Software Selection: The Last IT Frontier?
It has been a decade since my last CIO assignment, but when RSR partner Brian Kilcourse shared an email thread with me, it brought back a lot of memories. The exchange was with one of our vendor clients frustrated with the software sales processes he tries to navigate.
As a CIO, I was frequently involved in selecting new software applications for my company. More often than not, we were replacing something homegrown, and very, very old. The rest of the time, we were automating mostly manual processes. It was never pretty, and you could be guaranteed that in the first 120 days after implementation of whatever you selected, most every conversation with system users would start with these four words… “In the old system… “ followed by a complaint about something we’d left out.
Brian and I have observed on more than one occasion: even in a period when the consumerization of IT is transforming the retail landscape, the issues CIOs face are frighteningly similar to the ones we left behind. And so it is with software selection. As David Byrne used to sing, “Same as it ever was… “ but with a couple of odd twists.
- CIO’s often don’t own the budget anymore: Many of you have reported that capital budgets for new systems now live with the Line of Business (LOB) executive. This can make the CIO more a deal-killer than a deal-maker. In other words, the LOB exec may like a particular app, but the CIO can say “Well, this app will cause me endless integration headaches and an extra [insert the currency of your choice here] to complete. ” Down goes the deal.
- We have heard anecdotal stories of procurement types as gatekeepers. It was bad enough when retailers started using procurement types to source merchandise (I have a story about ‘the un-installable towel rack’ that dates to my AMR days), but to select technology? Seriously? Are any of you really doing this? And if so, how do you make it work?
- The NRF ARTS committee has created a set of industry-standard RFPs. I have mixed feelings on this one. On the one hand, it’s useful to start from some kind of template when creating and RFP/RFI. On the other hand, one challenge with RFP’s is “you don’t know what you don’t know. “ The Devil’s in the details, but a mountain of details does not really a thorough requirements definition make. In other words, neither Brian nor I are overly bullish on the RFI/RFP concept in general, standard or otherwise. It doesn’t generate enthusiasm among the users. In fact it’s painful for everyone involved, from the user to the vendor (as Nikki can attest). Plus, there’s no real way to be sure the thousands of questions in an RFP are answered accurately.
As far as we can tell, one method of system selection remains pretty much intact. A company can hire a company to do the selection for it. This usually involves a lot of interviews, a lot of paperwork (see RFP above), and a lot of money. It’s thorough, but again, the forest may well be lost in the trees.
So now that I’ve basically dissed almost all alternatives, you’re probably asking, “Well, what WOULD you do? ” Nikki is really bullish on conference room pilots. I like the more old-fashioned requirements definition documents (no yes or no answers, give me words), followed by a conference room pilot. We all like the idea of getting references from similar companies.
The bottom line is that as quickly as we can get mobile sites up and running, presence on social networks and brute our way to cross-channel fulfillment, we are still stuck in the selection model of the 90’s. It seems that software selection is the last frontier of retail IT. While integration remains a persistent problem, we certainly know how to solve it. It takes time and money, but it’s not really a mystery. Software selection seems to an art form rather than a science.
Which leads me to my last question. How do YOU do it? Next week, Brian will report on methods used by some retail winners. I’d love to hear what you have to say as well. Drop me a line… you can tell me I’m all wet, say “Amen, ” or discuss alternatives I’ve missed here. How do we cross the last frontier?