We’re just about ready to launch our 2014 Pricing Benchmark report, and two themes really jumped out at me. Well… in truth, more than two, but let’s say two NEW themes jumped out. First: Our industry has become fixated on competitive prices. Second: retailers are executing all kinds of sophisticated pricing schemes without strong confidence in the data used to drive automated and manual decision making. We’ve got to do better. We can’t use precision analytics on imprecise data. We’ll get awful answers.

It reminded me of my life in the shoe industry…and a story that I shared with Manhattan’s customers at a breakfast event the morning after the NRF Big Show. It’s a story about the shoe business circa 1990. I worked for a company named Morse Shoe. It owned the now departed Fayva shoes, and also ran the shoe department for most mass merchants – who have since mostly been killed off by Walmart. Names like Bradlees, Hills, Gemco, Caldor…companies once strong, now gone.

The shoe guys knew their data wasn’t very good. They had sku numbers on their tickets, but didn’t have a lot of confidence that they were being rung up accurately at the POS. They had some level of justification for this concern – half our business was done in those leased departments. It was easy money for the lessor, a clean 4% of sales, and no requirement to staff the department, invest in the inventory, or even gain a deep understanding of the size and color intensive shoe business. They also didn’t have a whole lot of skin in the game of accurate sku capture. So the merchants at Morse developed some interesting compensatory mechanisms. They’d look at the counts, compare them to the sales and would often say “That can’t be right…it doesn’t feel right. We need to change them. “

And change them they did. Their solution was simple. They’d take the chain total by sku and change it to a number that made more sense. Store counts were sent in from the field and captured by the data entry department, and they’d replace the system counts. But the detail and total didn’t foot, and no one really cared. Well, in fact the IT guys cared…because we had no stable base to do planning, allocation and replenishment with. And the merchants really wanted that.

My boss and dear friend, the late John Fiore went on a campaign, and we eliminated the count adjustments, eliminated the manual store counts and forced more compliance at the POS. Once that was done, we implemented first a planning, and then an allocation and replenishment system. We helped the company move into the late 20^th century. Sadly, an LBO in 1986 had sealed the company’s fate a decade hence. A balloon interest payment in 1994 drove the company into immediate insolvency, even though we’d had our best operating profits ever. But at least we could say we’d had a solid base of data, and at least I could say I’d accomplished the first successful implementation of the Arthur Planning System in the US, followed by the first implementation of (wait for it) MMS running on CICS for OS2. Those of you under 40 will not understand a word of the previous sentence. Trust me…they were both big deals.

So here we are, 20 years later, and we’ve got some amazing science. Long gone are the days when you’d push the “recalc ” button on the planning system and go out to lunch waiting for the new plans to appear. Everything happens really, really quickly. Yet 37% of us cite lack of clean price, competitor and purchase data as a top-three organizational inhibitor to better pricing solutions. That’s a big number, kids. Forty-seven percent of us say we can’t keep up with competitors’ prices or changes in price by manufacturers. That’s a problem when it’s one of the most frequently cited business challenge.

We know the solution is “data cleansing projects. ” Retailers have reported it as a tactical solution in study after study. But data cleansing isn’t a project, it’s a way of life. And it’s science, not a feeling. Seemingly simple things:

• Careful counting of inbound receipts
• Careful scanning at POS. Many companies don’t allow quantity overrides. They require the scanning of every single item.
• Keeping our hands off the inventory numbers. This is a bigger deal than it sounds. A core issue I’ve had with RFID initiatives in department stores is their fundamental incompleteness. No cut-offs, no preparation, and no roping off of areas. So I’ve asked multiple times: What do you do with the counts you get? Do you book them? Or keep the difference between perpetual inventory and counted merchandise as “memo shrink? ” I don’t believe we can do either one in good conscience.

In other words, beneath the science that we’ve come to rely on, we have a core requirement for consistent and clear processes across the entire enterprise. Unfortunately, all that counting costs money – money we’d all rather not spend. But if we don’t we’ll find ourselves back at the most basic computer tenet of all: “Garbage In, Garbage Out. “