On Marketing Efficiency, Immediate ROI, and the Hype of AI

For most DTC brands, the majority of marketing campaigns are minimally to moderately risky.  But then there’s the New York City transit ad that’s going to cost millions, and it’s your brands’ shot at scaling. No matter which camp your next campaign falls into, no one wants to waste money by choosing the wrong message or advertising to the wrong audiences. 

Last week, our CEO sat down with analytics and insights expert MaryBeth Maskovas to talk about how e-commerce brands can get the most (and the fastest) value from their data projects. One of the things they agree on most emphatically is that clean, unified data can result in major cost savings for brands, and is the first step for more effective and efficient marketing. 

MaryBeth Maskovas, CEO of Insight Lime Analytics (MB): From the analytics perspective, starting any data project right off the bat with clean data makes things a lot easier and faster. From your experience, after the data is in order, what are some of the projects analysts can do to drive immediate value, rather than waiting six months to see the return?

Daniel Brady, CEO of Orita (DB): Right off the bat, we know that if a brand is doing any sort of email marketing, they’re probably spending 15-20% more than they need to. As a brand grows, its email lists accumulate junk. The same person is represented 4, 5 or 6 times in your data, and you have people on your list who have never opened an email since they were added three years ago. By creating a unified history of all purchases from every platform over time, it becomes possible to prioritize the best customers on your list and stop marketing to the worst. 

Obviously this would be an arduous process to tackle manually, with a ton of room for error, which is why we use machine-learning to do the heavy lifting. With models that look at every email address, plus the open rates, link clicks, and a host of other metrics, it’s possible to get rid of accounts that are un-engaged. Coupled with removing duplicate accounts, and your marketing bill just decreased by hundreds or thousands of dollars a month. 

MB: One of the things we work on quite a bit is churn prediction. With clean data, we can be really confident in the customer lists and customer history, and we can predict churn effectively. Another thing that can have a quick ROI, and relies heavily on a good dataset, is cluster analysis. A lot of the time growing brands rely on customer personas for marketing, but when you really break it down, there might be two very different core groups driving revenue. There might be unique gift giving behavior. Understanding the real customer behavior can lead to a more effective activation strategy, and it can lead to better messaging for that major NYC transit ad. A lot of the time our customers are just starting to make the leap from digital to TV advertising, which is a big increase in cost. Relying on messy, aggregate data here isn’t a solid strategy. It’s really important to have a deep understanding of a brands’ actual customers before making these moves. 

DB: Absolutely. We’re at this point where marketers don't have to deal with the average anymore. There’s not one set of behaviors that fits everyone, so why market as if there is. One of the best uses of clean, unified data is figuring out who your best customers are. Understanding their behavior, and encouraging it. But also understanding your worst customers, and restricting the behavior that encourages those customers. 

MB: This is one of the areas that I truly believe AI is worth all the hype, which most of the time I don’t. Taking a few days to get your data cleaned and unified means a brand can be looking at an accurate cluster analysis in less than a month. Or can be understanding marketing efficiency and cutting costs in just a week or two. 

To see the full conversation, watch the on-demand webinar here. Or sign-up today to save money on your marketing automation platform before your next invoice.