I have been keeping track of labor times per operation for high-end custom cabinetry and I wonder how long some of these tasks take in other shops. I track layout, face frames, panel cutting, line bore and rabbit, edgebanding, cabinet assembly, cabinet route and sand, drawer manufacturing, roll out manufacturing, door manufacturing, kicks, shelves, wood tops, finish assembly, etc. I know that face frame times vary given the size, but it seems to average one per hour. Same with cabinet assembly. We do lock joint construction on drawers and I shoot for 3 drawers per hour and 2 roll outs per for manufacturing. How many of you actually track this stuff? We are a 4-man shop.
I tracked time using an automated time clock and I found that 21% of the total employee time was un-billable. This time was not for each employee, rather for all employee time combined. I totaled the number of hours on the clock and divided by the category used to track "shop" related things. This time included clean-up, machine maintenance, material unloading, things that were not tied directly to one job. When I figured the cost per employee hour, I factored in this time. When I track time for an operation I take out for the fifteen minutes twice a day for coffee breaks--when I figure the cost, I add this back in.
Instead of looking at operations in isolation, I try to concentrate my focus on the dollars produced against labor expended over a given period of time. Even when I break it down into individual machine times, I measure over a length of time sufficient to take all the activities into account.
Once the raw data was collated I made up a spreadsheet. The rows are the objects to be built and the columns are the operations. Simple formulas were set up for each cell with totals at the bottom for each operation and totals at the end of rows for a given object. What an eye opener. Very easy to see constraints and bottlenecks and time wasted. Most importantly, how jobs could be bunched, batched or otherwise improve efficiency. Also, what sequence a job should follow.
We now do doors and drawer fronts first, completely finished and ready to install. Only then do we cut panel. At assembly, the doors are hung and drawers and hardware installed. The cab is bubble wrapped and stored ready for shipping. A single cabinet is now only handled one time prior to loading. Very different from what we had been doing.
By eliminating unknown and known inefficiencies, we've cut our job times by 40%. We are much more competitive and now know exactly how long a given job will take. No more scheduling problems and more importantly, 100% on time.
We now regularly add rows as new items appear or we break down a given item into its component part. Also, times occasionally have to be adjusted to different employees and additional machinery. This has made our small shop into a highly competitive unit. And I'm sure it would be a boon for the larger shops.
Each shop works differently and utilizing one spreadsheet for all wouldn't work. But having reviewed some of the estimating software available, I found it to be to generic. Once the raw data was gotten it probably took 25 or 30 hours to refine the spreadsheet. Well worth the effort.
As I see it, there is only one real way to track this stuff. You have to do it, provided you know how to, and are competent at each task. Keep a notebook or whatever with you and write down start times and stop times. The reason you need to do it is because sometimes employees will fiddle with the times. Leave the clock running for a certain portion of your "non-production" time, i.e., going to the bathroom, clean up, getting a drink, and maybe a phone call or two. As much as we would like our employees to have their nose to the grindstone 8/5, it's simply not going to happen.
Once you have a benchmark, so to speak, you can check your numbers against what your employees are actually doing. Since I'm not in the shop every day, even though I know how to do all of it, I'm not as fast at some tasks as the guys who do it everyday are, so those times may need to be tweaked a little. Other tasks I'm very fast at, maybe a little faster than the guys who do them, so those may need to be adjusted a bit as well.
When you do this yourself, try to do it at a normal pace--don't go real slow/fast and forget you're on the clock. You want actual numbers.
Once you have some good solid numbers, when you hire someone new, you can generally tell quite quickly rather they are going to make it or not.
What's the first step in the process? Cutting? How long does it take? Make all the rest of them take that long, too, via extra or less labor, machinery or purchasing. Break the shop down into divisions first, and just concentrate on your worst to start.
Knowing no employee works 100% of the time at 100% capacity is the key. You're better off doing this test on a Friday afternoon than a Tuesday morning and you're better off not telling anyone you're doing it if you want true times. Double those times in case the guy doing it quits and a new guy has to learn, and you'll be about right.
We have gone to paper time cards that each guy keeps on a clipboard so he doesn’t have to walk to a time clock. The next day the data is entered into the erp software item by item so most of the mistakes get caught there. The software also tracks the quantity of each item run and the estimated amount of time for doing that operation so it becomes very obvious when there has been a mistake.