I recently came across an IDC report from early 2010 that predicted that the SaaS BI market will experience triple the growth of the BI market overall. As reported by InfoWorld, IDC projects a compound annual growth rate of 22.4 percent through 2013, although actual SaaS BI revenue totals will remain small compared to on-premise BI applications (emphasis added). Having conducted our own survey of SaaS BI adoption, having worked with clients who were considering SaaS BI, and having talked with leading SaaS BI vendors, I thought it might be helpful to share some ideas that may prove useful to companies who are considering SaaS BI – also known as On-Demand BI.
Innovation in Business Intelligence and Business Analytics
The history of IT is one of continuous innovation, some of which has led to spectacular successes. On the other hand, some IT innovations don’t really live up to projections, and some new developments in IT outpace businesses’ ability to assimilate the latest advances. Accordingly, it is not unreasonable to ask if SaaS BI is a flash in the pan or the killer app that gets BI across the chasm to become the “pervasive BI” people have been pitching for years. And as with any new IT product or service, it is not unreasonable to ask whether it is a fit for your company and whether to move now or wait for things to shake out a bit. We’ll share our approach to vetting SaaS BI in a moment, but first let’s briefly survey the context in which SaaS BI has arisen.
Value Propositions for SaaS Business Intelligence
There is no doubt that Software-as-a-Service (SaaS) has been successful in some applications. Think Salesforce.com (CRM), Mint.com (personal finance), and Google Docs (Office Suite) as examples. And there is no doubt that BI and data warehousing have been challenging to many companies – with the general business complaint being that BI/DW initiatives take too long, cost too much, and don’t deliver as promised. Setting aside the argument that ERP initiatives cost way more, were far more wrenching from a process change perspective, and often failed to deliver sustained measureable improvements to profit margins, it is true that BI/DW initiatives have challenged both IT departments and the business units they serve. The reasons for sub-optimal performance are many, some of which are detailed in our blog post – Agile Business Intelligence as a BI Release Strategy.
In this context, and given the technical ability to do things in the Cloud, it is not surprising that the SaaS model has now been applied to BI. There are a number of potentially compelling value propositions offered by proponents of SaaS BI, including:
- faster implementation
- lower costs
- lower technical risk
- simplified access to information and business analytics
- access to pre-defined business analytics
For companies who have struggled with getting value from a data warehouse and/or with successfully deploying business intelligence applications and business analytics, these value propositions may be especially attractive. On the other hand, companies who are good at business intelligence, data integration, and data warehousing may find the risks associated with SaaS BI might outweigh the benefits. Accordingly, companies considering SaaS BI need to take a holistic look to sort out whether it is a fit for their needs.
How to Evaluate SaaS Business Intelligence
The typical CIO and BI Director are under increased pressure to deliver more BI applications more quickly and less expensively, and thus SaaS BI becomes an approach that demands consideration. And the typical functional head in major companies has at least some material gaps in the availability of business intelligence and business analytics to help achieve business performance objectives. Knowing this, SaaS BI vendors are specifically targeting marketing messages to business leaders, who are often frustrated with the pace of BI delivery. The coverage of SaaS BI by well-regarded (albeit technology platform focused) analysts adds to the expectations of executives looking for a silver bullet solution for their BI challenges. Understandably, in many cases the temptation to jump into SaaS BI is powerful. That being said, it is still early days, as is evident from looking at the customer lists of leading SaaS BI players and from the fact that they don’t tout revenue numbers and growth. With this in mind, we recommend the following three steps for vetting SaaS BI.
1. Evaluate the strategic fit between your business strategy, BI strategy, and SaaS business intelligence.
Ideally, your BI strategy should be driven by the company business strategy. How does your company compete, and how effective are the key business processes through which you achieve your business strategy? With that in mind, your BI strategy should prioritize BI applications and business analytics that improve business performance. And your BI strategy needs to be consistent with your overall IT strategy. Is your IT strategy based on heavy outsourcing, and/or are key transactional applications already in the Cloud? Or is your company a do-it-yourself company with respect to IT? Further, how does your company feel about the possibility of having its critical business data under the day-to-day control of an outside party?
2. Evaluate specific SaaS BI offerings in relation to the top-priority BI requirements established by your BI strategy.
Assuming SaaS BI is a fit with your IT strategy and business strategy, it then makes sense to vet the degree to which SaaS BI offerings meet your prioritized needs for business information and business analytics. For example, say that the top priority BI requirement is for marketing analytics, including the ability to enable behavioral segmentation and rapid determination of trade promotion effectiveness. By definition, most SaaS BI applications are pre-designed, so we have to map our BI requirements against the data model underlying the SaaS BI application. If there is a good match, which there certainly could be, then SaaS BI could be a go for meeting the BI requirement for marketing analytics. Before pulling the trigger, however, it is important to assess a number of business and technical factors, examples of which are provided below.
3. Evaluate SaaS vendors in relation to relevant business and technical considerations.
SaaS BI is relatively new, and so are the leading vendors. Accordingly, it makes sense to evaluate the financial strength and business model of SaaS vendors. If BI is important to your company’s ability to execute its business strategy, then we have to know that the vendor we select is strong. We also have to know how the SaaS BI is actually delivered. For example, some SaaS BI vendors depend on Infrastructure-as-a-Service (IaaS) and Platform-as-a-Service (PaaS) vendors to deliver their SaaS BI offerings. This creates a multi-tier model and opens the possibility of finger-pointing among the various parties if something goes wrong. And on the technical side, our survey of SaaS BI adopters revealed these common challenges: scalability, the need to do a lot of up-front data integration work, the need to adapt the vendor’s data model, and the need to develop non-standard reports in order to meet the totality of business users needs. All these business and technical factors need to be evaluated before signing a deal.
SaaS BI could be a key tool in your business intelligence and business analytics toolkit. A straightforward due diligence process incorporating the steps above will help avoid the risks that attend any new IT approach. What companies we have talked with have found is that before they can move forward with SaaS BI, they need to address some BI fundamentals, like creating a BI and technical strategy. Once those fundamentals are in place, companies have a fundamentally sound context for determining whether SaaS BI is a fit for their business and technical direction