Predicting your customers’ future behavior through analytics

So, we’re recently back from the 2024 DTC Wine Symposium. We learned a bit, drank more than a bit of delicious wine, and met some engaging and industry-knowledgeable professionals. Bin 113 is comprised of marketing folks who hail from other DTC industries and are accustomed to much larger shows and service-oriented pools (e.g., Direct Marketing Association, Interactive Television Alliance, Electronic Retailing Association, etc.), so the intimacy of this event was truly refreshing.

While much of what was discussed was relatively old hat for us, many wineries — even the largest — were excited to learn about certain foundational principles that exist (and have existed for a long time) in the world of DTC. What became clear is that wineries need to use some level of data and analytics to acquire new club members and DTC customers, increase customer satisfaction, energize their most loyal supporters and — most importantly — to increase Lifetime Value. Customer satisfaction should — for any brand — be an ongoing and important process.

One popular and well-used tool used by every direct marketer is house file segmentation and scoring. This helps segment customers into clusters by examining recency, frequency and monetary value (RFM). Using past purchase behavior, it can help greatly enhance and lift your marketing response rates.

RFM can deliver the most relevant messages to targeted groups – including those at risk of lapsing. These customers are “scored” based on the recency, frequency and size of their purchases.

RFM scoring is more than 30 years old and was first used to verify what most DTC marketers intuitively understood: that 80% of sales come from just 20% of your customer base. The trick is to determine who that 20% is and to be certain that they are treated accordingly.

Although RFM is a lovely little tool, it portends to be predictive and relies on limited data to do so, which is its flaw. And so, modern, scaled, and sophisticated DTC marketers rely much more heavily on predictive analytics (RFM is categorized as descriptive analytics) and AI-powered modeling using significantly more variables than the three that RFM deploys to identify high-propensity customers. RFM is not necessarily antiquated, it’s just that more effective tools are now available. For wineries who are very new to the game, RFM is a perfectly rational segmentation tool.

There are a variety of RFM platforms to choose from — we actually saw a perfectly good one at the show — but what about an easier, simpler method of determining your high propensity (or dissatisfied!) customers; one that might not be nearly as sophisticated as RFM but may just do the trick in a pinch? Well, get ready for another acronym:

NPS, or Net Promoter Score.

Think about it, what better way is there to determine your most loyal and satisfied customers (what RFM basically does) than through a simple survey? NPS basically uses survey data to determine how your customers rate (from 1 – 10) their satisfaction. The buckets are then cut into those with high scores (“Promoters”), those with mid scores (“Passives”) and those with low scores (“Detractors”). As you might expect, there is a bit more sophistication to NPS than we’re letting on (yes, there is actually a way to calculate your NPS score), but you get the idea. Obviously, your Promoters are your biggest fans and the best candidates for upsells, cross-sells, special promotions, etc. and your Detractors, well you better start showing a bit of love.

To be frank, many DTC marketers will use both RFM together with NPS as there are qualitative differences between the two and, really, customer satisfaction should—for any brand—be an ongoing and important process. The neat thing about NPS is that it will also give you overall customer satisfaction scores across your entire customer list that can be benchmarked year over year.

In the end, Wineries need to use some level of data and analytics to increase customer satisfaction, energize their most loyal supports and—most importantly—to increase Lifetime Value.

By Jeff Giordano