A revenue analytics platform for a hockey club
Most of my background is in banking, not sports. This suite exists to answer a fair question: without a sports background, can I still do the modeling a club’s revenue team needs? The questions are real ones: will season-ticket holders come back, what should each game cost, what are partnerships worth, which fans are worth finding. Five models are built and interactive for a hypothetical Buffalo franchise; the rest are specified on the roadmap.
#55
Nielsen DMA
637,090 TV homes
16,421
Avg attendance, 2025-26
~86% fill · 31st of 32
19,070
Arena capacity
playoff dates sell out
~38k
Telecast audience
4.0 household rating
1st
US local NHL ratings
highest of any US market
The models are calibrated to the same public market data. The figures that matter most sit at the ends of the row: the country’s best local hockey ratings out of the 55th-ranked media market, against a building near the bottom of the league in fill. The audience is there; much of it watches from home. Most of these models end up working that gap.
Ticketing & attendance
Season-ticket renewal risk
Live85.0%
baseline renewal across a 2,400-account book
Scores every account, ranks the call list, and sizes the retention program.
Open the model
Game demand & variable pricing
Live+$3.87M
per season vs. a flat $85 ticket
Demand curves by opponent tier and day of week; prices the slate for gate plus per-caps.
Open the model
Predictive buying & lead scoring
RoadmapRanks the CRM pipeline by conversion likelihood: single-game buyers to plans to memberships.
Cross-sell & upgrades
RoadmapNext-best-offer for existing accounts: plan upgrades, lower-bowl moves, playoff priority.
No-show forecasting
RoadmapPredicts scan rate per game for staffing and as an early churn signal.
Fan & marketing
Fan identity resolution
Live2,375 → 886
raw records resolved to fans
Matches one fan across ticketing, retail, concessions, app, and email, with the precision-recall tradeoff measured against known truth.
Open the model
Fan lifetime value
Live$48.8M
ten-year fan equity across five segments
Discounted value per fan by segment, with the playoff cohort's year-two cliff priced and a save-program lever to close it.
Open the model
Segmentation & broadcast conversion
RoadmapBehavioral segments, including the region's large TV-only audience and what converts them into attendance.
Marketing ROI
RoadmapChannel-level return on marketing spend, with team performance controlled for.
Corporate partnerships
Sponsorship valuation
Live1.12×
delivered value vs. rate card, default package
Prices 30 partnership assets by media equivalency, structures the term, and suggests the ask.
Open the model
Sponsor renewal risk
RoadmapDelivery against contract, engagement, and category health for every partner account.
Category whitespace
RoadmapUnsold and underpriced sponsor categories against comparable markets.
Executive
Revenue scenario model
RoadmapTeam-performance scenarios carried through attendance, ratings, per-caps, renewals, and sponsorship into a P&L range.
Live-model figures are each model’s output at its default settings.
Why this order
These are hypotheses about sequencing from outside the building, not a prescription; a team living with the real data would know where they break. My read: coming off a losing stretch, renewal risk leads, because the season-ticket base is the largest forecastable revenue line and the data already exists. Coming off a breakthrough season, the order flips. Demand is doing the retention work, so lead conversion and pricing jump the queue, and the renewal model’s job becomes preparing for the year-two cliff, when new cohorts churn hardest. Sponsorship valuation seems useful either way, putting a checkable number beside the rate card.
The roadmap items wait on foundations: identity resolution before lifetime value, holdout experiments before uplift models, clean spend data before marketing ROI. And each model is built to end at a decision a revenue team could act on: a call list, a price sheet, an ask.
The club and its books are simulated; market inputs are public data. Not affiliated with any professional team.