Daily Archives: July 7, 2006

On Demand IV: And the SaaS Story Continues

Shortly after I finished writing my three part post on On-Demand over at e-Sourcing Forum (The Good, The Not-So-Bad, And the Coming Pretty …), I stumbled across this great article by IQ Navigator‘s John F. Martin (of Building SaaS fame) called How True Software-as-a-service Delivers More Value over at Supply & Demand Chain Executive.

In this article, he echoes many of the same points that I have attempted to make in my series of posts, and does so brilliantly. He also clearly emphasizes some of the significant disadvantages of legacy enterprise software in relation to True Software-as-a-Service and gives you some rules for identifying a legacy provider with an ASP model trying to disguise themselves as a software-as-a-service provider.

He emphasizes the following five significant advantages of true SaaS solutions as compared to legacy applications that I believe just cannot be stressed enough:

  • no significant technology investments
    with legacy enterprise application solutions, you have to shell out for significant technology infrastructures to support them; with on-demand, all you need is the PC already on your users’ desks
  • no assembly required
    legacy applications require customers to become technical experts in the software: installation, infrastructure configuration, integration, customization, issue diagnosis, and upgrades; this involves significant training, ramp up time, and paying for a large IT staff indefinitely
  • no lock downs
    once customized legacy software finally goes live, it becomes a “strait jacket” that prevents future innovation and improvements for 3-5 more years (as upgrades are deferred as long as possible due to significant costs to re-implement new versions)
  • speedy issue resolution
    with a legacy enterprise application, the software vendor must often replicate a customer’s unique environment (exact production versions of application, database, operating system, hardware drivers, etc.), which can take weeks, only to determine that the issue can not be diagnosed or fixed because (1) it’s due to a customization (2) it’s the responsibility of another software component provider or (3) the version of the software you are running is too old
  • true process expertise
    true SaaS software vendors are experts in the business processes they automate; they can provide ongoing assistance in using new software capabilities, process innovations, and best-practices

And when you consider that Gartner estimates that more then 70% of the total five year cost of ownership for enterprise software comes after implementation, when you consider on-demand is usually significantly cheaper then enterprise software to begin with, this is a powerful proposition.

He also indicates that you can often tell a legacy application on ASP in disguise by noting one or more of the following indicators:

  • a reluctance to pilot
    pilots represent a significant investment to a legacy application provider who will have to configure a new instance just for you; in contrast, a true SaaS solution provider can enter your name and flick a software switch and the pilot is immediately set to go
  • elephant hunting
    (where the salesperson tries to enlarge the deal as much as possible to maximize revenue before you discover the true benefit/cost ratio of owning the software); in contrast, a true SaaS provider will allow you to buy the bare minimum knowing that you’ll want to buy more when you discover how great the service really is
  • reluctance or refusal to discuss revenue
    or the question “what would happen to your profitability if you had no new customers over the coming 12 months?” this would cause a traditional provider serious grief and lead to significant downsizing; in contrast, an established SaaS provider would be able to maintain status quo
  • infrequent or new functionality in any given year
    in contrast, most true on-demand SaaS providers provide regular updates 3 or 4 times a year
  • hosted versions lag new releases
    in contrast, a true SaaS solution is always up to date
  • insistence on single tenant or multiple instances
    in contrast, a true SaaS provider will want to take advantage of the multi-tenant model to save you both $$$
  • delayed or long implementations
    a provider running a legacy application on ASP may require weeks or months before they can get you up and running; a true SaaS solution can literally turn you on the same day you cut a deal
  • simple customizations require single tenant instances
    in contrast, true SaaS implementations are usually built to contain a moderate amount of configurability from the ground up
  • end of contract unknowns
    with legacy in disguise, you never know if you’ll be able to renew, how much it will cost, if you can get your data out, etc.; in contrast, true on-demand SaaS providers will specify everything for you up front, usually in the contract

Remember, as John F. Martin says, “SaaS allows customers to focus on their core competencies and their business processes rather than becoming experts on software internals, technology infrastructure maintenance, or deployment methodologies.

Yet another take on SaaS can be found in Robert Bois’ recent AMR article The Rush to SaaS: Making Sense of the New Wild Wild West, where he presents The SaaS buyer’s guide. In it he presents three questions that the buyer should answer before selecting a solution.

  • What’s in an architecture?
    Mr. Bois points out that from a buyer’s perspective, that single tenancy vs. multi-tenancy should not be as big a concern as the pricing model, SLA, and customer support. Although I will admit that the SLA and customer support issues should be tops, I do not entirely agree with the first point. If price is important, then a multi-tenancy model should save both parties money. Furthermore, as Sudy Bharadwaj of Aberdeen eludes in his take of On-Demand Supply Management at e-Sourcing Forum, accepting a single tenancy model may cause you to miss out on the community benefits of a multi-tenancy model.
  • What’s the real TCO?
    This is probably the most important question a buyer should ask. If the TCO of an on-demand solution is high, I’d be willing to wager that there is a good chance that what you are actually being offered is a legacy ASP application in disguise, at which point you should refer to Mr. Martin’s indicators to find out for sure. I like Mr. Bois’ cost table, but should point out it is only accurate for the lifetime of your initial hardware and software purchases. The table seems to indicate that there are no annual subscription/license, hardware, or middleware/db license costs after the first year. This is not necessarily the case. If you want upgrades, you will have to pay a maintenance fee. Middleware providers only support their releases for a fixed time frame, and if you do not upgrade on a reasonable cycle, you will find yourself running an unsupported product, which will cost you a fortune if something fails and you need it fixed. Finally, even the best hardware will not last you more then three years without an upgrade.
  • To customize or not to customize? Mr. Bois makes a really good point here: “many companies believe their business processes are more unique than they really are, and they should weigh the advantages to more custom coded logic against the lower maintenance and upgrade costs of using business process tools and configuration instead of customization“.

Finally, I’d like to again point out Sudy Bharadwaj’s Six Step Framework for On Demand Supply Management. (login required) Part of Aberdeen Group’s Enterprise Strategies: Insight and Advice for Enterprise Executives, it overviews the Aberdeen PROFIT Framework for Supply Management as a Service (SMaaS). The PROFIT Framework was designed “to aid supply management evaluators in understanding how to deploy an On Demand solution to drive value“. It provides a series of Process, Regulatory, Operational, Financial, Intelligence, and Technology questions whose answers are designed to help you make the right decision.