image Source: Wikipedia, the free encyclopediaThe headline depicts my impression from a Research Note written by Burton Group's analyst, Craig Roth. The Research Note describes and analyzes the results of a new SaaS survey conducted by Burton Group and Ziff Davis Enterprise.
It is not just a survey. Some of the participants were asked to explain their answers. The explanations provide another layer of understanding: Why (explanations) in addition to What (the answers).
Some of the findings are in accord with my expectations (I assume that my expectations are not very different from expectations of other people who read articles on SaaS), but there were also findings I would not expect to find.
It should be noted that all the 318 participants reside in
USA and , therefore extrapolating the conclusions to other parts of the globe could be misleading.
Reasons for implementing SaaS
The obvious reason is cost reduction, but a lower cost was not the most frequent reason for interest in SaaS applications. The most related reason was "No in-house Maintenance", which could be positively correlated with lowering costs.
Surprisingly, the second, third and forth frequent answers are more related to Agility than to lowering costs (Shorter roll out time, Usable anywhere via internet, Faster improvement cycles).
Reasons for avoiding of SaaS implementations
As expected the leading reason for not implementing SaaS was Security risks. However, those who implemented SaaS experienced rarely Security breaches occurrences.
The second frequent reason was Unavailability in case of inability to connect by the internet.
and the third frequent reason (less customizable) also fit to my expectation.
The surprising findings are the low frequency of citing "Less functionality". I would also expect to find integration between SaaS and
applications to be cited frequently. However, that was not what Burton Group and Zif Davis Enterprise found. Data Center
The most popular SaaS applications are Web Conferencing and CRM. CRM is expected to be used frequently due to SalesForce.com effect. As expected, applications used for SaaS are applications which are standalone (e.g. HR, Surveys, e-mail) and are less used as connected systems.
Least cited applications were ERP applications.
I expected that the preferred charging model would be related to usage and or capacity, but actually most of the participants preferred Flat Fee Enterprise Model.
Probably, this model would be more expensive than charging by Usage and/or Capacity models.
The following bullets include my view of the surrey's results:
- Agility is a leading reason for selecting SaaS applications. This finding illustrates the similarity between SaaS and SOA, which is driven by the quest for Agility.
- The old Service Bureau charging models of the 60s are still valid. Service Bureau Customers preferred a simple and stable charging model over a more accurate and complex charging model. I do think that more flexible models are better fit for Cloud Computing. It looks as if SaaS consumers are not yet able to adapt Out of the Box thinking and therefore stick to the Data Center Enterprise License model.
- The issue of integration between local and SaaS applications was lacking from the list of reasons to avoid SaaS usage. This issue matters more to large Enterprises and especially for Business Intelligence purposes.
Why is this reason lacking from the list?
Is it because SaaS implementations are not yet mature?
Is this reason (Integration), which was not cited, at list a partial explanation of more benefits realization
by SMBs than by Large Enterprises?
- Types of applications implemented by SaaS model is a key issue. Core Systems implementations is an indicator of SaaS Maturity. Burton Group's analyst thinks it is not yet matured. The low rate of ERP usage sustains his view. NetSuite was the first company to offer SaaS based ERP. Other companies will probably follow NetSuite.
- Even though, SaaS is not matured it is going Mainstream. The majority of enterprises deploy SaaS applications. Mainstream technologies should not be confused with Mature technologies. Mature technologies could become non Mainstream e.g. pre-relational databases like IMS or Adabas. A very good example of Immature Mainstream technology is Enterprise Client/Server Integrated Development Environments (IDEs) in the middle of the nineties.
- Burton Group's analyst Craig Roth concludes that SaaS is immature. The immaturity is manifested in the type of applications used frequently in SaaS model. Maturity will be manifested by more frequent use of SaaS model for Core Applications such as ERP.