The First Year Post-Ari

By the time of the Better Collective acquisition in mid-2021, I was Director of Analytics at Action Network, leading a four-person analytics team and a three-person data engineering team across web, app, subscriptions, and affiliate revenue streams. The team covered every major vertical of the business: product analytics, affiliate partner reporting, financial modeling, and the data infrastructure that connected them.

The first months after the acquisition were a trial by fire. Much of our time was spent trying to understand our new European counterparts and integrate analytics processes. Communication challenges were formidable, and without a strong unifying leadership figure, we often found ourselves treading water. The struggle to unify metric definitions alone was surprisingly contentious, and we missed quick wins that should have come from an immediate synthesis of our data.

The most important lesson from this period: strong, aligned leadership is the prerequisite for effective cross-functional collaboration. We had a lot of smart people and no alignment.

What shifted things was a meaningful organizational focus change: from audience metrics to user acquisition. I was deeply involved in facilitating that shift, taking charge of new reporting and analysis to make it real. The result was an intricate user acquisition dashboard that tracked our conversion efforts in real time and guided future strategy.

Consumer Product Analytics Infrastructure

Before I could influence product direction, I needed to understand how users actually moved through the product. I built end-to-end tracking across the web platform and our native iOS and Android apps, connecting event-level user behavior to downstream revenue outcomes.

The Action Network had two distinct revenue streams. The first was subscriptions: monthly and annual plans that gave users access to premium content and picks. These generated predictable recurring revenue, and I modeled LTV using cohort-based churn rates. The second, and far more valuable, was affiliate clicks: when a user clicked a link out to a sportsbook, we earned a placement fee of $150 to $300 per activation. Because a single user could activate on multiple sportsbooks over time, their total affiliate LTV could approach $1,000. That asymmetry shaped everything about how we thought about the product.

My team built the model that connected in-product behavior to predicted LTV: which actions correlated with eventual subscription, which correlated with affiliate activation, and which predicted nothing. What we found was that the highest-LTV funnel was specific: get a web user to download the app, then get them engaging with editorial content (either from our writers or sportsbook-sponsored pieces), and they were significantly more likely to click an affiliate link within the same or following session. Web users who never downloaded the app had much lower conversion rates regardless of how much content they consumed.

This analysis directly informed product roadmap priorities. We were no longer building features for engagement's sake. Every content feature and onboarding decision was evaluated against its predicted impact on the funnel we had mapped.

Built with: SQL, Redshift, Segment, Google Analytics, Google Data Studio, Tableau, Python, SciKit Learn

The Value Clicks Project

As we shifted from focusing on Monthly Active Users to prioritizing revenue, nearly 40% of the company experienced a significant role change. This hit the content team hardest. Chad, a genuine savant in the sports media space, was completely supportive of the shift, but faced a motivational challenge. His team had previously had a clear view of their direct impact through article readership. With the new focus on conversion further down the funnel, it became difficult for them to see how their work connected to the bottom line.

In collaboration with Chad and his team leads (Katie, Steve, and Andrew), we developed the concept of a "Value Click": examining the subsequent action a user took after reading an article, and asking whether that action was predictive of a future subscription or a sportsbook promotional click. The metric was split by platform and session type, giving the team a granular view of their actual impact on revenue potential.

This became the company's north star metric. Both the content side and the product side aligned their roadmaps around it. We instrumented it in Tableau and Google Data Studio so it was visible to the whole organization on a regular cadence. For me it was the first project of this scope that I ideated and led end-to-end: I had to listen carefully to a large team's needs and deliver something that actually changed how they worked. We also discovered correlated metrics along the way ("articles read per reader" and "cross-sport readers") that had substantial downstream revenue impacts.

Built with: SQL, Redshift, Google Sheets, Google Apps Script, Google Analytics, Segment, Python, SciKit Learn

Better Collective Data Sharing

The teams at Better Collective in Denmark naturally approached me to gain access to our data. We didn't have a system fully ready for this, so I automated daily reporting to feed into their financial stack as a starting point. This early encounter with Better Collective gave me my first real taste of the contrast between a startup's "move fast" mentality and the more structured, deadline-driven approach of a scaled European media company. Both have value. They don't always fit together easily.

Built with: SQL, Redshift, AWS, Google Sheets, Google Apps Script

C-Suite Reporting Infrastructure

As Director of Analytics I owned the reporting layer that went to executive leadership. On a daily basis my team produced dashboards measuring performance against forecast across all revenue streams: affiliate activations by partner and state, subscription conversion rates, and content metrics tied to the Value Clicks north star. My role was to aggregate, synthesize, and present: turning the team's daily output into the weekly executive summary that went to the CEO, CFO, and leadership team.

I built the infrastructure that made this possible: a set of Redshift tables and Google Sheets models that standardized how each revenue stream rolled up to a single business view, with consistent period-over-period comparisons and automated variance explanations. That infrastructure meant I did not have to rebuild the reporting package from scratch each week. It also meant that when the Better Collective finance team needed data, the format was already reconciled with how they reported.

The discipline of owning a C-suite reporting cadence at the Director level taught me what actually matters in executive communication: one clear number, one clear trend, and one clear decision. The data behind it should be ready when they ask, not in the room when they are not.

Built with: SQL, Redshift, Google Sheets, Google Data Studio, Python

Affiliate Performance Reporting

Our acquisition by Better Collective was driven largely by strong affiliate revenue growth over the prior year. As that growth continued, it became clear that our affiliate performance tracking wouldn't scale. April and Amanda (hired at April's advocacy) played a pivotal role in building infrastructure to automate scraping and cleaning data from nearly two dozen affiliate partners. My role was leadership and quality assurance: making sure the reporting schemas met both Action Network and Better Collective requirements, navigating the competing priorities of multiple stakeholders, and trusting the team to deliver.

This was an important moment in my development as a manager. I learned what it actually means to trust a team, not just say you trust them, but step back and let them build.

Built with: SQL, Redshift, Python, CRON, Google Sheets, Google Apps Script, Google Data Studio