CHECKLIST: SUBSCRIPTION-BASED SERVICES.

Will subscription-based services finally solve the software industry's perennial problem of boom-and-bust sales cycles? That's still an open question, but a good many companies--including giants like Microsoft, Oracle, PeopleSoft, and SAP--are quietly testing the idea of turning products into long-term rental or subscription services. And so far the idea seems to be catching on.

In fact, Intel has just announced it will invest a cool billion dollars to set up public data centers to host remote services. Citrix has organized a new consortium to promote `application service providers' (ASPs), and research firms like Forrester and IDG are cranking out multi-billion-dollar predictions of the size of the server-based software rental and service business.

We happen to agree that the subscription model is almost certain to emerge as a huge win for many software companies, large and small. To be sure, it's a throwback to old service-bureau days. And yes, customers will end up paying more over time for subscription contracts than they now do for straight purchases (equally true of car leases and home mortgages, incidentally). But if customers decide they prefer to buy services rather than disks--well, it's time for sensible entrepreneurs to start learning how to make a service business work.

Trouble is, services can be a hard business to learn, and product-centric companies often fumble badly when they don't think enough about design, marketing, and economic issues. (Microsoft's Office 2000 and Back Office subscription pricing has already managed to annoy users who think it's just a ploy to extort more dollars for the same old products.) Ultimately, the real success stories in subscription-based services are likely to be brand-new service products that developers build from the ground up, not just old products that now run on a remote server.

So what does it take to build a subscription-based service? Here's our quick checklist of success factors:

* Is the concept idiot-proof? Since services are so intangible, it's easy for customers to get confused about exactly what they're buying (or describing to a colleague). So it's essential to distill the essence of the service into two or three explicit words--'payroll processing,' `anti-virus protection,' `expense account management,' and the like. The service product itself may have some fairly complex components (we all buy `annual engine tuneups' without thinking too much about spark plugs and oil changes). But if there's any ambiguity about the core concept, prospects will usually react by deferring a purchase decision, usually forever.

* Is the pricing model transparent? Figuring out how to price a service can be tough. But there's still no excuse for mystery prices or for what Larry Ellison calls `Turkish bazaar' pricing formulas that baffle even the sales force. The truth is, customers rarely quibble over service prices. If the price tag seems reasonable and easy to understand, most buyers will quickly move on to other issues. (As AT&T has shown with its new `Digital One Rate' service, a simplified price by itself can be a breakthrough sales proposition.)

* Is there an existing sales channel? This question catches lots of companies by surprise. We've been looking at a wide range of high- and low-tech service businesses, and it turns out that true resellers are remarkably rare anywhere in the service world. Instead, vendors tend to set up their own field sales forces, or else they rely on franchises, partnering arrangements, and telesales groups. To be sure, there are interesting theories about how VARs and ISPs will soon become enthusiastic resellers of technology services. But we're not there yet.

* What's the likely renewal rate? Experienced service companies religiously track one key indicator: contract renewal rates. That's because renewal rates have tremendous leverage on a customer's `lifetime' revenue contribution, and margins for renewal services are typically far higher than for first-time customers. If the renewal rate for a service isn't up in the 60%-80% range, too many customers are jumping ship and the service probably needs a complete overhaul.

* Does service start with a bang? Services are typically an extended experience, and over time many buyers literally forget they've become customers (especially when the renewal notice shows up). So it's important to give first- time service customers something tangible up front--books, software, binders, certificates, membership cards--and, even better, to touch base regularly with post-sale phone calls and e-mails.

* Can the customer cancel at will? From the buyer's perspective, the subscription model is often a way to reduce the risk of a large up-front purchase. Rather than play up cancellation rights as a selling point, though, many dim-witted vendors try to lock service subscribers into escape-proof contracts. In reality, hardly anyone cancels a service contract for frivolous reasons. Guarantees, escape clauses, and trial periods make customers more comfortable and shorten the sales cycle.

* Can you keep your service metrics honest? Magazine reviews and competitive shootouts keep product companies from too much self-deception. Service companies get much less feedback, and often their customer surveys come up with myopic, self-serving performance metrics (the same kind of data that convinces airline executives that we enjoy cramped seats and bad food). Start with focus groups and open-ended surveys to find out what customers think is important. Then make sure the whole organization sees the latest satisfaction scores.

* Can the billing system handle recurring charges? This sounds like a trivial issue, but billing snafus regularly torpedo otherwise competent product companies (including Microsoft) that thought it should be easy to cram thousands of small monthly payments through their accounting system. If a company can't straighten out its billing, most customers won't care about service quality-- they'll tell the world that the company is run by jerks and con artists.