Posts Tagged ‘forecasting’

Automakers are just different. Given their size, just what they do matters to people ranging from suppliers to policy makers. Thus it is news when a company like General Motors decides to be a little less transparent about just what it is doing (GM goes dark on output stats, causes stir, Automotive News, Jun 17).

For decades, GM and nearly all other major automakers have reported the number of cars and trucks produced at their North American plants each month, broken out by nameplate. The data get folded into numerous economic indicators, including ones published by the Federal Reserve, and are a benchmark for industry insiders to forecast GM’s future production.

But this month GM notified several research providers that publish production data — including IHS Automotive, the Automotive News Data Center and Autodata Corp. — that it will no longer give them those figures, providing instead only the number of wholesale deliveries.

Things brings up two questions. First, who was using this data? Second, just why is GM reporting wholesale deliveriesĀ instead? (more…)

Read Full Post »

My office sits in Nathaniel Leverone Hall. I never gave any thought to who Mr. Leverone was until on a visit to my parents in upstate New Hampshire I drove past Dartmouth’s Leverone Field House. That got me wondering who the guy was. It turns out that Mr. Leverone was from New Hampshire and went to Dartmouth but made his fortune in Chicago. He was a vending machine magnate.

I got to thinking about Mr. Leverone again because of a Wall Street Journal article on the technological evolution of vending machines (Restocking the Snack Machine, Aug 3). Companies who sell through vending machines, like almost everyone else, has been under pressure of late. Margins are thin and a number of typical sales points (factories and offices) have closed.

The response, at least among those that own a large number of machines, has been to go high tech. Imagine, essentially, an Internet enabled Coke machine that tells its owner what has been selling.

“It’s only natural that, when times get tough, people search for new and better ways to do things and take risks on doing things they might not have done when times were going well,” says John Mitchell Jr., an owner of Treat America Ltd. of Merriam, Kan., which has about 12,000 vending machines in the Midwest.

Treat America’s drivers used to stock machines with the most popular foods in their regions, based on sales tracked manually by category rather than individual items. It was difficult to know how different products were selling in particular machines.

Since January, Mr. Mitchell has outfitted about 40% of his machines with systems that record and transmit real-time sales data in each of a machine’s 45 slots. Mr. Mitchell says he now knows that about 40% of the slots are what he calls “dead spirals,” dispensing less than one item per week.

While packs of Cheez-It crackers sell the most of any item in a vending machine in a Kansas City, Mo., call center, they don’t even reach the top 10 in a hospital three miles away.

“You’re catering to a population that might be as small as 30 or 40 people,” Mr. Mitchell says. “The unique preferences of that population can drive sales significantly.”


Read Full Post »

%d bloggers like this: