What's going to happen to the application service provider market?
The first thing already has happened: the "e-diots" responsible for over-hyping ASP opportunities have been stepping all over each other to be first to eulogize the ASP approach. After all, even at a deathwatch, there's money to be made--especially if the transition from obstetrician to mortician is no more difficult than changing from white gown to black suit.
But most of the "death of the ASP" yapping that I've seen seems no better informed than the nonsense that made the ASP market, for a while, a hotter property than Russell Crowe. It was silly to think that the ASP market would adhere to the "new" Internet economics; and it's just as silly to be disappointed that it hasn't followed a meteoric adoption curve.
Boom products and services over the past few years almost all shared a common economic trait: their benefit to individual users grew with the size of the user network they embodied. Take instant messaging. The more folks I can engage with an instant messaging technology, the more valuable the technology is to me (unless I want to avoid being "IM'd" by others).
But "increasing returns", as this trait is called, is not an obvious feature of many ASP instances. Most traditional applications--say, a general ledger--are highly segmented. There's a clear limit to the number of folk I want connecting to my general ledger, thank you.
In other words, just because an application service can flow over the Internet, doesn't mean that it automatically accrues Internet-economy-type benefits.
The long-term prospects of the broad ASP market, however, hinges not on benefits, per se, but rather on cost advantages. Many large software suppliers, sick of the costs of managing an increasingly complex customer base, now tout the cost advantages that users could accrue by not owning their key business systems.
Clearly, there are at least some cost differences to be considered. If I don't own the servers, storage, system software, network, and application license necessary to run a business system, I can avoid a bunch of out-of-pocket expenses or risky asset acquisitions. And, if I can let someone else operate the system, I can avoid a bunch of people costs.
But ownership has its benefits. A home owner can wake up one morning and decide to remodel. A renter has to change addresses.
Application systems must be constantly enhanced and integrated--remodeled--to ensure that a business's behavior can be kept up-to-date. A system that is maladapted to competitive realities is a dangerous system indeed.
And this is the crux of the problem that must be overcome by the ASP market: how can an application service be delivered in common to a broad base of customers (to maximize seller profits), yet applied uniquely (to maximize buyer returns)?
I foresee four forms of economic organization evolving in response to this problem. The first is that ASPs lie down and expire. Some, of course, will. Quite frankly, a bunch should. A sizable portion of business systems just aren't suited to an ASP treatment.
But I'll hazard a guess here: within 18 months, the same e-diots denying that they ever forecast huge ASP market growth will start telling us that they were right all along--although their market growth estimates will be about a third of what they had been.
The second is that application suppliers in markets where broad-scale public connectivity is the source of application value morph into ASPs. The issue here is that service suppliers will tend to bundle "ASP-ready" offerings together, driving smaller players out. A few important application domains will become dominated by a few very large ASPs.
The third is that ASPs essentially become business partners for the large IT consulting firms, funneling clients whom must pay for periodic integration work to handle the problems of maladaption. The problem is that ASPs become like legitimate fronts for organized crime: the real money ends up in the hands of the integrators, which limits the stream of capital available for ASP business innovation.
The fourth is that ASP value proposition shifts from function shipped to function consumed. This is anathema to today's software industry, where most vendors try to jam as much esoteric function into their package as possible.
But really successful, innovative ASPs ultimately will be those that design offerings around the depth to which their service can be easily integrated into a client's business--and not the number of checkmarks on a product sheet.
If the ASP approach becomes broadly successful it will do so not because ASPs "Web-enable" a bunch of crummy existing software, but because they start driving new rules for software quality, integratability, usability, manageability, packaging, pricing, and customer service in the industry generally.
This is a tall request, but we'll see if the ASP market ultimately can stand and deliver. I have never been an ASP zealot, yet I am cautiously optimistic.
Peter Burris is an industry analyst and Meta Group research fellow.