Sponsored Research Reveals Payoff Metrics
November 6, 2008
SAP, as I have noted, is paddling its robåd in waters roiled by customer resistance to some new prices and increasing pressure from upstart alternatives to the industrial strength R/3 system. Aberdeen, a highly regarded consulting firm, authored ‘”Do More with Less: Merging Enterprise Applications with Desktop Tools” on with some support from SAP and Lawson (a maker of business software, not just human resources software). A summary of the report is here.
What I found interesting in this report were the facts highlighted in the news release; for example:
- Integrated technology delivers a 21% increase in gross margin in a 12 month time span
- Decisions are made 27% faster or what the Aberdeen management gurus and guruettes call ‘time to decision’
- And this quote from the news release: ‘technologies like enterprise search allow the Best-in-Class to spend 63% less time in non-productive tasks which include searching for and manipulating data’.
Metrics in search and related disciplines are difficult to verify. Aberdeen’s data exist in this news release is a similarly disembodied form. If I get a copy of the SAP and Lawson underwrite this study, I will look at the sample size, the data analysis method, and the context in which Aberdeen places these data for its paying customers and me, the addled goose.
Big companies, in my opinion, assume that the notion of explaining where a number comes from is silly when preparing a news release. I do not issue news releases very often, and I suppose my PR maven omits data tables.
Nevertheless, these Aberdeen numbers warrant several high level observations. Note. I am not ‘picking on’ Aberdeen, SAP, and Lawson. I am offering a personal opinion about a common practice in consultant news releases: using numbers to get editorial pick up. And it worked. I am writing about this news release in my Web log!
First, if the data are as substantial as they seem to be, why isn’t this type of pay off causing a stampede to the front doors of SAP, for example. The softening of SAP’s guidance, the annoyance of the SAP customers, and other issues suggest that the return is not commensurate with the cost. I may be an addled goose, but if I pay a reasonable price and get a big payoff, I am a happy goose, and I will do the quacking for the vendor.
Second, if the speed of decision somehow correlates with the efficacy of the decision, then I am all for speed. The problem is that in my pond, going fast may be the wrong thing to do. Here in Harrods Creek, I can decide to rush to the mailbox every morning at 8 am to see what my post mistress has delivered to me. But since she does not deliver mail until 3 pm each day, the speed of my decision is essentially irrelevant. The key to a good decision is context, accurate information, and judgment. Speed may have little or nothing to do with whether a decision is good or bad. Therefore, this goose wonders, ‘Why make a decision faster if a higher value cognitive function in incomplete?’ In the sophisticated world of SAP and Lawson, speed may mean more than it does to the goose in Harrods Creek, but I am skeptical about speed with or with out seemingly precise metrics.
Finally, a lousy search system can be a problem regardless of integration. Here’s another down home example. I index what is on my servers in my log cabin office next to the pond. The data I need are not available from a third party like Aberdeen or from the Internet index on Google. My phone is out of order. My fax is dead. In short, I am sitting alone and without a way to get the needed information. I root through what I know. Deprived of context and common sense, I can craft an uninformed paragraph even though I have an integrated system and I have my own subset of data and I have what is in my goose brain. My thought is, ‘The presence of a good or bad search system may have little or no bearing on the time I spend searching integrated information with an integrated system, manipulating the data, and analyzing the interesting numbers.’ Search is important, but if the context is wrong or the data incomplete, speed, integration, or whatever other factor I toss in invested does not make baloney into prime rib.
In my opinion, consultants and dinosaur companies are trying their best to justify the high cost of their systems. Savvy customers know that if the business process is flawed, the software will just make lousy business processes perform faster, not necessarily better.
I know that slapping numbers on independent factors that are not co related while out of context to boot may boost sales. A word of advice: vendors, please, don’t come to my goose pond to make a sales pitch with these faux data. The geese may take the vendors for an invigorating paddle across the pond.
Stephen Arnold, November 6, 2008
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
