Everything Skye Thompson knows about the plan-do-check-act framework (PDCA) comes from his wife’s uncle, Greg Howell.
Howell was a military veteran who served in Vietnam for the U.S. Navy as a civil engineer, Thompson said. His job was to work with his team to build a bridge or runway in enemy territory. With limited room for error and few resources, they relied on the PDCA framework to execute their projects.
What is PDCA?
Howell went on to become a thought leader in PDCA, Thompson said.
Today, Thompson oversees Human Interest’s projects as its senior program manager and relies heavily on the continuous cycle of PDCA to manage those projects.
While the stakes aren’t as high in the startup world as they were for Howell, it can still be paralyzing to take the first step in a project, Thompson said. The PDCA framework makes it easier to fail and learn from your mistakes.
Here’s how to pull it off at your own company.
Identifying the Problem You Want to Solve
Within every business, there’s never a shortage of problems that need to be solved. Your customers might be requesting a new product, the business needs to tap into a new market to grow or your code has too many bugs that cause release delays.
Each one of those issues could be tackled next in a PDCA cycle. The challenge for product and operations managers is figuring out which problem to focus your resources on. This is important for any company, but especially so for early stage startups, where every action comes with an opportunity cost, according to Reputation’s Senior VP of Product Pranav Desai.
Thompson compares building projects with PDCA as maintaining a set of spinning plates. Since your goal isn’t to completely solve a problem but instead to continuously improve upon a solution, each cycle is designed to keep the problems at bay (or the plates spinning). If you focus too much on one issue, however, another plate will crash. So you hop from one project to the next running new cycles the moment the solution starts to wobble.
“The first step is to set those objectives and figure out where you’re going to get the resources to put this thing in motion and do what you need to do to get to that objective.”
So, how do you figure out which problem to work on when everything seems important?
To start, Gocious Chief Technology Officer Maziar Adl recommends gathering a clear understanding of what your business wants to accomplish. This could be a goal like growing your revenue to $100 million in the next year or increasing your customer base by eight percent.
“The first step is to set those objectives and figure out where you’re going to get the resources to put this thing in motion and do what you need to do to get to that objective,” Adl said.
From there, you can talk to key stakeholders and review customer data to identify the problems preventing you from reaching that goal. The key is to figure out where you have gaps in your process and then ranking them based on which ones are the most significant, Adl said.
This will help you determine what problem to prioritize first in your PDCA cycle.
Breaking Down the PDCA Cycle
Put Together Your Plan
The planning stage in PDCA is all about looking at a gap you have in an existing product or workflow and then sketching out the solution.
Thompson starts every plan with a journey map.
The process begins in a room (or virtual room) with all of the stakeholders involved in the project and a whiteboard. If it’s an internal project, that group will include people from the departments affected. If it’s a new feature or product, they might include customer success to represent the customer’s interests.
The goal is to flesh out the problem and validate those ideas with data.
“As they say, a problem well-understood is half solved,” Thompson said. “So we focus on really understanding the problems. What are they? Are they real or are they just our ideas? Can we validate them? And how important are they?”
A good plan includes a description of the process you have today, the changes you need to enact to reach the goal and metrics for how you’ll measure the project’s success, Thompson said.
“Creating small spirals around [a project] is the best way to act on it.”
It’s important to remember that the bigger in scope and ambition the project is, the more it’s likely to fail. While you can have ambitious overall goals, the scope of your PDCA cycle should be on a tangible outcome within a manageable timeframe, Desai said.
Desai suggests thinking of your projects as a spiral. Your first cycle will start small with a mock-up or prototype that you can execute quickly. This allows you to make adjustments before you’ve invested too much in your product and it also keeps you agile to the whims of an ever-changing business landscape.
“If we do something nine months out for a release or a year out for a release, we don’t know what the world will look like or what extraneous factors will impact us,” Desai said. “Creating small spirals around [a project] is the best way to act on it.”
What does a plan look like?
At Gocious, Adl conducted a PDCA cycle to improve the engineering team’s day-to-day operations. It started with the realization that projects would get delayed at the last minute because there were errors in the code that needed to be fixed.
To solve it, they decided that the team needed to increase its test coverage and it set its target at 95 percent. Since the company couldn’t afford to hire more employees, Adl and his team focused on increasing testing automation. And they planned to accomplish that by having the engineers write test cases during pre-production to ensure there were no bugs.
Set Your Plan In Motion
Thompson considers this phase (do) the least exciting part of the cycle. At this point, you’re putting your plan in motion, either building the product or feature, or adapting a new workflow.
This stage is standard project management, he said, but it’s still important to get it right. Otherwise, you could face delays as people prioritize other tasks.
“If you have a clear way to express the problem and contextualize it, then you’ve created meaning for them.”
“The biggest disconnect is confusion around priorities, where they’re turning attention away from this project to another one,” Thompson said. “If they don’t understand what’s at stake and why the project is important, then interdependencies among teams can get off the rails and things can go sideways.”
To keep everyone on the same page, Thompson recommends doing a project kickoff with both the stakeholders directly involved in the project and those who are indirectly involved. His goal is to make sure everyone understands why the project is important and how it fits in with the company’s larger goals.
“If you have a clear way to express the problem and contextualize it, then you’ve created meaning for them,” Thompson said. “But you also want to communicate where this project fits in the overall strategy of the company. That’s motivating. People will want to knock stuff out of the park.”
Evaluate the Results
Once you’ve put your plan in action, it’s time to evaluate the results.
The check stage is how you determine whether your project is on the right path toward reaching your objective. The metrics need to be based on a tangible outcome tied to your overarching goal and tied to a short-term checkpoint.
“It’s better to keep it on a smaller cycle because it brings the cost of failure down.”
If your objective is to reduce errors in your code by 50 percent, then you might check after one sprint to see if the number of errors has gone down.
“It’s better to keep it on a smaller cycle because it brings the cost of failure down,” Adl said. “You don’t want to go six months and then realize, ‘Oh my god, I’m way off.’”
For external projects like new features or products, customer feedback can play an important role in evaluating your plan, but you have to ask the right questions, Desai said.
At Reputation, the product team will release a new prototype to a small group of power users that have agreed to participate in a customer advisory board. That group will receive a push notification asking if they would like to test out the new product, Desai said.
“We’ll find that even out of a high-fidelity prototype that we were sure of, 50 to 80 percent of the items we thought would be valuable turn out not to be valuable, but [the remaining] 20 percent of items are super valuable.”
Once they test it out, the product and user research team will select five to seven customers to interview — any more than that ends up redundant, Desai said. The team will typically ask how the customers feel about the product, whether the solution makes sense and if it makes their life better. The goal isn’t to sell them on the product but to learn and discover.
Often those interviews will guide them to the features that really resonate with buyers. And that shapes what they’ll do in the act stage.
“We’ll find that even out of a high-fidelity prototype that we were sure of, 50 to 80 percent of the items we thought would be valuable turn out not to be valuable, but [the remaining] 20 percent of items are super valuable,” Desai said. “That gives us an opportunity to re-evaluate the work.”
Ultimately, the check stage helps you decide whether you need to return to the drawing board or move forward with the plan.
Make Adjustments and Scale the Project
The act stage is where you put your plan in motion on a larger scale. Thompson, however, prefers to think of it more like an adjustment stage to reflect the cyclical nature of PDCA.
“It really comes down to: Did we learn something in that flow that compels us to act again? Or are we satisfied for now and we’ll move our band to the next important project?”
This is when he reviews the lessons he’s learned from the check stage. Did the results meet the objective? What did or didn’t work in the pilot program according to the customers or team involved? Those results will factor into a decision to either start a new cycle to improve upon the solution or standardize it.
“It really comes down to: Did we learn something in that flow that compels us to act again? Or are we satisfied for now and we’ll move our band to the next important project?” Thompson said. “We’ve done both and sometimes you learn something that surprises you and you want to capitalize on that.”
This is especially true for internal projects at fast-growing companies, Thompson said. Workflows can always be improved; you have to weigh the gains you’ll get trying to perfect it against the cost of not fixing another immediate issue.
“We’re only six years old [as a company] and a lot of our internal tooling needs to be improved,” Thompson said. “There’s a huge amount of opportunity to fix those flows and processes.”
If an internal project can be implemented on a larger scale, then the next step at this stage is to create a standardized procedure. Adl suggests starting with a pilot team and then expanding to another small group within the company. If you can showcase the advantages to the new workflow, then that can help drum up excitement for others and you can expand it more naturally.
For customer-facing projects, Desai recommends introducing the product with a pilot group of power users to start. Anything that doesn’t work or make sense for your power users will be an even bigger issue for your less dedicated customers. From there, you can expand the spiral and introduce new features or bring in additional teams like marketing or design to prepare the product for full release.
Still, it’s important to stay flexible at the act stage. Small glitches will come up; what worked for the pilot team may not work for the team at large, Adl said. There are always opportunities to keep improving.
“Sometimes you’ll have small glitches,” Adl said. “So these are the small corrections you make to make the [new workflow] part of the culture of the company.”
Keeping Your Plates Spinning
No project is truly finished at the end of a PDCA cycle.
The continuous framework is what allows teams to adjust and add to a project over time. But it also means you have to know when to return to a project to iterate on it, otherwise that progress could get lost.
That’s why it’s critical to keep organized documents and a system for when you plan to return to a project.
“There’s so many priorities and you can’t let any of them fall, but you have to keep moving to the next one.”
Thompson creates a wiki-style document in Atlassian’s platform Confluence for each project to help him keep track of his projects. The documents contain details on each project, highlighting what actions they took, what they’ve accomplished and what they’ve learned from their last cycle. It’ll also include any notes on what they still need to accomplish, providing the jumping-off point for the next PDCA cycle.
After each project, Thompson will also decide with the stakeholders involved in the project on when they want to revisit it for another PDCA cycle. He’ll then set an alert in his calendar to jog his memory.
While it can be a lot to juggle, it still beats investing months of resources on a project that was flawed to begin with.
“There’s so many priorities and you can’t let any of them fall, but you have to keep moving to the next one,” Thompson said. “That’s [life in] startups, even early stage startups.”