SaaS: Shelfware as a service?

SaaS: Shelfware as a service?

Summary: Software as a service is portrayed to be the future of information technology, but it isn't quite the cure-all it's cracked up to be. Shelfware and lock-in may not save you a lot of money.


Software as a service is portrayed to be the future of information technology, but it isn't quite the cure-all it's cracked up to be. Shelfware and lock-in may not save you a lot of money.

That's the message from Gartner analyst Rob DeSisto.

Here's DeSisto's talk at the Gartner IT Symposium in a nutshell:

The points from DeSisto are plentiful and are designed to be contrarian. Among them:

  • "Many of the bad practices that occurred in the on-premises world are also now moving their way into SaaS. The biggest example is shelfware. An early promise of SaaS is that you would pay for what you need, where in fact we see many companies over buying subscriptions leading to the new phenomenon of shelfware as a service. Shelfware as a service is analogous to on-premises shelfware but only now companies are renting the shelf from the provider instead of using their own data center."
  • No SaaS provider has come close to replicating's success.
  • Vendors claim SaaS can be quicker to implement, save money and produce more rapid innovation. However, it's not that simple, according to DeSisto. Unless you can reduce or avoid hiring software headcount, skip purchasing hardware and forgo buying database licenses you may not save all that much with SaaS.
  • SaaS forced upgrades may pose compatibility issues and not be absorbed the enterprise users.
  • Ninety percent of SaaS offering are not "pay for use," but set contracts.

Of those points, the shelfware issue is the most interesting. Does it really matter if you buy or rent shelfware? It's still a waste. DeSisto writes:

Shelfware as a service is the concept of paying for a software subscription that is not being accessed by an end user. This most commonly occurs in large enterprises, but it could occur in any company. Shelfware as a service also happens when you're forced to downsize usage. For example, if you cut your workforce by 25%, then you no longer need the subscriptions for those employees. Whether you're phasing in or phasing out, you will unfortunately be part of this new phenomenon of shelfware as a service. Not only are you paying for using the software, you're also paying for not using it.

And here are a few numbers:

The main point: The verdict is mixed on SaaS thus far. The positives are relatively clear: SaaS fits in the operational budget, cuts infrastructure and management overhead, cuts short-term cost of ownership, can be implemented quicker and promise more innovation.

On the downside, SaaS can lead to shelfware, lead to application portfolio management issues, lead to unwanted upgrades, can be a security issue and clash with on-premise software.

Overall, companies need to have a system to evaluate SaaS providers and management techniques. If your governance is lacking you'll have problems whether you buy or rent software.

Topics: Data Centers, Cloud, Emerging Tech

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Think the conclusions are a bit flawed

    I think we can all agree that SaaS saves operational expenses, increases agility and improves the user experience. I don't think the fact that has been successful can be held against the category.

    The shelfware point is a bit more subtle. The fact may be that shelfware can happen with SaaS. However, the contracts are of shorter length so can be renegotiated sooner, actual usage is transparent and easily measured, and the level of waste is orders of magnitude lower than with on-premise SW like SAP or Oracle.

    I think Rob's point is that SaaS does not eliminate some traditional problems and still requires management and governance processes. Not that SaaS creates new problems as this article seems to imply.
    • Actually...

      The post does not seem to imply that new problems are created, but simply that SaaS does not get rid of some of the oldest on-premise issues, as promised by current SaaS prodviders.

      I don't think either that the success of can be held against SaaS, the point being made though is if you look at the market as a whole, if something happened to, then that would drop the overall SaaS use by a big %, and actually possibly hurt movement to SaaS options.

      This is why I like Microsoft's approach, which is: Software + Services. That services and software include Windows Azure platform as part of that planning, for running, developing and even hosting those software + services concepts.

      Still, even with this more hybrid approach and large range of possibilities for how much software is on premise, and how much services are on premise vs. out in the cloud, still you can run into the same issues.

      To wrap this up, the point was to say SaaS has the same issues as trad. software. And BNara, you did get it right when you said this.:

      "I think Rob's point is that SaaS does not eliminate some traditional problems and still requires management and governance processes."
    • Yep

      The title is misleading. Gartner doesn't contend that SaaS introduces shelfware into an otherwise well-run shop. Point is, if your governance of on-premises software is inadequate, transferring poor practices to SaaS only redisrtibutes - or exacerbates - the problem. But a more sensationalist headline generates more hits, doesn't it?
  • RE: SaaS: Shelfware as a service?

    It doesn't seem like an issue of technology but of mismanagement from the SAAS buyers. However, lack of SAAS as a pay-by-use is an interesting point.

    R. Paul Singh
  • Typical mainstream analyst obfuscation

    This problem can be solved by SaaS much easier than with conventional software. Shelfware exists because of the buying practices and poor governance of enterprise IT, encouraged by the established software vendors and their analyst brethren.

    Most enterprise IT buyers insist on multi-year subscription contracts with SaaS providers, even when the SaaS provider is perfectly willing to bill monthly based on actual numbers of subscribers. The buyers want to pin down their budget for the year, so the flexibility of SaaS makes them very nervous. (Also, their legacy ERP systems can't cope with so many variables).

    So here's an analyst picking up a problem *created* and *perpetuated* by non-SaaS vendors and chalks it up as a reason not to buy SaaS.

    By the way, as his slide helpfully reminds us, the industry from which Gartner earns the vast majority of its revenue is still 90+% non-SaaS.

    phil wainewright
  • Need Better Manageability Tools

    To avoid over-provisioning of software subscriptions, we need better tools from all of the SaaS providers to track actual usage and easily adjust the number of subscriptions through self provisioning. This isn't a downside of SaaS, it's merely an opportunity for improvement that can lead to higher comfort levels, promoting wider usage.
  • RE: SaaS: Shelfware as a service?

    For me its not the shelfware thing, though there have been a lot of recent discussions around SaaS license costs, e.g.

    For me, it is this point that is the good one:

    * Vendors claim SaaS can be quicker to implement, save money and produce more rapid innovation.

    In my opinion the savings don't come alone from reducing headcount or hardware expenditure. They come from the fact that SaaS solutions are often designed 'differently' from their enterprise equivalents. Due to the nature of the beast, most SaaS solutions have to be ready to run without the developers and sysadmins having to get involved. This makes the solutions more configurable (less software development risk), but also means that they often codify the key, common and best practice ways that people need to work, without having to develop that from scratch.The time to go-live is much reduced, meaning that the ROI is likely to be ahead by months already, even before the reduction in cost of big software development projects.

    What you may not get with SaaS is every bell and whistle of an enterprise product. But the bells and the whistles have been put in enterprise software to appeal to Gartner, not because there is a proven market need for them. Less feature noise means reduced fees, licenses or whatever you call them, and typically a more reliable product. Yes, reliability and security becomes a big focus when the same company developing the software is the one that has to nurse it through 24 hours a day.

    So, is there a fear that SaaS is going to put inflated enterprise software license costs under pressure? Or the fear that SaaS commoditizes certain segments of the software market, so that the analysts don't offer their value? Or is it just the thing to distract us from another Windows cash-cow release?

    Phil Ayres
  • Cant get as bad as traditional shelfware

    Even though the problem of SaaS shelfware exists, it can
    never get as bad as traditional shelfware problem. This
    is because every month you have to shell out rental
    payments for your SaaS solution, and is hence a
    reassessment/wakeup call every month. With on premise
    software, you've bought it once, the costs are locked in,
    and theres nothing you can do about it.

  • shelfware free application develompent?

    Its a slogan I am thinking of using for our development toolset. Yes. this is a blatant marketing ploy but to date our customers produce only usable applications that grow with their needs. We have no shelfware being notused. Or.. i mean. we have no shelfware on the shelf. Something like that.
    EASYProcess at