I am working on a prediction of future growth of a workshop (50 – 100 manual workers). The idea is to develop how and in what way we would like to grow this facility. I am looking for a methodology or conceptual framework that would put me on rails, something that would give me a direction in developing this plan. Our company is a manufacturing company and we are using a form of Kanban so if you know any approach that is connected to this methodology then even better. I welcome any level of approach, from basic to complex.
-
1I am looking for a theoretical business tool (like S5, SWOT, charts etc.), not software– KrakenCommented Jun 22, 2014 at 10:11
-
1welcome to PMSE! I've edited the post to better reflect what it seems you're looking for. Feel free to re-edit if I missed the mark.– Mark PhillipsCommented Jun 22, 2014 at 13:32
3 Answers
TL;DR
You should continuously inspect-and-adapt your entire system as an integral part of running a successful Kanban implementation. To successfully manage growth, you will need to make changes to your processes if you expect to reduce (rather than increase) lead time as your input queues grow.
Workload as Empirical Data
Kanban is a great framework for estimating lead time, throughput, or cycle time. However, in the Kanban framework, "workload" is an empirical datum rather than a prediction. If we agree to define workload as the volume of stuff to be done within a given time frame, you can then measure the queue sizes within that time frame. For example, you could:
- Average the length of an input queue.
- Find the median of all input queues.
- Identify the minimum or maximum lengths of each queue.
Predicting Workload
In Kanban, you might predict future workload (e.g. queue size) by graphing queue growth over time. You might also simply make planning assumptions, such as "Workload will grow at a rate of 15% per quarter over the next 18 months." Either way, you would project the growth rate onto the size of your queues.
In a pull system, growth of input queues increases lead time, but shouldn't directly affect your cycle time or throughput. Therefore, if you expect growth but do not want to increase lead time, you should examine your processes in light of your growth projections to determine if:
- You can add more labor to increase work-in-progress (WIP) limits.
- You can add additional queues to reduce lead time for a given sub-process.
- You can improve throughput (by reducing waste, not by management fiat) in order to meet your lead-time targets.
You might also find other ways to manage the work. As long as you are continuously inspecting and adapting your processes, and making optimal use of the metrics your framework provides, then you are more likely to successfully absorb the additional work into your processes. There is no silver bullet.
If you have historical data on how many production orders you get, number of workers and the output per week/month, then you can use simple excel pivot & chart to see your progress.
Excel charts have a great feature that shows trend lines and predictions for next couple of periods based on the information at hand, you can use these to have an insight.
Since you mentioned that you are running a workshop i would also like to point to you "Law of Diminishing Marginal Returns". It tells you that your output will not increase by having more workers.
Maximize the contribution margin from your constrained resource
To grow such a business you want to maximize the contribution margin. Contribution margin is the selling price minus the variable cost per unit. In a job shop business of this type your key decision point is the desired product mix you shoot for. For example, if you would like to get more orders for Product A, you can create promotional materials for it, give better incentive to your sales team and so on.
How do you determine which product to push and which one not? In general, you will push those products that give you the most contribution margin. If you do that though, you may be up against some bottlenecks also known as "constraints" in accounting jargon! For example, in a job shop that I worked for earlier, our vertical boring machine was the constraint. It was the most expensive, required the most skill and could also be the most profitable. So, you want to maximise the contribution margin from such a constrained resource. You have to figure out which is the constrained resource in your shop. Once you do, here is a good article that shows you how to model that.
After you have booked the orders, as @CodeGnome pointed out, you apply Kanban to continually improve your processes.