More retail data woes
I got a lot of mail on the piece I wrote last week about retail data collection. The mail ran 2 to 1 in support but it was still interesting to gauge the reaction of those who think the notion of collecting data like zip codes, drivers license numbers and the like is kosher.
Perhaps the best response to those who think all the numerology has a point came from Kevin Hillstrom who has a blog called, “The MineThatData Blog.” In the blog Hillstrom wrote that the piece “…correctly illustrates the uselessness of the information” collected. He goes on to note that retailers are at the mercy of a lot of market forces today including what I call the de-massing of the mass market—there is no “mass market” any more, just a lot of large interest groups that each need to be satisfied differently.
My favorite bit of mail came from someone who will remain nameless who supported the status quo with, “…how good a job most retailers do of that [capturing customer attitudes] is a fair question. But you shouldn’t say they’re wrong to try.”
Actually, I have lots of issues with this idea. It’s one thing to make a valiant effort when you think you are doing the right thing but once that effort is proven to be ineffective it takes real good denial skills to persevere. Check out what’s happening in Washington these days and you will get my meaning.
On the heels of all this, Stop and Shop the operator of 385 grocery stores in New England, New York, and New Jersey revealed that its security had been breached and that customer credit and debit card information may have been compromised by tech thieves. Stop and Shop is part of the Dutch company Ahold which operates about 1,300 supermarkets in different regions of the country operating under various names.
Let me say again in plainer language, collecting my zip code, driver’s license number, and credit card number and the like and not securing it while thinking you will learn something about my shopping preferences is a fool’s errand. The only thing you will learn for sure is how irate I can be at this intrusion into my privacy. Actually, you might also learn how to say ‘class action’ and it surprises me that people who have been harmed by lax retail data security have not become better organized.
On the flip side, one major retailer wanted more detail on what I meant when I said there are better ways to collect more meaningful information about customer attitudes and behaviors and I am happy to provide my ideas here. They boil down to the customer community of interest or just ‘community.’
Communities, if run well are marvelous things. I have studied communities of very rich people and special demographic groups and each one worked amazingly well to reveal what’s on people’s minds. In the hands of good product and marketing people, those ideas turn into better products, services, and messages.
What might surprise a traditional marketer is that the groups I have had exposure to all represented a demographic that was hard to reach by conventional marketing methods. For example, 18 to 24 year old men are notoriously hard to pin down; they have more important things on their minds than giving a lot of time or thought to marketers. Yet in a group I studied the participants couldn’t wait to tell the community leaders about their lives and preferences. The same was true of rich people talking about their investing habits—you could barely shut them up.
The reasons are the same in both groups. People participate because they believe someone is listening and corny as it sounds, they believe they are making a difference. Compare that situation to the gum snapping, surly, and over worked clerk behind the customer service desk at your local retailer and you get a stark contrast.
What’s interesting about communities is that they can ultimately serve two internal business processes, product development and pure marketing. In other publications and on our Web site I have begun referring to these processes as “deep marketing” which simply means what you do long before you engage with the customer to ensure you have the best chance of understanding the customer’s needs and then of consummating a sale.
There are three to secrets of building a community—trust, freely revealing ideas, and authentic communication and retail data collection falls flat on its face in all three areas.
It’s in the authentic communication, though, that old fashioned retail data gathering is best exposed. In most retail situations that I know there is no authentic data gathering—you can’t call capturing a zip code or any other information reluctantly handed over at the cash register anything more than data extraction.
If retailers took a community based approach to customer data gathering, the headlines would quite possibly be different when there’s a break-in. Rather than being worried about liability, the retailer would have a true sense of loss knowing that information that it had uniquely gathered might have fallen into the hands of others. At this point, too many retailers have the same information about you and me so how valuable can it be to anyone looking to do something honest with it?
The sound and fury of retail
The not-so-pretty underside of CRM became apparent again recently with the revelation that TJX Companies, parent of more than 2500 outlets including TJ Maxx, Marshalls, and other retailers, had a break-in that may have compromised customer data. Increasingly, retailers are becoming the targets of thieves because they capture and store huge amounts of sensitive customer information like drivers license numbers, credit card numbers, and more. To put it charitably, many retailers have become better at capturing data than at securing and preserving it.
Experts in the retail and IT industries have been quick to round up the usual suspects meaning that all kinds of security and encryption solutions have been recommended to staunch the flow of customer data to the bad guys. But there are at least two low or lower tech ideas that ought to be at least as prominent as we all grapple with the downside of digital consumerism.
The absolutely lowest tech solution involves my favorite word at a checkout and I honestly don’t know why more of us don’t try it. When they ask for your zip code give them letters instead of numbers—two letters to be precise—NO. There’s no reason why a customer with money to spend has to run a gantlet of retailer demands in order to make a purchase. Either your money or your plastic is good or it isn’t and if it isn’t the retailer better have a waterproof reason for turning you down. Refusal to provide a zip code will not suffice.
If you are using cash, there are laws that stipulate that you can’t be turned away from a public accommodation (which covers a lot of ground) and if you are using plastic there is nothing in the merchant agreement that says the merchant has a right to collect your other private information. Just say no and make a stink of it if you have to.
The lower tech solution, and frankly one that is over due, is asking some basic questions about why we collect the data we collect. The usual answer is a variant on the theme of knowing customer behavior so that the retailer can better focus promotional efforts but I don’t buy that. How is knowing that a woman bought a man’s sweater going to be useful in the future? There just isn’t enough data there. You need to know if the woman is married to the man or has some other relationship, if it’s a present, or if she just liked the style and it fit.
At the end of the day the logic of this kind of data gathering doesn’t make sense. A logician would call it a priori reasoning which basically means that if something happened before it will happen again. It’s like saying that if you flip a coin and it comes up heads that the next flip will also come up heads too, but that’s not true. The two events are independent and subject to the same laws of probability. Similarly every visit to a store is a unique event and needs to be treated that way. Still retailers love to collect historical data hoping it will turn out to be some sorcerer’s stone.
There are lots of variations on that theme but here’s my point—we spend too much time and effort capturing data about behavior and hardly any collecting information about customer attitude yet it is customer attitude that will ultimately bring the customer back. I suspect the reason is that it’s easier to collect numbers at the checkout and associate them with specific behaviors than it is to capture attitudes. When you capture attitudes you might discover things you don’t want to know and that can be scary.
Capturing attitude provides the very important ‘why’ of a transaction as in why the customer likes to shop at a particular store, at a particular time, for a particular reason. Attitude can also tell a lot about selection and placement preferences and a lot more if you let it.
In the bad old days capturing that kind of information was tediously slow and ruinously expensive; it involved focus groups, surveys, and other field research and retailers went looking for it only infrequently as their budgets allowed. Today, though, capturing attitude can be swift and accurate leading to numerous insights, some of which have not even been contemplated.
What’s keeping retailers and other vendors from capturing more useful customer attitude data is a complex subject; it involves the established ways of operating a business and a general reluctance to be the first at implementing new technologies. Retailers might be able to create a moat around their sensitive data and government might at some point require them to do so but all that is a long way from the original intention of better understanding customers.
Even the best security systems won’t change the fact that many retailers are simply not collecting the right information to achieve their goals. If Shakespeare were alive today he might call the current regime of retail data capture the “sound and fury” of retail but as Shakespeare also wrote, it is a tale told by an idiot and it signifies nothing.
Systems Integrators and the New Garage
Salesforce.com issued a press release last week that by now looks rather ho-hum by their standards. They announced that Deloitte-Touche had signed on as a systems integration partner and that got me thinking about how many things have changed, and how many have not.
It wasn’t that long ago that Marc Benioff was comparing Salesforce with Siebel and the different amounts of integration resources it took to get the respective solutions up and running. His point was about the complexity of client-server and the simplicity of on-demand. That was then, today there is still plenty of complexity in the IT department with all of its legacy systems and integrating with them can be a challenge, so why not teaming up with the integrators makes as much sense now as it did then.
Almost three years ago I forecasted that the integrators would have to become involved in the on-demand market for the simple reason that the market was pulling them in that direction. The integrators have great relationships with the largest corporate IT users that go back many years and multiple implementations—if you plan on integrating your on-demand solution with the rest of the legacy applications in the glass house, they are necessary.
Nonetheless, the association was not a slam dunk for the SI’s for the simple reason that there isn’t as much to be gained for them in deploying an on-demand solution. The turning point has been the need that users have for both lower cost of ownership and the reality of their legacy systems. As I said at the time, and still believe, there is still a lot of work for the SI’s to do even if they never take on another traditional integration project.
SI’s could probably make more money sticking with conventional software deployments but the number of customers willing to take on conventional software is in decline, hence the apparent rush to hook up with the on-demand players.
To be sure, on-demand vendors have done a lot in the last two years to make their products appealing to the integrators. Platform technology, open architectures, and openness to custom coding are all responsible for the SI’s interests and the openness comes down to the same basic motivation. It will be years before any enterprise goes to a 100% on-demand IT profile, if ever.
In the long interim, there is real money to be made hooking up modern on-demand solutions with increasingly archaic solutions that originate from enterprise software vendors and internal IT shops. Someone has to manage all that spaghetti and the logical choice is the systems integrator.
The New Garage
Integrator involvement with on-demand is the last part of the roll out of what I called the “New Garage” in 2004. I am still waiting for the term to catch on but idea was that innovators needed more control over their destinies and it needed to become possible once again to build software in your garage or spare bedroom and eventually build a company around it without all the high stakes games played with venture capitalists.
Venture capital is good and necessary but I think we got into a mode during the bubble years—and since to a great degree—of thinking that if you couldn’t get funding for your idea that it wasn’t good enough to pursue. Just looking at all of the emerging companies that are working from the AppExchange makes you realize that not having capital is no reason to scuttle a dream.
On the flip side, a lot of companies that got funding back in the bubble years crashed and took a lot of money with them. There had to be a more efficient way for the market to work so for a lot of reasons it made sense to me that the next move would be back to basics.
VC’s are still an important part of the picture and like SI’s they are morphing. For example, Salesforce.com launched its incubator a few weeks ago with which it hopes to provide “garage” space for innovators using its platform. Incubators like this should enable innovators to make less expensive mistakes and enable the investors to see more before they put serious money into a venture.
This also signals a new kind of VC model evolving too. Rather than investing in myriad companies, investors might begin looking for themes such as specific business processes to support. That approach worked well in the back office and I think it might be time to examine how it works in the front office.
We have already seen the beginnings of a transition from conventional software to SaaS even for the largest enterprise software vendors. SAP recently announced its investment in a new business model for the mid-market, which surely will grow up to take over the company’s largest customers, and Microsoft has been dabbling with various on-demand models for a while, not wanting to put all of its eggs in one basket. Oracle has its own ideas that range from Siebel On-Demand to a hosted facilities management option.
For all that though, it’s the business models of the major software companies that need to change most. It’s one thing to tell the world you can offer an on-demand solution and quite another to handle all the details of a successful partner community which is a necessary part of it.