CASE STUDY — SPORTS BROADCASTING

NEXUS knew the audience better than the existing forecast

A Fortune 500 media company already had an established system for forecasting sports audiences. NEXUS reduced its prediction error by 33%.

PREDICTION ERROR

33% less

COMPARISON

Incumbent baseline

GROUND TRUTH

Aired audience

GROUND TRUTH

Aired audience

01 — THE CHALLENGE

Decisions made before the audience arrives

Broadcasters make enormous commercial decisions about audiences that haven’t arrived yet. Advertising inventory is priced and sold months in advance, based on forecasts of how many people will eventually watch. Get those forecasts wrong and the consequences are real: overestimate demand and you can create make-good exposure; underestimate it and valuable inventory may be left underpriced.

02 — THE APPROACH

A forecast with nowhere to hide

What makes this case particularly compelling is that there was already a strong incumbent to beat. The company had an established forecasting process and a locked pre-season baseline. NEXUS independently predicted the same broadcast audiences, and as the season unfolded, there was nowhere for either forecast to hide: both could be compared directly with reality.

Incumbent

Forecast locked before the season

NEXUS

Forecast independently generated

Ground truth

Actual audiences as broadcasts aired

03 — THE RESULT

33%

less prediction error than the existing forecasting baseline

As actual audiences came in, NEXUS tracked them substantially more accurately than the company’s incumbent forecast. The chart makes this particularly tangible: the locked baseline, NEXUS forecast and actual aired audience can be seen diverging and converging as the season progresses.

04 — THE IMPACT

The money has already moved.

For a broadcaster, better audience prediction flows directly into the economics of the business. It creates better information for pricing inventory, making advertiser commitments, spotting where audiences may be over or underestimated and managing the risk between what gets sold and what ultimately gets delivered.

The audience hasn’t arrived yet. The money has already moved. That’s why predicting it better matters.

Related Cases

World Cup

NEXUS called the World Cup before it began.

NFL Player Value

NEXUS predicted what NFL players would be worth before the market did

Formula 1

NEXUS predicted the behavior of an F1 car before it hit the track

Football Club

One of Europe's biggest football clubs used NEXUS to answer three questions: who's about to leave, who can still be won back, and which players are worth more than the market thinks.

SMB Lending Fintech

A small-business lender needed to size up default risk on brand-new applicants — before they'd ever connected a bank account. NEXUS ranked that risk more accurately than the model already in production.

Credit Underwriting

For repeat cash-advance customers, a small-business lender needed to know who was likely to default on their next draw. NEXUS out-predicted the incumbent model — with a fraction of the engineering behind it.

Energy Demand

Forecasting demand across wind, solar and total load to guide real-time trading decisions — sourcing, timing and pricing. NEXUS beat the incumbent model by 71%.