To quote the favorite phrase of a friend of mine, it's like jabbing a pointed stick in my eye. Yesterday I got a call from the owner of a local manufacturing company asking my advice on whether to implement SAP. This is a single-site $5 million operation with one nice big contiguous manufacturing floor, making rather uncomplicated widgets that have a raw material lead time of a couple days and a manufacturing cycle time of a couple hours. The next sound he heard was the bam bam bam of my head banging against the wall. Obviously he hadn't read my post on the False God of the Almighty Algorithm, otherwhise he could have guessed my opinion.
Of course SAP is not really designed for companies of that size, and the cheapest implementation I've heard of is in the $500k range. But the same can be said for almost all ERP systems. Some form of system may be needed for large multi-site or complex operations, but they are often not appropriate for smaller companies, and the are almost always overly relied upon at larger ones.
I talked to this guy at length about what he really needs, the additional complexity that ERP (let alone SAP) inherently creates, etc. He remained unconvinced... he had read that he must have SAP to grow. So I sketched it out for him, which I'll share here.
The SAP (or any large ERP) Option:
First you need a team to decide on the requirements and identify an appropriate ERP solution. If they're good, they will first take a look at your processes to see where they could be improved. Then the software is ordered.
Then you'll need the implementation team to actually install the software, configure it to your processes, and do the training. They will migrate the data, do all the integrity checks, and get it up and running. This usually includes some customization by the vendor, which will make future planned upgrades interesting.
At that point you'll probably realize that it's too big of a job and too complex, so you'll hire a consultant. That person will seem like a wizard, and will appear to get everything working fine, probably without transfering any knowledge of how it happened.
And finally, after everything is working, you'll need an ongoing support team to keep the whole bloomin' thing running, pick up the pieces when it tells you to make a few million extra chocolate bars, and to recommend software "feature enhancements" that will be too expensive to implement.
In the end you'll have something that looks like this. But I'm sure someone will be able to figure it out, right? You can always call the wizard back. The line "pay no attention to the man behind the curtain" comes to mind.
Total cost: $500,000 plus some assorted medical benefits and chocolate bars.
The Staples Option:
Let's keep in mind that this is a relatively small company, single site, simple supply chain, simple widgets. So instead of SAP, how about...
First get some people together and go through a value stream mapping exercise. Nothing fancy... just figure out what you're currently doing. Take a few hours to think about it, to identify non-value-added activities, and kill those wasteful activities.
Next, run down to Staples. Buy a nice big whiteboard. While you're there, pick up a some markers and erasers, a webcam, and a copy of QuickBooks (Manufacturing Edition). Grab some pizza for your team to celebrate their accomplishment (and a keg of beer if your HR group hasn't tied you up in your underwear yet).
Back at the plant, mount the whiteboard on the shop floor where it will see the most traffic. Draw the flowchart on it. As a lean company you want to pull jobs from customer orders, so think about how the chart flows. Put little labeled magnet tags on the flowchart to indicate jobs. Remember to include your suppliers so you can trigger raw material orders. Get everyone on the shopfloor together and explain this visual method, and how it is everyone's responsibility to manage the board.
Now take the webcam and aim it at the board and connect it to the company's network so anyone, such as sales and customer service, can remotely view the board at any time to see the status of any job. Take your copy of QuickBooks, spend a couple hours setting it up, create a couple of user accounts for shipping and receiving people (of course receiving is done right to the point of use on the shop floor...).
Cost? A couple grand... maybe.
I saw this "system" in operation a couple months ago, and since I wrote about it I've heard about three other implementations where the whiteboard is THE scheduling system. I also know of a couple implementations in much larger companies where the whiteboard is used for all shopfloor scheduling, with their ERP system (which can't be killed as it would put a lot of finance folks on the unemployment line) simply managing materials into and out of the floor, and creating vast amounts of paper and reports in the process. Variations include side-by-side kanban, andon lights that trigger a small LED on the whiteboard, and a second whiteboard with metrics and the like.
The key is that it is simple, visible, and especially that it is owned by everyone on the floor. The last company I saw using it had no production leads and the supervisor spent all her time with customers... the shop floor folks managed themselves. With one glance you can see which processes are becoming bottlenecks, and which operations need to have people shifted to them. Flow becomes scarily smooth.
Simplistic? Absolutely, and damn proud of it. A new idea? Of course not.
Visual systems have been around for a long time. But sometimes we need a reminder to take a step back and think about what we want to accomplish. Many of us, especially engineers like myself, can become so enamoured with technology that we forget to think about whether it's really the best solution.