The truth is that many have found that CRM systems have soaked up so much money that they have little money (or energy) left to spend on important things – like communicating with customers.

CRM has proved to be a bucket with a large hole in it.  Customer data goes out of date at an alarming rate.  Replenishing it costs a fortune.  And I wonder how far huge CRM systems can really predict customer behaviour at all?  Essentially, they rely on analysing past behaviour to map the future.  Yet we live in an unstable and fashion-conscious world.  Who could have forecast the success of the Atkins diet and its impact on consumption?  Three years ago, $400-a-night hotels were the niche every major hotel chain was chasing.  Today that niche has become a tomb.

In any case, people change.  The head of database management at American Express once told me that, however hard he profiled people, 25% of them jumped from one category to another every six months or so.

Of course, various forms of CRM have worked for the lucky few.  Through a judicious combination of database-mining, good direct marketing and the use of call centres, some financial-service outfits have made money.  There can be no doubt of the role that CRM-driven customer loyalty schemes have played in enabling British retailer Tesco to increase its market share.  And some pharmaceutical companies are using CRM to successfully map influencers.

But these are exceptions.  In particular, I think, CRM has proved a huge disappointment to anyone who sells mainly through intermediaries.

So why has CRM proved to be an undeliverable nirvana?
Firstly, I think we should all have the courage to admit that, in most industries, marketing remains an art.  Most marketing relies on the flair, imagination and intuition of individuals.  And most marketing, particularly in business-to-business sectors, is either not measured at all or is measured entirely by your thumb.  Against this background, the very idea that we can construct hugely complex databases predicting our customers’ behaviour is laughable.  It is like asking Stone-Age man to drive a motor car.

Secondly, segmentation remains a huge problem.  All CRM systems depend on companies coming up with proven, effective segmentation strategies.  Without segmentation, companies cannot expect to derive any customer insight, no matter how much they spend. 

Yet segmentation is still a dark art that few practise successfully.  All too often, it is along the lines of little, middle and large.  A small minority of companies come up with segments that go further.  Still fewer go on to successfully deploy campaigns around such segmentation exercises.

Catastrophically, large CRM systems encourage over-segmentation.  This is an easy trap to fall into.  The truth is that coming up with a coherent and successful direct-marketing campaign is pretty difficult in the first place.  Coming up with twenty such campaigns, each of which supposedly addresses a different segment, is overwhelming. 

The test of good segmentation is that each segment should be Identifiable – can you put names to customers? Reachable – can you access the segment through channels? Relevant – what are you going to communicate and offer that is unique to this segment? And Cost-effective – Can you generate returns that justify the costs of addressing this segment?

Instead of investing tens of millions in CRM systems, companies should be spending more time on getting the basics of segmentation right.  And, if you sell through intermediaries, it is vital that you consult them early on in this process.  Your channels are not just about communicating, but also about selling, delivering and supporting (post-sale) offers.  The priority is to develop segmentation that drives the entire process.