MM: The next thing I’d like you to talk to is the optimized workflow.
With these basic principles of accountability and no enabling bad behavior or sloppy work. Staying online. Getting it right upfront.
With that set of principles, your current-state process, as I recall from our conversation—entailed about 300 or so discrete steps. Is that right?
TM: That is correct. 300 individual steps.
MM: Then how did you start to develop the optimized future-state capability?
TM: Well, as you mentioned—the visual representation on the wall. We kind of created, after we had all of the artifacts out there, we either took pictures or put them into a document. We reduced them so that they could fit on half of that wall. Then we drew a line under that.
The top of the wall was the current. The “current,” being with a few modifications that we felt we could implement right away. Then we drew a line.
On the bottom, we started the new process. “What do we do now? What is the process that we should lead with? Look. You’re doing this on the backend, and we know that there’s pain back there. We want to move this up into the process.”
Subject Indexes in Two Hours vs 30 person-weeks
MM: So for example, there were two things that I want you to address. First, how your team builds the subject index in the back of the catalog and, second, how your ream conducted profit analysis of each page in the catalog.
Let’s talk about the index, first of all. In a 900-page catalog, creating an index can be quite the chore. It typically happens at the end of the production cycle.
So there was like a two-week period at the very end of the production cycle, where 15 to 20 people sequestered themselves into a room and basically called out names and told what page number those particular items were on. That’s now the index was built.
Again, you’re talking 10 to 15 people for 8 hours a day, for 2 weeks. That’s a lot of man-hours.
MM: Yes. That’s 30 man-weeks.
TM: It had to be done. So we recognized with database publishing how electronically that information is captured within the database, and—through a simple export—that same result can be more accurately and quickly exported within a matter of 2 hours.
But more importantly, we reduced the cycle time by two weeks.