Athlete evaluation 2026: how the scorecard I built at Wasserman would change today.
Two years on, three additions: AI-augmented sentiment, audience overlap against first-party brand data, and a contingency view for sponsor risk.
The framework I built in 2023 was designed to answer one question: does this athlete's story, audience, and character fit the Nationwide brand well enough to justify a long-term ambassador commitment? We scored candidates across four dimensions. On-field performance and projected longevity, off-field narrative and media persona, audience demographics against Nationwide's customer profile, and character risk. It worked. The framework helped sharpen a decision that became the Saquon Barkley "So Much More" signing.
Two years on, I'd make three additions.
AI-augmented sentiment from unstructured sources
In 2023, sentiment analysis meant social media monitoring and a sweep of press coverage. That's surface-level. A better version pulls from podcast transcripts (sports radio, athlete-owned shows, mainstream media appearances) and beat-writer archives, where the real texture of how an athlete is perceived in their market actually lives. Structured NLP models can surface patterns a human researcher misses at scale: how often is an athlete associated with negative storylines, what emotional register does their coverage fall in, what's the sentiment trajectory over the last 12 months versus the prior 12? The signal-to-noise ratio in earned media is higher than social. It's also harder to game.
Audience overlap against first-party brand data
The original framework used demographic and psychographic proxies to estimate audience fit. That's still the baseline when you're early in a search and screening a long list. But for a brand like Nationwide, one with millions of policy and financial-services customers, the right move at the finalist stage is direct audience-overlap analysis: match the athlete's follower set against the brand's own customer data via a data clean-room (LiveRamp, Habu, or similar) without exchanging raw PII. The question shifts from does their audience look like our customer? to how many of their followers already have a relationship with us, and how many are adjacent prospects we're paying to reach through other channels anyway? That second question changes the ROI math on the fee significantly.
A contingency view for sponsor risk
The original scorecard treated character risk as a binary flag. Something is either present or it isn't. That's not sophisticated enough for a Fortune 100 brand. A more useful version models three scenarios at the evaluation stage: clean execution over the contract term, a minor controversy (a social media incident, a soft press cycle), and a hard termination event. For each scenario, what are the exit provisions, what's the likely media radius, and what does contract length mean for exposure? Building that in at evaluation rather than at legal review isn't pessimistic. It's the discipline that earns buy-in from the finance and legal stakeholders who ultimately sign off. It also tells you which contract terms to fight for.
The underlying logic holds. Score across dimensions, weight against brand priorities, pressure-test character risk. The data layer around it has just gotten much richer, and the tools to work with that data are now accessible to practitioners who aren't data scientists.