VRM, Customer Data and Competitive Advantage


Share on LinkedIn

The original aim of writing my post on Four Fallacies of Vendor Relationship Management was to challenge VRMers to shed some light on what I see as fallacies in, admittedly, a fairly extreme model of VRM. As Doc Searls said in his rejoinder, nobody has really defined what VRM is yet, as it is an idea that is emerging (and evolving) out of the many discussions swirling around the subject.

So far I haven’t seen a strong argument against any of the fallacies that I wrote about.

Fallacy No1 is a case in point. As Rob Knight points out in an excellent post on VRM: Rent, Don’t Buy, companies with large amounts of transaction data (thousands of transactions per customer) can mine it for their own competitive advantage. Tesco leads the UK and possibly the world in supermarket retailing because of Dunn Humby’s excellence in data mining. Capital One leads the credit card world because of it. All the retail banks, mobile telcos, airlines and utilities are trying them damnedest to extract competitive advantage, no matter how temporary, from the data that they have collected during transactions with customers, that they have organised in vast data warehouses and that they thus (legally) own.

As Akerlof, Spence & Stiglitz point out in their 2001 Nobel Prize for Economics lecture, asymmetric information of the type that large data owners have gives them an advantage in their dealings with and over customers. They know more about customers past behaviour than customers do, (although they still know nothing of the context in which customers buy). And once they have developed their statistical next-best-product models, smart-customised products and implemented them through highly granular customer segmentation, they can reap the benefits at a low per transaction cost. Tesco reputedly manages 20,000 different customers segments. Some Nordic telcos manage 10,000 customer segments. And the master of data-mining, Capital One, makes over 50,000 marketing experiments each year.

There may possibly be more value from dealing with customers individually on customers’ terms but the complexity and associated transaction costs of doing so rapidly get out of hand. So why should large data owners bother when they are doing so well without actively ceding any control to customers?

One thing I just don’t understand in the talk about data ownership in VRM is the ‘rental’ idea: where customers manage their data and rent it to sellers when they want to buy something. I can understand how that might work for high-complexity, high-cost products that people are passionate about and where the asymmetry is on their side. Racing mountain bikes for example, where each bike is individually customised and can cost thousands of dollars. Here VRM through a reverse market for buyers, or a multi-side market for buyers and sellers probably makes sense as a potential solution to finding desired components.

But I can’t see it working for something like a supermarket where there are hundreds of low-complexity, low-cost transaction items per week, or a mobile telco where there are potentially thousands (and where the customer doesn’t even have access to the CDRs).

  • How would the customer collect the data?
  • Where would they store the data?
  • How would they make sense of the data?
  • How would they offer their data to sellers?
  • Is it really worth it for commodity products like butter, baked bean and toilet rolls?
  • If it is too difficult and expensive for small retailers to do, what chance to customers have of doing it?

I think all these questions provide real barriers to VRM for everyday, commoditised products. And almost by definition, it is these products that makes up the vast majority of transaction data about customers.

I would be interested to hear from readers what you think about this interesting topic. It is time that VRM was clearly and unambiguously defined and discussed out in the open. And that average Joe customers were asked whether it makes sense to them. Most of the VRM discussions seem to have been largely devoid of customer input so far. And as all those who work in innovation know, not involving customers at each stage of the innovation process is often a prelude to another expensive flop in the market.

Graham Hill
Customer-driven Innovator
Follow me on Twitter

Further Reading:

Graham Hill, Four Fallacies of Vendor Relationship Management

Rob Knight, VRM: Rent, Don’t Buy

Doc Searls, A little understanding goes a short way

Akerlof, Spense & Stiglitz, 2001 Nobel Prize for Economics

Graham Hill (Dr G)
Business Troubleshooter | Questioning | Thoughtful | Industrious | Opinions my own | Connect with me on LinkedIn https://www.linkedin.com/in/grahamhill/


  1. Hi Graham,

    I’m enjoying your stirring of the pot on VRM, but i’m with Doc in that we need a bit more time, and to put the code in place before going to research and banging the drum more loudly.

    My own view is that VRM without a robust ‘personal data store’ is like CRM without the ‘single view of the customer’/ analytical database; i.e. it’s just tinkering around the edges.

    The personal data store is easy to talk about, trips off the tongue very nicely – but by no means easy to deliver in practice; at least not without validation and verification capabilities that will be required for meaningful interactions and transactions.

    The build of that capability is where we are headed with this social enterprise (www.mydex.org). We have just finished the build of a proof of concept (which, yes we then take to both individuals and organisations in research mode).

    Hope that helps clarify where we are on at least one aspect of VRM – drop me a note if you want to discuss further/ have a look at our proof of concept.



  2. Hi Graham

    Thanks for your interesting pushback. A couple of thoughts.

    First, you are concerned about definitions. Well, we cannot define everyday terms like space, time or gravity so I’m not too fussed about being perfect here. But, to me, it’s pretty simple. There are two core pillars to VRM.

    First, ‘personal data stores’ which individuals use to gather, store, analyse and use the information they need to manage their lives better and more efficiently: manage my home, my money, my health, etc. That doesn’t sound astonishingly outlandish to me. I suspect you use an incipient personal data store every day – the contact list on your mobile phone. As to how much data people will want to populate their data store/s with, that’s up to them – whatever they feel is useful. But the central departure point here is simple. This is a service to the individual without any regard to vendors’ data management strategies, priorities etc.

    Second, ‘volunteered personal information’ or VPI. This is an inevitable knock-on from the personal data store. If I want to manage my home, my money, my health, etc better and more efficiently then I need to interact with suppliers. Which means doing things like telling them who I am, what I want or need, what my changing circumstances and plans are, making arrangements etc. VRM helps people do this in a structured, efficient way. Once again, pretty straightforward, I would have thought.

    So let’s turn to your four fallacies.

    Your first fallacy is that companies are not going to cede control of data easily or willingly. I absolutely agree. But the real issue is the data they don’t have any control over, such as my transactions with another supplier in the same category, the fact that I have moved home without telling you, or that ‘I’ve decided to buy a new car some time in the next three months and am actively researching this purchase – but you don’t know anything about this because I haven’t made a transaction yet and I don’t see any reason why I should tell you’. Companies can’t keep control over data they’ve never had access to.

    Your second fallacy is that customers don’t want control over their transaction data. The issue here is not control for control’s sake but value. I currently have three current accounts with different banks for different purposes, three or four credit cards, some savings schemes, a loan or two, and a mortgage, plus a pension plan or two. I have relationships with at least a dozen different financial services suppliers. If I want to manage my finances better, I need to get my own ‘single view’ of all my dealings with these suppliers, which means I need to put the information together. To do this, I need transaction histories. Pretty obvious really. Yes, I may want transaction histories to manage my finances better but I may not be interested in groceries. That’s up to me. I’m not going to assiduously collect transaction data that I’m not going to use. On the other hand, as companies have already found, there are a surprising number of ways of getting value from transaction data. Why shouldn’t individuals do this when and if it adds value?

    [By the way, the fact that I have all these different relationships also explains why so much ‘propensity modelling’ and other supposedly sophisticated data analysis is so much nonsense. Each one of these organisations has a ‘single view’ of me based on what business I decide to give them. To my savings suppliers, I’m a saver. To my lenders, I’m a borrower. They each build their picture of me based on a highly misleading, isolated sample of data. That’s why their predictions as to what I’m going to do next are so awful, and why the returns on the marketing initiatives based on these predictions are also so awful.]

    I have to say I don’t understand where your third fallacy about customer managed markets (and your fourth one about planned economies) has come from. In all my discussions in all my many Project VRM steering committee conference calls and meetings I’ve never come across anybody who has said anything like this. So I just draw a blank here.

    Your fourth fallacy is interesting in one way, however. Of course, you are absolutely right about the importance of transaction costs in defining the nature and boundaries of the firm. But what this school of economics ignored was the individual’s transaction costs: the transaction costs of going to market and of trying to organise, administer and integrate the multiple, siloed, isolated suppliers I need to work with to move home, for example. VRM helps individuals reduce their transaction costs. Why is that such a silly idea? Do individuals like high costs?

    In fact, transaction costs are probably the core economic opportunity opened up by VRM. In the course of trying to manage their lives better – researching, making and implementing decisions – individuals generate huge amounts of new information about what they want, from whom, when. VRM helps capture this information, and enables individuals to pass it on to selected suppliers if they think it’s going to add value. From the supplier point of view, this offers them the opportunity to eliminate huge amounts of guesswork about who to market to, about what, via what channels, when – thereby cutting costs and generating greater returns at the same time. Lower costs, greater value and a win for both sides. That doesn’t sound so off the wall to me. In fact, it sounds like ‘same old, same old’: the three defining features of all successful business models.

    Best wishes

    Alan Mitchell

  3. Graham,

    I put half a day into a long and thoughtful comment to your original Fallacies blog post. It never appeared. Probably my fault. I have no idea. But I also don’t have the time or energy to respond at length right now. And I think Alan Mitchell has done a fine job in any case. (Thanks, Alan!)

    Two quick thoughts in the meantime.

    First, VRM has never been about “customers manage their data and rent it to sellers when they want to buy something.” That’s just wrong.

    Second, the arguments against user control of data are all posed on the seller’s side of the marketplace. In the entire Industrial Age we have never had a market condition where customers controlled most of their own data. The Internet not only makes that possible, but inevitable. The Internet has blown up many other asymmetries. (Just ask the recorded music business.) This one will get blown up too.

    Arguing that vendors will never give up trying to control customers with asymmetric data is like arguing that AOL, Compuserve, Microsoft, Apple and Prodigy will never let the Internet happen.

    But the Internet did happen. And now that it’s here, serving as a platform for symmetric business relationships, VRM will happen.

    It’s critical to understand that VRM is not about customers vs. vendors. It’s about customers + vendors. It works with AND logic, not OR, or the more box-officy VS. The only thing that will work is VRM + CRM. In fact I’ve been told that the top official at one of the big CRM companies has said “Whoever wins at VRM wins at CRM.” Why not?

    All we’re talking about here is new tools for customers: ones that give them much better means for relating to vendors — to hold up their ends of relationships. It’s not much more complicated than that.

  4. Graham:

    You said — “I can’t see it [VRM] working for something like a supermarket where there are hundreds of low-complexity, low-cost transaction items per week.”

    I can. And I’d really really like it… like, now. After I tell you about my “Average Joe” trips to the supermarket, I think you’ll understand why.

    According to the social contract I have worked out with my wife, I cook all meals on the weekend, so that’s usually four or five meals. Over the years, I’ve managed to develop a short list of dishes she likes on a repeat basis, but she likes to try new stuff.

    So on Thursday and Friday, I pull together a list of recipes from various sites, like Epicurious, AllRecipes, RecipeZaar. I cut and paste the ingredients lists into an email I send to myself so I’ll have it on my Blackberry.

    If I’m smart, I’ll also remember to add the list of stuff we always need, like the type of milk my wife likes, the diaper cream my kid needs, and so on. This is my shopping list, right? Pretty darn average.

    Now, there are four decent-sized supermarkets on my way to the house. Some are closer to the road, some farther; some cheaper, some fancier. I don’t have any particular loyalty to any one of them.

    My goal is to spend as little time (and gas and cash) as possible shopping. Ideally, I get everything I need at one store. I hate having to go to two places. Three stores = total nightmare.

    So as I leave work on Friday, I can do two things:

    1. I can call customer service at one of the supermarkets and ask about some of the more exotic items I need. This past Friday it was a kind of hot sauce specified in a recipe for gourmet Buffalo wings. (Yeah, you WISH you were at my Super Bowl party.)
    2. I can risk it and pick one of the supermarkets without calling first.

    Both options suck. It’s really time-consuming for the customer service folks to look up items in their systems; sometimes they even need to call other departments; they often make mistakes or are unclear. But when I don’t call, I often end up going to two or three places. In fact, it’s not uncommon for the store I randomly pick to not have one of the ‘commodity’ items that I figured they would, like the diaper cream.

    So this is what I want:

    I cut and paste the ingredients from the recipes, plus my standard always-need list, into a shopping list VRM tool. I ‘expose’ this need-data to the four stores (or, frankly, any along my driving route), and I get back a grid showing which stores have which items at which prices, plus a total price for each. You save me a trip to a second or third supermarket, this Average Joe is very happy.

    Now obviously there’s a lot more the tool could do. If it knew my loyalty card numbers, it could show me their ‘personalized’ discounts for various items on the list, maybe even for items I have bought in the past that aren’t on my shopping list this time. That could be useful too.

    And if it kept track of the entire list of stuff I’d bought in the past, it would make it easier to specify things I’d liked in the past but I’d forgotten the brand when making the list.

    So, yeah, BuyRite and Wegmans and Giant would have to expose their inventory systems to this VRM system. But they are doing that anyway, at much higher cost and lower service quality, when I call up the customer service guy. Right?

  5. Jonathan,

    Great scenario!

    Isn’t a loyal customer’s actual shopping list more interesting to a seller than guesswork about what it might be? Shouldn’t loyalty mean more than an inconvenience with a barcode that slows down checkout lines, requires dual pricing schemes, and justifies itself by imagining that it increases “switching costs” for customers?

    The answers are all yes. Which is why VRM is a fun challenge.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here