A neglected retail IT trial is worse than none at all

A neglected trial is expensive in ways far beyond out-of-pocket costs.


A throwaway reference in a Reuters story last week showcases a critical retail IT problem, which is that chains in the hectic holiday shopping season are quite willing to try new technology but are almost never willing to support it. That means insufficient or non-existent signage, associate training or simple attention to details to make sure it works reliably.

"Recent visits to a Staples store in New York City found that a kiosk set up to allow people to order online wasn't functioning, and at a JCPenney store in the city, the Wifi didn't work," the Reuters report said. "Both companies said the problems encountered were unusual and that they have backup systems in place."

Both companies were lying. First, such problems, especially during the holidays are extremely far from being unusual. As a retail IT reporter for 16 years, I can't think of any holiday shopping season where technology trialed by major chains didn't fail when tested. It's when it works perfectly that it's noteworthy.

The second claim — "that they have backup systems in place" — doesn't even need refutation. If they in fact had backup systems, the reporter wouldn't have experienced those problems, no? The most charitable interpretation of those comments would be that they meant "We have a plan to fix it — eventually, when we get around to it, like perhaps Jan. 2."

Typically, the problem starts with a senior IT exec, often the CIO, who asks, "What will it cost to run this trial?" And a vendor will sometimes reply to a large chain, "Nothing at all. We'll handle all of the hardware and software costs, and we'll send teams to install at every one of your selected stores." To that IT exec, it seems like a no-brainer — and it is, in the sense that agreeing to such a deal is evidence of the decision-maker having no brain.

This train went off the tracks because the key question asked ("What will it cost to run this trial?") was the wrong question. The question that needed to be answered was "What will it cost to successfully run this trial?" And  "successfully" needs to be defined as having lots of your customers use the service and for those customers to enjoy a pleasant and powerful experience. And for those pleasant experiences to deliver whatever metrics the company wants, including potentially increased revenue and/or profits, behavioral changes (buying multiple items instead of one or adding more accessories, etc.) or bringing in new customers.

Vendors are famously optimistic about their offerings, with the vendor CEO thinking that merely having its offering in a Macy's or a Walmart will itself be a magical thing. What those vendors don't realize is that allowing the trial to happen with insufficient support will all but guarantee horrible results. In some cases, it will be unwarranted horrible results. In short, cutting corners on a trial isn't doing any favors for anyone — the vendor, the retailer or that retailer's customers.

Much of this comes down to a headquarters versus store mentality. Under the best of circumstances, HQ has made sure that people (often from the vendor) are available to support the stores in case anything goes wrong. That's great, insofar as it goes. But do the stores know this? Have the associates been told? If the system glitches, does whatever assistant store manager who is in charge at that time know who to call/email and what details to look for?

Let's now get realistic. Let's say it's Dec. 19 and that assistant manager is dealing with lines of customers in the middle of an ultra-hectic shopping mall. If a newfangled mobile mechanism stops functioning, is that assistant manager going to search for and find the memo detailing repair process? Or is it more likely said assistant store manager will ignore it and proceed as he/she has always done?

Here's the key point: How many sales will it cost the store for that assistant manager to stop and focus on this? Knowing that making sales is the manager's only goal, what should he/she do? Which decision will place his/her job in greater jeopardy?

The only way a trial will work is to set it up in a relatively tiny number of stores — and for each and every one of those to house someone from the vendor who will make sure the system is functioning and will have the tools to immediately fix it.

Here's what that IT exec should say to the vendor: "The only way I am going to sign off on doing this is when I see a plan for us to do it right, to launch in such a way that will deliver on all of our objectives. I want signage in every store and lots of it. Your teams will personally train every associate at every store, and you will assign people for all shifts at all of the stories where we run the trial. If we're not going to do this successfully, let's not do it at all."

A neglected trial is expensive in ways far beyond out-of-pocket costs. The biggest is missed opportunity. That's when a neglected trial delivers almost no customers, which your team (incorrectly, but understandably) interprets to mean that shoppers have no interest in this capability. You then opt to not use the technology. A proper trial might have shown significant interest. You may be about to walk away from technology that could sharply increase your profits, all because you ran a neglected trial.

This article is published as part of the IDG Contributor Network. Want to Join?

The march toward exascale computers
View Comments
Join the discussion
Be the first to comment on this article. Our Commenting Policies