VRM: Rent, don't buy

Jan 27, 2009

Recently I wrote a post outlining some of the basics of vendor relationship management - a new-ish concept in how individuals and organisations can relate to each other and communicate their wants and needs.

Since writing this, one idea has been rattling around in my head with increasing velocity:

The potential business benefit of VRM could be for smaller companies which cannot afford the costs of massive data-gathering and CRM operations, but could trade off ownership of the data in exchange for access to it. Many smaller companies, most likely internet-based startups, would be very happy to allow individuals to maintain control and ownership of data, since they could never gain access to it any other way.

Also in the last couple of days, I’ve been reading an interesting post by Graham Hill, in which he identifies the ’Four Fallacies of VRM’. I may return to some of the others later, but right now I’m interested in the first in the list:

The Fallacy of Ceding Control - For data rich companies like mobile telcos, credit cards and utilities, cutomer transaction data is a source of huge commercial advantage. It creates literally billions of dollars of incremental revenues for them. Why would these companies and the many others like them give up their expensively collected data and the power it confers, to a rabble of customers? Would you if you had paid millions to collect it?

In many ways, I don’t disagree with Graham’s point here. Existing CRM operations do provide a competitive advantage to the organisations in possession of the data. There’s no doubt that Tesco’s growth to retail dominance has, in large part, been due to the company’s superior data management, as Kevin Meyer (quoting BusinessWeek) pointed out in on the Evolving Excellence blog:

Analysts say that Tesco’s big advantage over major international rivals, which also include Germany’s AldiLidl, is its unrivaled ability to manage vast reams of data and translate that knowledge into sales. While data crunching may sound dull, it has given Tesco two major advantages: an unmatched ability to operate multiple retail formats—ranging in size from convenience stores to hypermarkets—and the market knowledge to offer what many analysts say is the best and broadest range of house brands from any retailer.

So it’s quite true that companies like Tesco aren’t going to be abandoning their CRM systems any time soon. And they are joined in this by many leading companies in other sectors. But Tesco and their equivalents are successful companies at the top of their industries, with little need to think of reversing what has been a very successful strategy. The same cannot be said of many smaller players that are presently finding life much more difficult. Data mining on the scale of the largest operations is hugely expensive, prohibitively so for many of the companies who might think of competing with the biggest. The homogenisation of the retail industry has been driven by the fact that the biggest companies can gain incredible economies of scale and can successfully perform better than their smaller rivals, and not by ‘exploiting market power’ or using loss-leading prices to drive rivals out of business - they can produce perfectly legitimate superior performance by virtue of their greater knowledge of the marketplace, a knowledge simply unavailable to smaller companies.

But personal data is basically like any other costly asset. It has some obvious attributes:

  • Expensive to acquire
  • Depreciates in value rapidly unless maintained at further cost
  • Only truly valuable in significant quantities

In short, acquiring this kind of data is only worth it if you can commit to the up-front cost of initial data gathering, plus the on-going cost of maintaining your data collection. Many smaller companies do not gather data simply because it’s too expensive for them; on a recent pitch for an e-commerce project, it was mentioned that the company in question haven’t invested in segmenting their customer base to any great extent because of the cost. And the idea of this kind of data processing is out of the question for startup retailers. CRM represents a considerable barrier to entry and innovation in some areas of retail.

What VRM offers is a different means of using data to drive a business, where you 'rent’ the data from its owners - the customers - rather than 'buying’ the data through mining. Just like renting, you don’t have to worry about the asset depreciating in value because you’ve failed to maintain it - in a VRM scenario, you get automatic updates to the personal data of your customers so long as they continue to consent. The costs of acquisition are negligible; it might be necessary to offer some incentive to customers to provide their data, but this can be in the form of discounts or other benefits which only apply when you’re generating cashflow in the first place. And as the quality of data improves, the need for overwhelming quantities of data is mitigated.

That’s the case for small companies, but would we ever see larger businesses giving up control of their data assets in the way that Graham is sceptical of? Like all innovative ideas, it appears impossible right until it appears inevitable, and I have to admit that I have my doubts too. One big reason for inertia is the sunk cost fallacy: the fact that we’re often unwilling to give up on something we’ve invested a lot in, even if doing so would be beneficial in the long run. Graham alludes to this by asking “Would you [cede control of data] if you had paid millions to collect it?”.

So, it’s my view that Graham’s argument (at least Fallacy #1) rests on the fact that VRM might not work for companies which are enjoying the status quo. This is true, but it’s only a reflection of the fact that it’s a lot easier to make predictions based on known quantities. The future, however, will contain surprises. In particular, it will contain small, new, innovative companies which will be seeking a means of getting access to data on something closer to the 'rental’ model, because of the huge acquisition costs of running one’s own data mining operation. I’m willing to bet that they’ll be a lot more open to VRM.