Since its release in 2011, the film “Moneyball” (based on Michael Lewis’s 2003 book “Moneyball: The Art of Winning an Unfair Game”) has been celebrated as an inspiration for the adoption of data-driven decision-making in the marketing space.
“Moneyball” chronicles the 2002 season of the Oakland A’s, in which the team won the most games in baseball, despite having the third-lowest payroll in the league. According to the film, the A’s were able to do that primarily by shunning traditional scouting (i.e., subjective human-led/eye-test evaluation of players by watching them play), leaning instead on an analytical approach that highlighted the players’ “advanced” stats, trying to find predictive patterns.
That’s an easily digestible narrative, but it’s not what happened. What separated the A’s at the beginning of the century was not that they were analyzing stats but the stats they were analyzing. While other teams focused on traditional statistics, the A’s identified metrics that were much better predictors of players’ performance and leveraged that to sign undervalued players to small contracts.
Identifying the right metrics is also a central theme in CRM/Retention marketing. This blog post will outline the futility of relying on traditional KPIs when trying to get an accurate picture of a marketing campaign’s performance and detail how you can modernize your analytical approach – and beat the market just like the A’s did.
Sometimes old-school is just old
Let’s face it: traditional marketing stats are dinosaurs. Too many marketing teams have glorified tracking email opens to the point that entire campaigns are evaluated based on it. This approach has two problems: 1. Apple’s and Android’s recent privacy updates blocked marketers’ access to email open rates and other old-school metrics; 2. Even if they were still available, they couldn’t provide an accurate picture of a campaign’s impact.
Successful marketing is about engagement, performance, and sales. Tracking only email stats is missing the forest for the trees. Not to mention how marketers who still rely mainly on email interaction data will have to remodel their CRM approach due to new privacy regulations. In other words, they’ll have to start focusing on scientific and actionable data metrics when evaluating the effectiveness of campaigns.
One crucial part of such methodology is shifting to a control group strategy.
A control group strategy requires setting up different recipient groups and tracking each group’s long-term behavior – while always having one small group of a similar audience not receiving any marketing messaging at all. Like A/B testing, the idea is to track campaign results based on the content presented to multiple groups, compare it to the control group and look for the uplift. And you’d better be measuring revenue-related metrics while you’re at it.
In a more long-term, bigger picture way, to make sure your strategy is heading in the right direction (even as per-campaign result may ab and flow), you should focus and expand on the following three KPIs:
- Customer-Base Coverage
- Number of Channels
- Number of Target Groups (“segments “)
And then again, the way to keep improving and optimizing your CRM Marketing strategy according to these KPIs and truly gauge the campaign’s performance is going through looking at figures, customer retention, and other truly meaningful, business-related KPIs.
The more micro-segments, the better
Now that we’ve covered the macro (shifting to a control group strategy and focusing on the right things), it’s time to discuss the micro; micro-segmentation, to be precise.
Micro-Segmentation, i.e., leveraging customer data to create as many groups as possible, stands at the heart of any intelligent retention strategy. The idea is not to group customers only by common purchase histories but to go deep and create micro-segments based on time and frequency of purchases, messages responded to, channels used, etc.
Cluster analysis should play a key role here. This method uses mathematical models to discover groups of similar customers based on the slightest variations. It’s a basic methodology used when Optimove’s onboarding team builds a bespoke customer model for a new client.
But, how many customer groups should you aim for? Well, the more, the better. And the right technology will help you scale it slowly but surely, while also making it possible to manage even hundreds of them.
An Optimove study found that companies with more segments in their retention campaigns earn significantly more than companies with fewer segments.
Granted, setting up dozens or hundreds of micro-segments takes more effort than sending a standard mass email campaign and waiting for click-through stats. But combining AI-based automation with human input and supervision will yield great results. It’s also the only way to achieve such scale – and the boost in Customer Lifetime Value that comes with it.
The next step in modernizing your analytics approach is experimentation. For example: using control and test groups as an ongoing experiment to determine what works best for which micro-segment.
CRM experimentation and comparison can provide profound tactical insights that sometimes border on prophecy. They can also help with the thorny issue of product pricing. In fact, the insights gained from experimentation configurations and methods can help with all the facets of your business.
A constant bird’s-eye view
While micro-segmentation and experimentation are vital in today’s retention marketing campaigns, they are meaningless without measuring and quantifying your test/control experiment methodology in finite terms.
The right actionable-CDP/multichannel marketing hub should provide complete marketing performance evaluation metrics, taking into account all data, campaigns, personalizations, and segmented groups. This constant bird’s-eye-view gives marketers the data-based confidence to tweak where necessary, learn from and maximize positive habits while eliminating negative ones.
A crucial parameter in this regard is statistical significance. Statistical significance indicates whether the campaign recipients’ behavior directly resulted from a specific campaign. This is one of the most critical insights marketers can hope to obtain. A fundamental “Moneyball” metric, if you will.
From daring to track business-related metrics through using control groups, scaling the number of your target groups (=segments/micro-segments), and constantly treating campaigns as experiments, adopting a science-first approach to Retention Marketing is proven to be the best way to boost customer lifetime value. But it all starts with getting into the right mindset. And moving away from tactics. Especially the ones people use “because it’s how we’ve always done it.”