Saturday, October 17, 2009

Future applications: SaaS or traditional?

After hearing a podcast of Dana Gardner's interesting interviewing Aneel Bhusri Workday CEO,
I tried to figure out what went wrong in the transition from traditional applications suites composed of large components to Service Oriented based application suites.

In 2003 I had to present Giga Information Group's view in a local ERP conference in Haifa.
A presentation by Giga's Analyst (and afterwards Forrester Research Analyst) Byron Miller titled: "ERP Technology Is Going Mainstream" was used for that purpose.
The presentation depicts three possible future ERP states:
Suite State – proprietary vendors suits.
Industry or Vendor Ecosystems – more evolved standard based suits.
New B-O-B World – standards based Best of Breed state. Use of ERP components from different suits is possible because of the standardization.

New B-O-B World is the most desired state; however according to Giga's prediction the most probable future state is Suite State.

I am not sure if SAP's and Oracle's initiatives to transform their ERP suites to Service Oriented suites are attempts to move to Industry or Vendor Ecosystems state or to New B-O-B World (I suspect that their goal is to build a Service Oriented Ecosystems and not a New B-O-B World), but so far the initiatives are far from being successful.

One of the best ways to save money during the recession period is to delay ERP upgrades. Delaying ERP upgrades could save millions of dollars. In a Case Study presented in Gartner Group's Architecture Webinar, an organization saved 2M$, by delaying an ERP upgrade for 3 years.
The reasons to postpone an ERP upgrade during Recession are obvious: It is not strategic and it is expansive.

Compare these upgrades with Web 2.0 or SaaS applications upgrade with no user responsibility, resources and efforts. Aneel Bhusri talked about a version upgrade of Workday Hr application about once in 3 or 4 months. Google as well as other Web 2.0 vendors do not even have versions for their SaaS applications at all (see my post Web 2.o for Dummies: Part 4: What is Web 2.0?).

I do think that ERP and other Core Applications vendors have no choice: In order to be Agile, Future applications should be Service based applications, but how would these applications be transformed to services.? The hard way is by decomposing current Component based ERP systems to services by Fusion project, partially Java rewrite of ABAP applications and similar initiatives. The easy way is to start fresh Service based single ERP or CRM application such as, Workday or Netsuite and extends it gradually.

The other side of the coin is the functional gap between current ERP and the SaaS based solutions, but this gap could be closed in the long run.


Pete said...

Excellent comment on ebizQ's Forum. To join officially, which means you'll get the questions emailed directly to you, please email

Sreekumar J said...

I think SaaS will be considered as the most preferred software delivery model for small business.

Mid sized companies and large scale outsourced service providers may try SaaS tools but there won't be a complete shift from installed software and legacy solutions to SaaS, 5 years down the line

Avi Rosenthal said...

Sreekumar, thanks for your comment.
I do not think that large enterprises applications will be completely SaaS based in foreseeable future. There will be always some unique or proprietary systems, so there is no disagreement between yours opinion and mine. However, as far as Mid size companies are concerned, I am not sure that SaaS applications will not be their most important paradigm. It depends on the definition of Mid Size companies. Some people's Mid Size definition could include companies Enterprises.

Public Cloud Core Banking: Hype or Reality? - Revisited

  More than 4 years ago I was asked if Public Cloud Core Banking is a Hype or a Short Term Reality? If you had read the post, you would prob...