I started with Super Bowl viewership (around 100M) and worked down through layers: who has a phone out, who's actually paying attention during the ad, who bothers to scan, and who completes the promo.
Break the problem into a funnel: total Super Bowl viewers → those who noticed the ad → those who scanned the QR code → those who signed up → those who used the promo. Use publicly available data (e.g., Nielsen ratings, social media buzz) and reasonable assumptions to estimate each stage, then multiply through to get a final number.
Pro tip: Anchor your estimate with a known benchmark (e.g., typical QR code scan rates for TV ads are 0.1-1%) and then adjust for Coinbase's unique context (e.g., tech-savvy audience, controversial ad). This shows you can combine external data with product-specific insights.
Use Nielsen data or public reports to estimate the average live audience (e.g., ~100 million viewers in the US). Consider that not all viewers watch the ads (e.g., some skip or are not paying attention).
Assume a certain percentage of viewers actually saw and remembered the ad. For a high-profile Super Bowl ad, recall can be 30-50%. Then, estimate what fraction noticed the QR code (e.g., 50% of those who saw the ad).
Use industry benchmarks for QR code scans from TV ads (often 0.1-1% of viewers). Adjust based on factors like ad creativity, call-to-action strength, and audience demographics (Coinbase's target audience may be more likely to scan).
Not everyone who scans will complete the sign-up and use the promo. Assume a conversion rate (e.g., 20-50% for sign-up, then 50-80% for promo usage). Multiply through to get final number.
Compare your final number to any publicly reported results (e.g., Coinbase reported 20 million hits in one minute, but that's website visits, not unique users). Adjust assumptions if needed and present a range.
AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.