What three-letter IT investment is huge, expensive, complex, risky, enterprise-critical and just about inevitable? If you think the answer is "Enterprise resource planning (ERP) and we've done it already!" then it's time to phone a friend and get the bad news. The correct three letters are, of course, CRM - customer relationship management - and even if you've started a CRM project you certainly won't have finished it.
At the CRM show and conference in Olympia last month, there were a lot of companies that are keen to plunge right in. And, judging by the sea of IT suppliers on hand, there are plenty of products to choose from. Put the letters CRM in your sales pitch - even better, prefix it with an 'e' - and suddenly a lot of users want to know you.
But from a user point of view, the sheer numbers of IT companies claiming to be in CRM - in a market that is increasing at 40% a year and will reach, says market analyst Frost &Sullivan, $7bn (£4.4bn) in Europe alone by 2005 - makes finding some kind of sane reality even harder.
So how do you make sense of it all?
First, accept that CRM affects a range of company activities, from making more of your call centre or tracking customer movements on your Web site to keeping the salesforce in order.
Second, understand that CRM is moving into new areas - most notably the Web, of course - and occupying new niches. As Paul Homan, principal information systems architect at the Post Office, points out, "The CRM space keeps growing faster than vendors' offerings."
Moreover, in each niche, there are half a dozen suppliers.
This broadening range of areas where CRM can be applied and products to use means that with CRM, unlike ERP, there is no single-product solution. Even the largest suppliers do not and cannot cover the entire range of CRM activities. CRM is, therefore, inherently component-based. You will need more than one product to do it all.
The corollary of this is that you will need to make the separate CRM products work with each other. Furthermore, you will need to make your CRM systems work with your back-office systems as well and, as supply chains tighten and business partners proliferate, you will probably need to make your CRM systems work with those of other companies you do business with.
So, do the CRM products available all integrate and interoperate well with each other "Of course not," says Michael Juer, director of CRM consultancy Sales Pathways. "The interfaces are awful. You'll increasingly need enterprise application integration to do CRM."
Nor will it just be the problems of technology integration. There will be organisational ones too. "Sales, service and marketing all sit in separate silos," points out Juer.
And each department will have its own political agenda for CRM with which the IT department will have to contend. This highlights the need for a high-level senior business sponsor for CRM who can knock departmental heads together early and often.
There is yet another caveat when it comes to implementing a multi-product, multi-supplier CRM architecture, warns Paul Lethbridge of marketing consultancy Peppers & Rogers, it is essential to avoid the stovepipe trap, where each individual CRM system, addressing its discrete part of the CRM spectrum, has its own customer database and business rules.
Instead, he urges, users need to understand that in a CRM architecture, some elements can be unique and standalone, but some must be common. Common elements comprise the definition of customers, the normalised database of information about customers, the analytical engine trawling that database, the campaign management software and the business rules that drive it. Standalone elements can be the variety of source data, including operational production systems and third-party external data that feed the normalised customer database, and the vertical channel systems running all the customer facing systems including the Web, call centre, branches, interactive TV and sales.
But do you have to do all CRM all at once?
"It's terribly important to have all the channels covered by technology," says Juer.
Homan also emphasises the comprehensive nature of CRM. "Yes, you can improve individual areas such as the front office and sales with CRM, but unless you do end-to-end CRM you won't get any real benefit," he says, "but doing it end-to-end is a major headache."
If the strategy has to be enterprise-wide, the implementation should not be. So it makes sense, says Homan, to deliver in the areas of most pain and gain first.
Lethbridge agrees. You can get value from point solutions addressing just one or a few niches on the CRM spectrum, he concedes, but you get the greatest gains from end-to-end. You still need concrete deliverables, of course.
Nevertheless, one school of thought argues that cherry-picking is the best strategy. Research from Andersen Consulting suggests that being selective in where the CRM investment goes, can pay off the best (see boxCRM successes). In the chemicals and communications industries, for example, Andersen found that only five out of 54 possible points of CRM investment across sales, service and marketing had any real impact. That impact, however, could be hefty: improved CRM performance from average to superior could be worth $100m (£68m) in extra profits to a billion-dollar chemicals company.
Of course, it's difficult to tell up-front just what will have the greatest impact.
As ever, whether the strategy is cherry-picking or wall-to-wall, the key factor is to ensure that there is a well-thought-out underlying architecture, even if not all the slots are filled at once, giving flexibility for change, growth and replacement.
"You can design CRM in various ways," says Homan. "You can design it around your organisation, such as asking your salesforce what they need and building it. You can design it around the available technology, picking one of the leading companies recommended by analysts. This is a good way to gain ground quickly, but it only gets you parity with competitors."
But the best way, he says, "is to design it around your customers. That gains you competitive advantage."
An innovative, but definitely brave, company would design it with the help and advice of the very customers it seeks to manage better, but this is a high-exposure option.
"No one I've spoken to has engaged their end-customers in helping to design their CRM systems," says Homan.
More likely, he says, your system is designed with your business partners in what is increasingly an extended enterprise marketplace.
Whatever approach to CRM you take, all the commentators agree - expect to spend a lot of money.
"CRM is expensive," says Lethbridge, "because implementing big, complicated enterprise software that is highly scalable and can do thousands of transactions is not cheap."
The key aim is to ensure that what you spend is well spent. Already, commentators warn, a lot of users' money is disappearing into software suppliers' pockets. Homan calls the licence costs of one leading CRM supplier "obscene".
Across the board there is an expensive tendency for over-engineering CRM products.
"There's a massive amount of gold plating," warns Homan. "You're buying stuff you can't use, such as screens that can be configured 1,800 ways."
This level of flexibility means, he says, that, unlike ERP where you just do inventory management the way the module does it, with CRM "configuration and customisation blur completely".
Take it as read, Homan says, that "all vendors lie about their products. There is a huge amount of arrogance and legalism in this marketplace."
Implementing a complex architecture in an immature product marketplace almost inevitably means users will need a helping hand, whether from a consultant or a systems integrator. Beware of both, warns Homan, "There's so much school bussing - consultants who have come off the vendor's product training course the week before. You have to hire on the cv [of each on-site consultant] and retain them, or they suddenly drift away [to another customer]."
"It's a big issue, because there are not that many people who have done CRM and done it well," he adds.
Similarly, systems integrators - whose experience of CRM projects is in the same ball park as consultants - are, says Homan, aware that "notoriously many CRM projects get canned - so they want a quick payback.
"Watch resource levels - they're frightening."
Finding a CRM guru among the suppliers, consultants and integrators can be invaluable to the success of your project, says Homan, if you can find one.
Not all the costs will come in from outside. One huge source of time and cost is sitting right inside your organisation. It is a safe bet that every organisation possesses a large number of databases in which customer information is already stored, and which will need to be pulled together into the CRM customer database.
"Try and find a single definition of a customer in all your databases," suggests Lethbridge. "It'll be a challenge. I've seen anything from eight to 28 databases in an organisation and each database has customer information and no links between them."
Homan concurs. "I've heard fantastic horror stories about data cleansing. One company found it had 28,000 Mickey Mouses. Another was just about to send out a sales letter to its major account customers, it took the names straight out of the marketing list and found they'd been addressed as 'whinging git'."
Mice and gits apart, creating a common, collated customer database is a major challenge, both time and money-wise, even more so if that database is to take in third-party data such as demographic lists and customer data from business partners.
Lethbridge is not joking when he says the first task of an IT department challenged with corporate CRM is to "find out the number of databases" their company has.
Even worse, don't expect all the relevant data to be in databases in the first place. Sales staff, for example, are notorious for not pooling their information. They'll either keep their prospect data firmly tucked into their personal organisers or even more firmly tucked inside their skulls, he warns.
This attitude may well continue right into implementation. Given the traditional tendency of sales staff to hunt on their own, it isn't surprising to hear statistics such as 60% of sales force automation (SFA) projects fail - but not because of the technology. Sales staff will vote with their feet not to enter the data that, they believe, gives them their own competitive edge.
"Sales people don't want to know," says Juer. "They don't want to communicate with the customer or each other or be accountable to their employers." It will take a new generation of sales people to accept that SFA is here to stay, he believes.
Until then, "only 40% of the SFA licences sold are being taken up and used in production. There's a lot of shelfware," Juer says.
Lethbridge agrees. "There's a lot of CRM product licences out there not being executed."
So, all in all, the news from the CRM frontline is sobering. While the benefits of CRM from a business point of view are difficult to refute, from a technology point of view the difficulties of making it happen - cost-effectively, sustainably or even at all - are more difficult to refute. Any IT department that sees enterprise CRM heading its way knows it's going to be in for a bumpy ride. "Beware of others blaming you," Lethbridge says. "CRM is one of the most complex and challenging programmes your organisation will have even thought of, containing many, many career terminating events."
And, just like ERP, expect that there'll be a lot of CRM foul-ups everywhere you look - if you foul up, you'll be in good company.
So what is customer relationship management?
From a business point of view CRM is exactly what it says it is - managing your relationship with your customers. The purpose of doing so is threefold:
"It's about extracting more value out of a customer," says Michael Juer of consultancy Sales Pathways.
The concept of customer share rather than market share is of growing importance in business, and reflects increasing recognition that the focus of business is changing.
"Most businesses compete on three things," says Paul Lethbridge, of marketing consultancy Peppers & Rogers. The first is operational efficiency but these days, he warns, with most of the corporate and supply chain fat already sweated out "a multimillion-dollar spend improving operational efficiency gains a minute improvement in profitability".
The second is product performance, where competitive advances are all too rapidly eroded.
The third is on customer management - making your customers want to do more business with you because they get better service, the products they want, when they want them, via the channels they want to buy them through, and so on.
To do that efficiently and effectively requires, of course, technology.
Be warned, says Lethbridge, CRM can either be "the exploitation of technology to enhance and personalise customer interactions with the objective of acquiring and retaining customers" (and making more money out of them along the way), or it can be "the random deployment of expensive and elaborate technology in a desperate bid to be more exciting than your competitors".
There are, overall, two types of CRM technology, says Mike Ferguson of Database Associates - operational and informational. The first is process-focused and the second is customer intelligence-focused.
The first type attempts to impose the same degree of automation all the myriad customer-facing processes in a business as has been traditionally applied to internal processes. While enterprise resource planning (ERP) systems run the business "under the hood", CRM systems touch the customers or those who face the customers. If ERP automated the back-office - finance, logistics, personnel - CRM automates the front office - marketing, sales and support.
The second type has at its heart a data warehouse and analytical software to store and explore all the information gathered from all the other systems - CRM plus others - about customers, their behaviour and desires, to discover how best to target their wallets and increase their profits to the company.
How to minimise the CRM burn
Speakers at the recent CRM 2000 show pinpointed the following tips on achieving a trouble-free CRM implemenation.
Research conducted by Andersen Consulting into the chemicals and communications industries identified five big winners where investing in CRM really paid off:
Paul Homan, Michael Juer and Paul Lethbridge were all speaking at CRM 2000, organised by Business Intelligence. Mike Ferguson will be running a workshop on designing and building intelligent CRM solutions in London on 7-8 December see irmuk.co.uk
Problems you will encounter with CRM