Software as a Service Business Intelligence to the Rescue
It's precisely in today's economy that SaaS BI, with its promise of rapid and inexpensive deployments shines
- By Stephen Swoyer
- February 18, 2009
With uncertainty in abundance and with organizations placing a premium on ROI -- especially when it comes to getting the greatest value from their existing investments -- the timing has never been better for software-as-a-service (SaaS) BI, argues Anil Chitkara, senior vice president of sales and marketing with SaaS BI player Oco Inc.
In such uncertain economic times, SaaS BI's promise of rapid -- and comparatively inexpensive -- deployments shines, Chitkara argues.
"What we've tried to do with customers is be very focused on both payback and ROI. Our claim to fame is that we get up and running very quickly, typically in six to 10 weeks," he says. "The challenge I see in the BI space is that a lot of people make that claim, so you have to get a few layers deep to figure out where they're getting their numbers. We have customers who have publicly talked about getting a 30-day ROI after their deployments."
Oco has a lower ROI hurdle to clear than its on-premises competitors, Chitkara argues. "A customer can get a solution in place for a thing like inventory, and it could cost them $100,000 to $120,000 a year [for an on-premises product]. But if they can get something in place inexpensively -- that's the deployed solution, including all the support, including the maintenance, including everything else -- that's very attractive to them, especially today," he comments, citing the case of juice and soft-drink purveyor Welch's. "They spend about $50 million annually on their transportation [costs], and they had all their information for their data on transportation … in three different systems, one of which is Oracle and two are legacy systems, but they couldn't get the visibility to be exactly what they needed it to be," Chitkara continues. "In that case, we pulled resources together … so that they could increase the utilization of how full these trucks were and [also] change some of their [internal and external operations] to basically get a rapid ROI: eight weeks to deploy and then thirty days to ROI."
Oco doesn't pitch itself as an all-things-BI-to-all-people kind of player, either. It has strengths and expertise in key verticals -- particularly in manufacturing -- and it's in this segment that it's enjoyed its greatest success, Chitkara says. "We have modules focused on the operational side of manufacturing, things like fixed asset management, service and performance for those companies that have services businesses, then … modules for retail and direct channel [sales]," he explains.
Nor does Oco bill itself as a turnkey BI. ROI and value are to be had in every deploymet, Chitkara stresses -- but they aren't always going to be easy to get at. Instead, he says, ROI is buried in multiple data sources, lurking -- to use another metaphor -- in the deepest, darkest recesses of data warehouse-dom. It's a prescription that -- although specific to Oco's SaaS BI methodology -- can and probably should be applied to any BI effort, Chitkara concedes.
"What we're constantly doing is trying to find more advanced analytics. For us, 'advanced analytics' is not a tool to do data mining or a tool to do predictive analytics," he explains. "It isn't a tool of any kind. [It's more an issue of] let's combine different sources of data, operational and financial data, for example, and come up with some new analytics that are going to really help you find where the value is sitting in that dark corner of the warehouse."
The key is to separate potentially big-ticket ROI projects from smaller (incremental) projects, according to Chitkara. To do that, you first have to understand where your data is, what it's supposed to represent, and how it can be meaningfully combined or synthesized with data from other sources.
"What we do is we spend two days with our clients just to understand all of the business aspects of reporting: the data that's going to drive that reporting and then the dimensions of the data in the system. After about two days, we understand where the data is [and] what needs to be pulled together for the reports. Then we take a data dump from them [using] Oco Connect, our data discovery and identification tool, and that helps us understand where in the data is the information [that] we need to pull into the warehouse and populate the reports."
It's an iterative process, Chitkara continues. "Typically, it takes two to three generations to get the data right, so there is a combination of a sort of proprietary process and methodology that we use -- an example of which is our [data] profiling, [our] connectivity technology -- and then working with the folks," he says. "Rarely do we see data that is always clean. That never happens."
Big business intelligence and data integration (DI) players tout an enterprise-wide approach on BI and DI. Oco's is a more tactical approach, according to Chitkara. "We say, 'Let's look at the data that comes in and see how business-relevant it is.' There might be products -- a whole host of products -- that make up one percent of the revenue. Fixing the data for those [products] is not going to move the needle for the problems they need to solve. Instead, let's look at the '80/20' -- the smaller set of products that drive most of the revenue -- and make sure that the data associated with those is clean and good and accurate."
It's a vision that Chitkara claims dovetails nicely with an emerging trend: the importance of extracting value from existing investments. In the past, he says, customers might've been willing to cut infrastructure investments some slack, giving them extra time to mature or otherwise come into their own. In the current climate, however, shops are battening down the hatches -- in some cases freezing planned infrastructure expenditures -- and are instead focusing like a laser on squeezing value out of existing investments.
"There are still a lot of it folks who would like to stay with the major investments they've standardized on. There's almost a reticence to leave some of those tools. There's still a major allegiance to those pieces of software that are in place. I think that what we're finding is that there's such pressure on IT to support [these existing] business initiatives, and that that's causing them to look at other solutions," he comments.
"Even with the sort of Big Three [i.e., IBM Corp., Oracle Corp., and SAP] in place, there are other areas where companies want targeted solutions," Chitkara points out. "BI is not a place for one standard set of tools. I think BI is too complicated and the needs are too different to just have one tool to solve everything. I do think that we're going to see [companies embracing] some of these smaller, niche-oriented tools to solve their problems."