On NBA Opening Day, A Marketer's Salute to Advanced Analytics
The NBA season starts tonight. I’ve been an avid fan of the game for years and years, and my favorite players today aren’t who they would have been when I started watching.
Why? It’s not that the game has changed; it’s my appreciation for the way we measure impact in the game that’s changed.
When I was a kid learning the game, I loved volume stats: points, rebounds, and assists. To me, those were the ultimate indicators of performance; there was no consideration of how those stats played out on a team level.
When I was just cutting my teeth in digital marketing years ago, my approach was about as nuanced. If I couldn’t directly attribute conversions or revenue to investments, I didn’t make those investments. This meant that, back in the day, I was hesitant to recommend platforms like YouTube or display – essentially, any options in the dreaded “brand awareness” bucket that didn’t directly lead to down-funnel metrics.
As the NBA season tips off, I’ll be closely watching a lot of players who won’t show up on statistical leaderboards – after a day spent helping my clients achieve great performance using all kinds of campaigns, including brand awareness.
So what am I hoping you’ll learn from the evolution of my NBA fandom? Let’s start with a look at one of the “best rebounders” in league history.
DeAndre Jordan, Rebound Gobbler
From 2013 to 2018, DeAndre Jordan racked up thousands of rebounds, standing out as one of the most statistically impressive players over any consecutive five-year period. Throughout this span, he led the league in rebounds twice and never placed lower than second. The following graph illustrates his year-over-year rebound per game averages:
If the General Manager of an NBA team had a specific goal of building a strong rebounding roster, signing DeAndre Jordan would be a logical choice. His remarkable output and consistent performance place him among the rebounding elite.
In comparison, Steven Adams, while a respectable rebounder, does not surpass the rebounding prowess of DeAndre Jordan at first glance. Over the last decade, Adams has never ranked higher than eighth in rebounds per game – solid but not as obviously outstanding as DeAndre Jordan.
A deeper analysis of Steven Adams’s track record shows that his teams have consistently excelled in the realm of rebounding: a top-five ranking in rebounds per game for seven out of those nine years, a feat that transcends multiple franchises. This consistency demonstrates Adams' undeniable impact on a team's rebounding prowess. The table below displays Adams’s rebounds per game and his teams’ overall rebounding ranking:
Contrast that to DeAndre Jordan – whose teams, despite his superior individual performance, were consistently in the bottom half of the league:
The contrast in team rebounding rankings between DeAndre Jordan and Steven Adams stems from their differing playing styles. DeAndre's pursuit of personal rebounds sometimes hinders team performance, while Adams's focus on facilitating his team's rebounding efforts (by boxing out and helping position his teammates for success) boosts their overall rankings. DeAndre prioritizes individual stats, while Adams prioritizes team success.
The marketing measurement parallel
In the realm of marketing, it's not uncommon to encounter situations where a campaign seems to perform well within a specific platform when using last-click attribution – yet doesn’t translate into a significant improvement in overall business performance.
This often occurs with remarketing campaigns. While these campaigns may claim credit for a significant number of conversions, a closer examination often reveals that many of those conversions weren't truly incremental. In other words, they might have happened without the remarketing campaigns, so the campaign conversions didn't substantially contribute to the business's bottom line or overall success.
In such cases, it's crucial to move beyond surface-level metrics and last-click attribution to focus on more comprehensive performance indicators that align with the specific goals of the business. Some marketing activities may create the appearance of success within the platform but may not genuinely impact the business's key metrics, such as revenue, profit, or customer lifetime value. This underscores the importance of measuring and optimizing for the platform-agnostic metrics that are most relevant to your business objectives, rather than relying solely on data from individual platforms or channels.
In my experience, remarketing and branded search campaigns often resemble the DeAndre Jordan of digital marketing, as they tend to "cannibalize" conversions and engagement from other touchpoints. These campaigns may dominate in last-touch attribution but not necessarily contribute to incremental growth.
Conversely, top-of-funnel marketing efforts often act like the Steven Adams of digital marketing. They focus on persuading and educating customers, akin to Adams's emphasis on "boxing out" opponents and letting other teammates swoop in for the rebound. These campaigns work collectively to create space for other touchpoints and contribute to a more balanced and comprehensive approach to customer engagement. They prioritize long-term growth and customer nurturing over immediate conversions.
Advanced analytics in action
At a publicly traded tech company, we initiated testing for account-based targeting with a top-of-funnel offer. While we did anticipate a higher Cost Per Lead (CPL) compared to our reliable and ongoing lead generation efforts, it turned out to be four times higher than our typical CPL. This presented challenges when making budget decisions.
As a last-ditch effort to gain a deeper understanding of ABM's real impact, we conducted a test using Google's Causal Impact framework. What we discovered was that the actual CPL for our ABM initiatives was significantly lower than what we had initially reported. We found that users that were exposed to this media later converted through different channels (Brand, organic, direct) at a far greater rate than users who were not exposed to the media. In fact, we found that we reported on only 10% of the actual impact of this campaign. The table below tells the story:
In another example of the power of digging into advanced analytics, we conducted a YouTube prospecting campaign for a Direct-to-Consumer (DTC) client. Again, we used Google's Causal Impact theory to measure its effectiveness, and the results were quite promising. In the regions where the YouTube ads were shown, we had initially expected 2,390 orders. However, the campaign generated 2,570 orders with a total ad spend of $11,000. This translated to a Cost Per Acquisition (CPA) of $88, which was significantly below our target CPA. In essence, the campaign not only exceeded order expectations but also proved to be cost-effective for our client. This is a Steven Adams-y campaign.
Conclusion
Just as Steven Adams is an unsung rebounding wizard, upper-funnel campaigns and ad types can carry hidden impact you must look deeper to uncover. In the NBA, statistical models have advanced to help GMs recognize non-splashy players who take over the game in subtle ways. In marketing, folks who truly want to drive business-level impact should create and operate within a framework that reveals true incrementality. Won-loss records – on the court and in the business arena – will be better for it.
