Data has quickly become the most powerful tool in today’s marketing world. As social media and other digital platforms proliferate, purchase intent analytics help marketers target consumers and contextualize offers in real time. We check in with Jeff Rosenfel, VP Customer Insight and Analytics, The Neiman Marcus Group who reveals how retail advertising and marketing are changing how they allocate media investments and generate new creative concepts informed by data.
Gwen: We’ve talked about how data has become the most powerful tool in today’s marketing world. Jeff, can you give us some context on how you’ve changed in your approach to targeting customers?
Jeff: Neiman Marcus had a catalog business that started about 50 years ago. The reason that I bring that up is because to run a catalog business you have to be paying attention to segment-level detail, and analyzing it to improve. I think that culture of analytics positioned us well for when the website started around 1999, as data really began to grow very rapidly in a much more complex way. Fortunately, numerous technologies were advancing at the same time that we leveraged to help us analyze that data and act on it. Those technologies, both in-house and vendor partners, are enabling us to use things like artificial intelligence to iteratively learn how customers respond to further personalize. That journey has really led us from high-level segments years and years ago to today, where on a weekly basis there’s over millions of messages that are going out to literally just one person at a time.
Gwen: Building on that, we have terms today we’re calling “predictive, prescriptive, selling.” How are your teams approaching this?
Jeff: I think the better we can predict what the customer’s needs are, the better we can better fundamentally serve what she’s trying to achieve. I think we all know from experience just navigating and trying to get done what you want on a mobile device is much more challenging than doing so on a desktop. And so we ask ourselves, “How do we solve that for the customer by leveraging data and things she’s told us through her behavior?” One of the things we’ve recently implemented in our emails is a feature we call “Quick Links,” where we predict the likely places a customer will try to navigate to and put those personalized links directly in her email so she can go straight there and doesn’t need to navigate extensively on her phone to find what she wants. At the end of the day, the goal is simply trying to make it easier for the customer.
Gwen: Building on the relationship theme, are you seeing a shift from the traditional media segmentation to more relationship-driven strategies? I think of relationship-driven programs as a hallmark of The Neiman Marcus Group.
Jeff: In my mind, that relationship-driven strategy is almost the next frontier. Where we say how do we not just manage for what we think she’s going to do now even if we can customize that message, but really what’s the optimal mix of things to send her over a length of time. I think there are not a lot of folks out there that I’m aware of who are doing that very well. We’re starting down that journey. I think a number of other companies are as well, and it’s really exciting. As some of the technologies develop, and the ability to optimize them becomes very complex. I’m really excited to see what comes in the area of capturing the customer journey.
Gwen: Do you categorize types of customers in “behavioral buckets” rather than each and every behavior of an individual?
Jeff: We do have some high-level segments still that are used to inform reporting and strategy-type decisions. At the end of the day, I think the focus is really trying to get as close to personalizing that communication, then, eventually in this relationship-driven piece, personalizing the relationship. Some of that even involves, again, having our sales associates as part of the mix. Our sales associates have apps, and we can send them information to help them expand their reach. A human can only remember so much. If we can supplement that with data on the customers they don’t interact with as much, that can also help enhance that relationship.
Gwen: The speed has probably increased in terms of how much data you have.
Jeff: When it comes to ROI, especially in media, we look at how what is each dollar of spend driving to our business. We started on that journey probably about four years ago or so. Really as we’ve done that, we have made fairly sizable shifts in some of our media allocation between different channels. What I’ve found exciting is over that journey we’ve managed to drill down to specific channels and for some channels down to specific partners. I think as we’ve progressed on that journey one thing that we found useful is to start to be able to look at optimizing multiple different metrics. There are times where you’re going for ROI, there are times you are going for sales, there may be times when you’re going for new customers trying to optimize the longer-term value. I think that’s one of the important things to take a look at. How do we align the analytics to match the strategy and be able to support multiple objectives?
Gwen: Can you talk a little bit about the difference of this media-mix modeling versus some attribution modeling that’s being talked about today?
Jeff: When I think of media-mix modeling, I think of it as a very high-level approach looking at spend buckets. It is especially good at managing the mix of digital and traditional media, pretty effective at very high-level budget planning. On attribution, there are multiple types. There is multi-touch, last click, and first click. We use multi-touch, which I think is best for most use cases. In our case, we work with a vendor partner who helps us but is essentially algorithmically allocating credit at a customer level across the various media channels.
Gwen: I wanted to ask you to comment on how you see your organization, in particular your marketing department changing when you have to take in all of these new points of information. Have all of these new tools, have all of these new analytics. How do you see your staff changing? How do you see your capabilities evolving? What does that look like and feel like to you?
Jeff: The rate of change is absolutely amazing. I started at Neiman Marcus a little over 12 years ago in the analytics team, and at that point there were about 1 or 2 toolsets that we used. Now, the team is managing 10–15 toolsets and needs to be learning new algorithms and new approaches almost on a monthly basis. It’s a real challenge to keep up, and I think it forces us to have to really prioritize and identify what the things are that we want to develop ourselves versus what the things are that we want to partner with the best-of-breed vendors out there to really help escalate our roadmap.
Gwen: How are you looking at data to fuel personalization?
Jeff: Before I address that about the data piece, I like to think about just the personalization piece. We have great sales associates in our stores, and they’re a great analogy of personalization. A good sales associate is going to personalize the experience. They’re going to observe what the customer is doing. They’re going to listen to what they are saying, see their facial reactions, and listen to their preferences. Then they’re going to act on that information by recommending products or a different item that might fit differently. And to really benefit from the relationship over time, they’re going to remember information about that customer. So now, if we translate that to your question on the data piece and go online, we can go through those steps and ask, “How do we observe online?” We observe through things like views and clicks, how much time they’re spending, how they’re interacting with marketing. But that’s just data collection. The action is a little bit harder, because that’s a basic heuristic, like if this is a “new customer,” I’m going to do X, or this is a “best customer,” I’m going to do Y. But it’s here where we see the strength of the algorithms. I think this is what I’m most excited about—seeing the advances in the technology and how we’ll combine the rich behavioral and purchase data with the style of a great associate to really “wow” the customer and make her life easier.
Gwen: Thank you for this insight. It’s worth pointing out that while data becomes the new oil for retailers, consumers are already using algorithims as they navigate their day. Apps like Wallet.AI and SPIRE allow them to be financially responsible and health-aware. Retailers will certainly shift marketing efforts as they sell to preferences rather than demographic groups and post-purchase experience rises in priority. We’re looking ahead to more innovations in technologies, changes in media consumption, and more predictive algorithims. Hold on to your hats!