A Nucleus Research report about how Oracle and SAP fare among small and mid-sized enterprises has sparked a dust-up between the two--again.
It's no surprise that the Nucleus report has raised a bit of a ruckus given how both companies are targeting small and medium-sized companies to fuel future growth. And this isn't the first time Nucleus has dinged SAP. I can see the Nucleus report being used in an Oracle ad any day now. And now it's a shareholder issue as the Nucleus report anchored a Wall Street Journal article today (subscription required). SAP called the Nucleus report junk science in the story.
The reality is a bit more nuanced than that. I read a copy of the report and there are many key items worth discussion. Indeed, Nucleus only interviewed 56 customers (27 from Oracle and 29 from SAP). I'm not sure that sample size equates to SAP's junk science comment, but perception--especially customer perception--matters. Perhaps a bigger sample size would even up the percentages, but the anecdotes are similar to what I've heard customers say over recent years. Oracle has won over customers in the field.
Here are some of the need-to-know items that didn't quite make it into the Wall Street Journal story:
- The 56 customers interviewed in the Nucleus report were the ones that agreed to do an in-depth interview. Nucleus found many customers that filled out surveys, but relatively few that wanted to go deeper. This fact means that the 56 interviewed were a self-selecting bunch. Nevertheless, you can find customers to gripe about both companies without too much difficulty so it is a bit surprising that SAP took the biggest hit.
- Oracle had home survey advantage when it comes to the mid-market. The Nucleus survey included Oracle customers that had deployed Oracle, JD Edwards and Siebel products. SAP customers had deployed R3, All in One, Business One and mySAP Enterprise. So why does Oracle have a home survey advantage? JD Edwards primarily caters to smaller companies and in my experience vendors that focus on smaller companies do better. As long as Oracle doesn't screw up JD Edwards installed base it'll do fine. Mid-sized CIOs think that large applications companies dumb down their big enterprise products for them--and to a degree they are correct. Meanwhile, the perception is that big software giants will drop the mid-market in a heartbeat to close a much larger sale. Is that the reality today? Probably not, but the perception lingers. In addition, Oracle's acquired customers are typically happy these days. Why? None of those doomsday scenarios about Oracle treating acquired customers like dirt panned out. As for SAP, its big midmarket app, A1S, just got off the drawing board.
- Nucleus found that 34 percent of SAP customers are unlikely to ever achieve a positive ROI without changes to the deployment. All Oracle customers were in line to get a positive ROI. Why the difference? My hunch is it's the consultant factor. Large enterprises rarely install SAP without an army of consultants that can make the software work and customize various modules. Without the budget and resources smaller companies don't have that luxury. A 2004 Nucleus study, reported in this Baseline story, found that more than half of the costs to deploy SAP go to consultants and personnel. There's a reason SAP is building a lightweight out-of-the-box application like A1S--there's a need for it.
- Sixty nine percent of SAP deployments were on time while 78 percent of Oracle deployments were on time. Nucleus says statistically it's a draw. However, there's a disconnect between time and money. Nucleus found 45 percent of SAP deployments were on budget while Oracle customers hit the budget 63 percent of the time. My take: Both of those figures stink, but SAP customers seemed to struggle the most with predicting deployment costs. Hmmm. Consultants?
- The money question from the Nucleus report is this one: Would you do it again? Nucleus found 66 percent of SAP's SMB customers would recommend deploying SAP to peers. Oracle's percentage was 89 percent. From the report: "When Nucleus probed the customers further, we found that 44 percent of Oracle customers felt strongly enough about their success that they were actively recommending Oracle, while only 10 percent of SAP customers felt the same."
As for vendor support, SAP and Oracle customer ratings--3.4 vs. 4.0, respectively, were in the same ballpark. The key difference: Oracle customers got most of their support from Oracle. SAP customers got support from a partner or parent company.