The majority of catalogers are mailing too many catalogs to web customers. -- Kevin Hillstrom, MineThatDataKevin Hillstrom is one of the leading thinkers in the area of multichannel marketing analysis. Here's a transcript and a podcast of an enjoyable conversation we shared last week. Listen to podcast: RKGblog_Interview_Kevin_Hillstrom.mp3
Alan Rimm-Kaufman:It’s a tremendous pleasure to be speaking today with Kevin Hillstrom, president of MineThatData. Kevin is a virtuoso of database marketing, with deep work experience at Nordstrom, Eddie Bauer and Lands’ End. Kevin, could you start off by telling us about your background and what you’re doing today? Kevin:Sure. I started off as a lowly statistician. I got a statistics degree from the University of Wisconsin. I spent a couple of years in my first job analyzing corn and sorghum hybrids for a little company out in Iowa, and learned basically how to do SAS and SPSS programming in that job. That helped me transition into a job at Lands’ End, where I was a statistical analyst, and I figured out who should receive catalogs for Lands’ End. That progressed into a job at Eddie Bauer, where I was the director of circulation. That became a job at Nordstrom, where I was vice president of database marketing, and spent a lot of time at Nordstrom understanding how channels interacted with each other. That kind of gave me the impetus to want to start my own business, focusing on helping people understand how channels fit together. That’s how I arrived here at what’s now called MineThatData, my own business where I basically help CEO’s figure out how customers are interacting with all the different channels. Alan:How customers interact with all the different channels -- that is the billion-dollar question in our industry. Collectively, we are all figuring out the relationship between catalog and web and store. Kevin:Yeah. It’s the big question right now. Alan:You came out with a tremendous book. It’s bright yellow, and has a beautiful flower or shell on the cover. It’s called, Multichannel Forensics. I very much enjoyed reading it. In your book -- jump in and correct me if I’m getting this wrong -- you take a corporation and break it up into business units. Web, store, catalog, and so on.You look at the annual retention rate of each business unit. And you bucket those rates into low, medium and high. Then you look at the migration between the business units. You suggest four different migration states. So there are three different retention states, and four migration states. In kind of an RFM kind of way, you get 3 x 4 = 12 unique combinations of those. The real beauty of the book is you say that once you know which cell each channel falls, you know how to manage that business unit. Did I get that right? Can dropping each business unit into 1 of 12 bins provide that much insight? Kevin:I think you are on target with that. Basically, I make the premise that as a business, you’re gonna keep either a lot of your customers from last year, or you’re not gonna keep a lot of customers, and that’s generally not good or bad. For instance, a lot of multichannel retail businesses will keep 75 to 80 percent of last year’s customers this year. A lot of catalogers and online businesses might keep 40, or 35 or 30 percent of last year’s customers. Knowing that, first of all, is really important, especially if you’re an online business or a cataloger, because it tells you what kind of customer acquisition activity you have to do. For a catalog or online business, you find more often than not that customer acquisition plays a disproportionate role in driving the future and health of that business. Then, there are migration states, that you mentioned, that I like to look at, as well. This is what’s really important, is want to understand if customers are willing to switch between channels, or if they’re just going to stay focused on one channel and that’s all they’re going to do. Basically, I look at people who are stuck, and they’re never going to move in channels, and I call that Isolation. I have customers who might be willing to try different channels. They’ll dip their toe in and try buying something on the Web, or might go into a store, or a Web customer might try to buy something out of a catalog. That’s called Equilibrium. Then, the most interesting dynamic of all is called Transfer, and that’s when customers switch. They go from one channel to another, and then they stay in that channel in the future. In some instances, you have customers that switch back and forth. I call that Oscillation. You basically end up, like you described, with an RFM scheme, with 12 different cells. If you’re an executive, knowing these 12 different cells tells you what you do with your business. I’ll give you an example. In our time at Nordstrom, we realized that our Web site wasn’t really great at holding onto customers. It was great at acquiring customers, especially in new markets or places where we didn’t have stores. What we found was those customers over time switched their loyalty to our store. Once they bought in the store, that’s where they stayed and they didn’t go back to the Web site anymore. Within that framework, you look at your Web site very differently, and you say, “My Web site is a place to get new customers, and then I want to make that experience great for them so that they can have a good in-store shopping experience.” It changed how we look at the business. You don’t focus anymore as much on conversion rates and things like that, the traditional metrics, because you want the customer to migrate through that process. It then gives the Web site a role within the business, and you realize that without the Web site, we knew that our sales would decrease by a certain percent each year. It basically gave that channel a role. It also told us, because our retail customers wouldn’t convert back to the Web site very easily, that we really couldn’t do a lot to basically get a store customer to convert back to the Web, so we didn’t try as hard at that because it wasn’t in the customer’s natural inclination to do so. At Nordstrom, it helped us figure out what that process would look like. For catalogers, we have a very similar kind of situation now, where customers receive a catalog but they order something on the Web. We need to figure out do we mail those customers more in the future, or do we not mail those customers in the future? This framework can help you understand what role a catalog plays at growing a long-term, healthy online business. Alan:You’re on the road a lot, helping different companies and getting to seeing a lot of different companies' results across the industry. Today, how far off-base is it to just treat the web, catalog, and store channels as independent? Kevin:There are some results that are kind of mixed. I break things up into more channels than the client thinks they have. Alan:I How so? Kevin:For instance, a client of mine might have customers who receive a catalog and send their order in through the mail, like you would back in 1975, and they include a check in their envelope. Then you’ve got customers who receive a catalog, and they order over the telephone. You have customers who receive a catalog, and enter a key code on the Web sites, and place their order on a Web site. You have customers who receive a catalog, and don’t tell you anything on the Web site about the fact that they got a catalog; they just order something. Then, you’ve got customers who use paid search and other online advertising forums to order online. At the end of the continuum, you’ve got customers who just love ordering online that really don’t need advertising. Within that framework, you’ve got really six or seven channels. Customers who are early in that process, either mailing in their orders or ordering over the phone, you’re not going to get them to switch over and buy on the Web and doing a lot of referrals from paid search, as an example. If a customer is further down that path, they might be likely to do that. What I find is that a lot of my catalog clients right now are right in the middle of that process, where their customers receive a catalog, and they order using a key code, or they order and through a matchback analysis, we find out that at least the catalog was mailed during the past 30 or 60 days. Catalogers are at this big inflection point, where they could eventually spend too much money on catalog mailings, or their customers are going shift online really rapidly, and they’re not going to be ready from a Web site standpoint to handle this migration of customer behavior. Most of the catalog clients we work with are kind of right in that middle stage. Alan:Do you think that catalogers typically are still over-mailing and relying too much on the book? Or are doing it about right, perhaps not mailing the Web folks enough? Kevin:I would say that the majority of the folks I run into are mailing too many catalogs to the Web customers. Alan:You had an interesting blog post describing your time at Lands’ End, where each business unit mailed their very best customers, and as a result, each customer got every single Lands’ End book. That’s reminiscent of what might be going on today with mailing Web buyers, perhaps? Kevin:It’s very similar. Most of my clients are in a position where the catalog is their heritage. That’s what they love. They really want to believe that the catalog is what’s driving those orders. They’re mailing these Web customers a lot of their catalogs, hoping that Web customer will just keep buying as a result of those catalogs. At the same time, they’re also sending those customers emails, and those customers are also kind of self-serving themselves. They’re going out on the internet and finding the information they want. There’s kind of this little bit of a trap, where you get customers, your very best people, the ones who like to shop across the multiple channels, but you end up, from a marketing standpoint, spending a lot of money on them. As a result, they end up being actually a little bit less profitable. There are going to be ways here over the next couple of years to optimize that relationship and that spend that you do, so that maybe you cut back some of the catalogs, and you maybe invest a little bit more in paid search. Over time, then, you kind of push the customer a little bit toward her helping herself, if you will. Alan:Do you think catalogers are at a disadvantage because of their emotional connection to their heritage? Kevin:Absolutely. Absolutely. I was at a conference earlier this week, meeting with a group of CEO’s and every single one of them loved their catalog. They all brought their catalogs to the conference. They were passing them around and showing everybody. Alan:Like pictures of the children. Kevin:Exactly. They weren’t showing pictures of the Web site, you know? Yet, over half of their business on average was happening on the Web site. There’s definitely this emotional attachment to how they’ve marketed to customers over time. Alan:Interesting! For folks interested in these issues, I highly recommend your new book to all the folks listening to this podcast. For 2008, what would be your top tips for retailers navigating the multichannel waters this year? Kevin:I think we’ve touched on a couple, and then there’s a couple of additional ones. When I’m meeting with clients or I’m talking at conferences, I’m talking about this continuum, where people think they have a telephone channel and a Web channel, or a catalog channel and a Web channel. Realistically, they’ve got many more channels than that, and the customers start revolving through these. If a customer is still ordering over the phone, that customer is still going to need to receive catalogs. If you have a customer who’s purchasing, and you’ve seen that referring URL, that it’s a paid search or it’s any kind of search engine optimization activity that’s driving that customer, that’s a customer you treat a little bit differently, and so to kind of look at things across that large continuum, as opposed to just thinking, “I need to mail all my customers.” I think another thing that’s important is Matchback analyses. I think the best way to describe this is in a matchback analysis, you find out that you drove a lot of orders to the Web site by mailing of a catalog. You can take another group of customers and not mail them a catalog, and then 30 days later or 60 days later, submit them into the matchback analytics process and see how many orders the process suggests were caused by a catalog. What I mean by this is you know you didn’t mail the catalog to the customer, and yet the matchback process will say that some orders were driven by catalogs. That tells you kind of what happens organically. That organic thing is the thing you really want to learn, because that tells you, “If I don’t have a catalog business, how much of my business is going to be protected?” That truly tells you what your strategy should be from a catalog-mailing standpoint, so I think that’s a big deal. Another thing that I think we will have to, as an industry, do over the next few years is link visits to the Web sites to purchases in other channels. Right now, for instance, in cataloging, we look at the fact that a catalog drove 20 percent of all the orders in the Web site during a certain period of time. We would look at a cataloging very differently if we said it drove 70 percent of the visits to the Web site during that period of time. The Web has a responsibility of converting a customer to order something. I have clients who maybe have a six percent conversion rate on their Web site, and they see that they do a really good job of having catalogs drive business to the Web. I’ve got other clients that have a three percent Web site conversion rate, and they say their catalogs don’t work as well at driving business to the Web. But it’s really the Web site that’s causing that effectiveness to either happen or not happen. Alan:Interesting. Kevin:There’s going to be this process, where linking Web visitation behavior to marketing activities at the purchases is going become increasingly important. At Nordstrom, we knew, for instance, that during the course of a month, a multichannel customer would visit the Web site three times, would visit a store twice, and buy something once, and they would use any channel to buy. It told you how integrated everything really was for our best customers, so I think that’s going to be a big deal. I’d say the last point is really for catalogers to have a vision for where their business is going to head over the next five years, and the same thing for retailers. Earlier this week, I was meeting with executives, and we were trying to see where the business is heading. With prospecting challenges in cataloging, and with catalog choice slowly gaining momentum as a way for customers to not receive catalogs, you want to look ahead to what your business might look like in five years. And then decide what your marketing strategy should be over the next five years to get you to a place that’s positive. I’d say those are the things that database folks and catalogers, in particular, are going to be looking at. Alan:Excellent. Thanks for all your time today, Kevin. Kevin:No problem. Thank you!
Listen to podcast: RKGblog_Interview_Kevin_Hillstrom.mp3
Join the Discussion