With SAS Global Forum coming up next week, I expect my Google Tracker to start pinging as SAS’ PR team gets to work. In the meantime, a few news hits.
KXEN sells something. Hooray! You know you’re in trouble when somebody buys your product and it’s news. While the folks at KXEN celebrate a win, I must note that their claim in the press release to simultaneously increase the client’s sales by 230%, maximize campaign ROI and increase the conversion rate by 230% is logically impossible. I’m guessing that KXENs’ PR agency figures nobody actually reads the releases, or that the rubes are too dumb to notice an absurd claim.
KXEN’s black-boxy approach to analytics makes sense when you’re embedding analytics into a larger application and selling solutions to the C-suite. It makes no sense at all as a standalone analytics platform, since the folks who actually buy those still care about the math.
So it’s no surprise to see reports of a possible acquisition in the near term by the likes of IBM, SAP or Salesforce.com.
***Update: KXEN has changed the story, and now just claims to have optimized the campaign audience and increased the conversion rate, which is plausible; in targeted marketing, you optimize the audience by selecting the best prospects, which increases the conversion rate but reduces total sales. (To maximize total sales, you spam everyone on the list).
However, an increase in the conversion rate can be caused by many things — the value proposition, the creative, changes to the landing page, and so forth. The only valid measure of KXEN’s analytics is to compare the performance of a targeted group against a control group, with all other variables held constant. That’s Targeted Marketing 101.
SAS has a nice rock collection. Really nice, including dinosaur eggs. Memo to SAS PR: if you’re trying to position SAS as an innovative company leading the world into a bright new future of advanced analytics, try to avoid using the words “SAS”, “dinosaur” and “fossil” in the same sentence.
IBM “reassigns” S&T exec. Ginny Rometty is not happy with the Q1 numbers, and Rodney Atkins pays the ultimate price by getting reassigned to Corporate Strategy, a move that says something less than flattering about IBM strategy. (I believe they call this “failing upward“). Ginny also laid down the law that from now on IBM employees will respond to customer queries in twenty-four hours or less, a diktat that is (a) unnecessary for IBM’s competent reps, and (b) pointless for the others.
So is it sales execution, product strategy or something else? Enterprise hardware isn’t exactly flying off the shelves lately: the hardware businesses for EMC, H-P, IBM and Oracle are all down year-over year, by double digits in some categories, a phenomenon variously attributed to the growth of Cloud computing, virtualization, commoditization of the server market and a slow economy. Anyone not living in a cave knows that Cloud is growing big time, but those Cloud operations buy servers, too — there’s a reason they call them “server farms”. Virtualization? Meh. There’s more evidence supporting the commoditization theory — IBM took the biggest hit in mid-range Power systems, while H-P’s sales declined slightly and Dell’s Enterprise segment grew slightly.
But I say “it’s the economy, stupid” and the fact that much of the recent growth in server workload consists of people doing stuff that doesn’t matter in a pinch. Last year, I worked with a banking client whom I shall call Enormous Bank. Enormous Bank has SAS running on big AIX machines and, like many SAS customers, the machines are constantly maxed out (the “Freeway Effect”), to the point that Enormous could not run its Basel II reports.
What happens to a bank that cannot run its Basel II reports? Answer: it soon becomes an ex-bank, like the parrot in the Monty Python skit.
SAS offered its standard guidance for customers with performance issues: Buy More Servers, a strategy heartily endorsed by the IBM Client Executive, who totally agreed that buying more servers would be a grand thing for Enormous Bank to do. Did Enormous Bank buy more servers? They did not. Instead, they simply kicked every SAS user off the system, ran the Basel II reports, then let the SAS users come back to play. Call it the Soviet style of workload management, if you will, but crudely effective and a lot cheaper than Buy More Servers.
Would Enormous Bank do that if those SAS users were doing something productive? No, they would not. Perhaps Mr. Atkins should ponder that over in Strategy.
