Socrata

In the 1990′s I was a software engineer, writing software to automate distribution centers (most of us would just call them warehouses, but whenever I say warehouses people assume I mean data warehouses in that I work with data so often).

One of the more interesting systems we deployed was for a very large northwest outerwear manufacturer selling ski parkas, hiking pants and the like. Their customers would order whole cases and/or individual items. The individual items would be picked by people, who would lay them bar code up on a conveyor belt. The items would be conveyed up to the mezzanine level, which had 80 packing chutes arranged in a giant U shape. The items would drop one by one onto a tilt-tray sorter, which was a continuously moving train of trays that looked like flatbed cars. The bar code scanner would scan each item and then when the tray passed over the appropriate chute, the tray would tilt, dropping the item down the chute.  When an order was complete, one of the 10 to 20 workers would pack all the items from a chute into a box and apply the packing label.

The initial problem we encountered was that a worker would waste time determining if a chute was ready to pack. If all the items hadn’t been dropped down the chute yet, it couldn’t be packed. The simple but elegant solution was installing a few large monitors in the mezzanine area. The monitors simply displayed a chute number of a chute that was ready to pack. If more than one chute was ready, the monitor cycled through displaying each ready chute number every 2 or 3 seconds. The workers no longer had to figure out if a chute was ready.

The same technique can be used to focus any kind of workers on any kind of metric. At blist, we try to run the business as numerically as possible. We measure lots of things – bugs by priority, bugs by age, retention rate, # of new users who signed up yesterday, # of active users, etc. We can and do generate a nightly stats package which is emailed to all employees and the board of directors. But if you’re a busy software engineer, it’s easy to get caught up in other things besides reading the nightly stats email.

Recently we installed a metrics monitor on one of the walls in our main engineering area. We loop through some, but not all, of our key stats, swapping the metric every 30 seconds. It’s a great way to draw attention to the stats that matter. So now when an engineer is waiting for some code to compile or just trying to relax his eyes by taking his eyes off his own monitor for a minute, he can get up-to-date information about the key metrics by which we run the business.

You might consider doing the same. It’s a cost effective, visual and fun way to call out the key business metrics.

This entry was posted on Wednesday, December 3rd, 2008 at 4:53 pm and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.