Recently I was having a discussion with some friends about process improvement in organizations. (Sure, we also talked about sports, food, etc., but when you get two or more QA types together, it seems we can’t help but talk shop at some point.)

As we continued discussing our past experiences in process improvement, we agreed that many times the greatest improvements originate from the grassroots of an organization. You know, someone or a team of people see something that needs to be fixed in the process and they fix it. Or, they have an idea for a new and better way of doing something, so they just do it.

The thing I find interesting enough to write an article about is that this flies in the face of conventional wisdom. In fact, I have taught people over the years that “improvement must come from the top” or it won’t last.  I often make the distinction between “management support” and “management leadership.”  To me, “support” is too weak. It really takes leadership to get something done.

However, this leaves me with a nagging question, “What do you do when management refuses to lead?” Or even worse, “What do you do when the leadership is going in the wrong direction?”

I have often commented that after being inside some of the largest and most influential companies in the United States, I’m amazed sometimes that we even have a gross domestic product. When I think of all the waste, lack of leadership and poor customer treatment, it’s really a miracle that some companies stay in business at all.

The Problems With Top-down Process Improvement

As I think about why top-down approaches often fail (think Total Quality Management or “TQM” here) it could be that people just don’t take them seriously. People are naturally resistant to change, unless perhaps it’s a pay raise.

Dr. W. Edwards Deming used to caution against big pushes and programs for process improvement. He said that quality improvement is not like the United Fund drive – just another program than comes and goes, complete with balloons and T-shirts. However, this is exactly how most companies tried to implement TQM back in the 90’s. So don’t blame Deming for that!

Once I was at a company performing an assessment of their testing practices. I noticed on the walls were all kinds of TQM posters. One poster read “Quality is a Part of Everything We Do.” The disconnect was that their nationwide network went down daily due to software errors the company had implemented with inadequate testing. The day came for me to deliver my findings to about forty senior IT and business managers. During my delivery I remarked that I had noticed the posters and that if quality truly was a part of everything they do, it must be a small part! Fortunately, they laughed instead of throwing things!

If You Don’t Like the Weather, Just Wait!

Instead of embracing a new direction set by senior management, most people have grown cynical over the years toward major process improvement efforts and have started to think “Why bother? This, too, will pass.”  I can’t really give a strong argument against that attitude. Look at all the reorganizations that happen in companies all the time that wipe out entire groups of people and render improvement efforts meaningless.

Also, there’s the issue of who knows best how to improve a process – management or the people performing them? I would pick the people performing them just about every time.

What About the Models?

So what about all the process improvement models such as ISO 9000, Six Sigma, the CMM, and more recently the CMMi?  I just read an article about how the U.S. Department of Defense is struggling to keep IT contractors under control. Costs are spiraling out of control and quality is slipping. Isn’t this exactly what the CMM was designed to improve?

Without getting into all the mechanics of these models, they are just a roadmap for improvement. I find value in all of them, but there are also drawbacks and risks in each of them. Just like a real roadmap, just because you follow the route indicated does not mean you will have a great trip. Your car might break down, you may get into an argument with your traveling companions, you may even get lost or stuck – even with a map!

Risk Avoidance

Risk avoidance is another issue that seems to short-circuit top-down approaches. It seems that the higher you go in the management structure, the less willing some people are to take risks in making changes, even small ones. After all, there are risks in process improvement, such as lack of process adoption, unexpected impacts, etc.  It boils down to the issue of not making a mistake and having a “failure” on your record. With large top-down efforts, it’s easy to spread the blame to many people.

My belief is that unless you fail, you won’t see true improvement. However, not everyone sees it that way.

The Strengths of Bottom-up Process Improvement

I think the strength and value of grassroots process improvement are in the empowerment of people to get the job done well.

A Strong Sense of Ownership

It is true that people support what they help create. When a process is broken, the people that perform it generally know how to fix it IF they are given the resources (time, tools, people, etc.) and empowerment to do so.

This ownership can transcend organizational structure and therefore is more likely to survive the next reorganization.

Agile Response

A grassroots improvement effort is often based on quick experimentation and is done without permission. People have learned that it really is easier to get forgiveness rather than permission. Permission often takes time. In fact, people can get so deep into analysis that the opportunity is lost before a decision is made.

A Limited Investment at Risk

If an attempted approach doesn’t work, there is very little investment at risk. There are no million dollar licenses of processes or courseware at stake. No huge teams of people – just a small team trying something new on the fly.

What About Control?

A reasonable question at this point is, “Doesn’t my manager deserve to know what we’re trying?”  Sure. There’s nothing in a grassroots approach that prohibits that. However, you may be in a situation where your manager may refuse to take a risk and prefer to live with a broken process. The problem is that it is the people most directly impacted by the process (perhaps you and your customers) that feel the pain. In that case, it may be easier to try something in a low-risk context and then tell your manager how it worked (if it worked!).

What if We Fail?

The best companies and individuals have learned that the only way to truly learn is to fail. Failure is not a matter of “if”, but “when”.  Some cultures don’t deal with failure in healthy ways and ironically that propagates even more failures. If you are in one of these companies that punishes failure, you can work to change the culture or perhaps update your resume.

When the threat of punishment is hanging over people, think about what happens. People don’t try new things or take risks, they blame others, and they hide mistakes whenever possible. Sounds like a fun place to work!

When you fail, always take time to reflect and learn from the failure.

Next Steps

Grassroots process improvement starts by having a team conversation and asking some basic questions, such as:

  • What’s broken?
  • What is time-consuming?
  • Where do most of the problems originate?
  • How do other people deal with this?
  • Is there anything we could copy that works?
  • How can we add greater value to the people we serve?

I have detailed this approach in two articles:

Also, if you would like some thoughts on developing a healthy view of defects and process improvement, read Defects are Good.

Then, you will need to make your own “Top Ten” list of improvements. I suggest dividing the list into short-term actions (0 – 3 months to complete), mid-term (3 – 6 months) and long-term (over six months to complete). The short-term items help you build momentum to sustain the effort for the mid and long-term actions.

Next, pick low-risk situations to pilot your improvements. By the way, low numbers do not necessarily mean low risk. For example, you may pick 2 or 3 customers to try a new process. However if these are your 2 or 3 best customers, they would be your highest risk. You certainly would not want to do anything to make matters worse for them.

Finally (and continuously), analyze the results and make adjustments, if needed. It’s the risk – fail – adjust cycle.


I hope this gives you some encouragement that improvement is possible and it doesn’t require a massive effort and expense. Improvement can also be accidental. Some of the most innovative inventions came to be because a seeming “mistake” was made. (The PostIt note was innovated because the glue was too weak for the original purpose!)

So, pick a problem and give it a try. Improve something today!

Powered by Spearhead Software Labs Joomla Facebook Like Button

Buy The Book!


Randy's book, Surviving the Top Ten Challenges of Software Testing, will help you solve some of your toughest testing problems: people problems!
Now in Kindle format!
Click on the image to buy it from

Go to top