Look up W. Edwards Deming on the Internet and you typically will learn first that he revolutionized manufacturing practices at giant corporations, particularly in Japan after World War II. But even if you’re a remodeler with three workers rather than a factory executive overseeing 3,000, there’s a concept Deming popularized that you can apply.
It’s called PDCA, for plan / do / check /act. It’s a way to improve the processes a company implements by making sure those processes produce ever-better outcomes. At Ford Motor Co., that meant a better-built transmission. For you, it could mean better service, more productive workers, or increased efficiency. Let’s examine what each step requires:
- Plan: Decide what you need to do.
- Do: Execute that plan.
- Check: Measure results and calculate how results vary from the plan.
- Act: Figure out what to change.
Let’s say you want to increase profitability by doing three things: Schedule better, train current field employees, and make sure all materials are on the job when needed. So you write a six-week plan calling for you to buy and learn a scheduling program, set up an on-site with Tim Faller to train workers, and start ordering materials four days ahead of needing them.
Suppose you did all those things. Congratulations: You accomplished the “Plan” and “Do” parts of PDCA. You might even have found that your efforts nudged profits higher. But without analyzing the effectiveness of each of your three tactics, how would you know which is responsible for your success? And if they didn’t work, what part of your plan failed? You might end up having spent a lot of money and time without any visible results.
You're Halfway There — Important Next Steps
The key to success in PDCA lies in the last two components: check and act. Checking requires that you collect information from the “Do” portion of the cycle. What are the hoped-for outcomes of each of the plan components? How will you measure them?
Let’s use our scheduling challenge as an example of what to do. In that case, compare estimated to actual schedules both before and after you executed your plan. Perhaps before the change you scheduled 180 days for a project and finished in 210, thus pushing actual time over budget by 16%. Checking whether your plan worked requires that you set a reasonable but ambitious goal before starting — in this case, cutting scheduling slippage by, say, 10% every quarter. Now that you have a goal, you can check the results.
Similarly, for the training part of your plan, you might compare the work produced by a full-time equivalent employee (FTE) before your trained field workers and then six months later. If your company produced $1.2 million in the second half of 2012 with four FTE, each produced $300,000. If six months after training the same four FTEs produced $1.4 million, each would have produced $350,000.
As for checking the value of your four-day advance purchase plan, look at the number of receipts you have from suppliers that are under $100. They probably reflect last-second purchases. If six months later the number is much lower than the number before you began, the change probably worked.
But you’re not done yet. You need to act again. If the results are what you expected, build your changes into the company culture to maintain the gains. If they’re not, decide whether or not to tweak the current plan or scrap it and try something new. This cycle is repeated continuously as the company matures and different problems arise. That’s why Deming’s PDCA cycle also is called “continuous improvement.”