equity in Athletics Disclosure Act survey.

The model

The model tests whether audience is associated with pay after accounting for position, production, and program scale. These controls reduce differences caused by comparing athletes at different positions or in programs with different resources.

Pay is log‑transformed, so every result reads directly as a percentage change in pay. The four factors are put on a common scale so their sizes can be compared, and each is reported with the range the data cannot rule out.

What this study cannot tell you

Top of the market

The sample includes only athletes with publicly reported contract values. It does not represent the wider population of scholarship athletes, many of whom receive revenue-sharing payments without a published amount.

Some positions only

A comparable production statistic exists for quarterbacks, running backs, receivers, tight ends and basketball players. Linemen and most defensive positions have no equivalent, so the fully controlled model runs on 66 of the 112.

One platform

Audience here means Instagram audience. Other platforms were not measured, and an athlete whose following is concentrated elsewhere is understated.

No position-level numbers

Only quarterback carries enough athletes in this sample to support a figure of its own. Every other position would be noise, so none are reported.

Association, not causation

This measures what pay and audience do together in observed data. It does not establish that changing an athlete’s following would change their pay.

Sample size

The sample limits how precisely we can estimate the relationship between audience and pay. A larger study could detect an association that this one could not.

---