, ,

Why Government Should Not Go Lean

Just after I posted “Business Process Management As If People Mattered: Adaptive Case Management” the latest issue of the Harvard Business Review downloaded to my Color Nook. This was a special issue devoted to talent management and so I flipped through it pretty quickly until I came across “Lean Knowledge Work” (Staats and Upton, October 2011) So, had someone finally found the secret to applying Lean and Six Sigma to knowledge work? Did this mean that adaptive case management (ACM) was dead before it even caught on?

The short answer – no. The authors, Staats and Upton, claim that Wipro Industries has successfully applied Lean to their knowledge work but upon a closer reading of the article, their case for better knowledge work through Lean seems weak. In fact, the authors’ findings were more encouraging for ACM than Lean. My two major argument against Staats and Upton’s conclusions is that they didn’t demonstrate substantial improvement and two of their assertions were unfounded.

To start my critique, we need to consider the methodology. Staats and Upton studied just one company, Wipro Industries. This is a large Indian IT services company which has offices in many countries. They had a good population of projects (1,800+) to study when examining lean knowledge work versus traditional knowledge work. Even so, the results are underwhelming. The lean knowledge work projects – on average – came in 5% under schedule versus the traditional knowledge work projects that – on average –came in on schedule. The lean knowledge work projects – on average – were 9% under budget versus 2% under budget for traditional projects. It would be helpful to see the full datasets because I have a suspicion that when you factor in statistical error, any advantages of the lean knowledge work will be much smaller or even non-existent.

There are also numerous questions that call the results in question but my biggest objection is that this is one company and there may be something in the culture, work practices, labor force, etc. that could account for the differences. For example, maybe the lean projects had the more experienced project managers and teams assigned to them or the projects received more attention. Until you see the same results in a variety of organizations, all you have is essentially anecdotal evidence. And the plural of anecdote is not data.

Putting that aside, the authors give six principles they derived from their study:

  1. “Continually root out all waste.”
  2. “Strive to make tacit knowledge explicit.”
  3. “Specify how workers should communicate.”
  4. “Use the scientific method to solve problems quickly.”
  5. “Recognize that a lean system is a work in progress.”
  6. “Have leaders blaze the trail.” (Staats and Upton, October 2011)

Principles One, Five, and Six are also principles in Adaptive Case Management. Principle Four was not explained that well in the article but it sounded much like the core practice of ACM. This is where workers look for common issues that arise in a number of cases and develop templates to handle these common issues. Staats and Upton argue that workers should use analytics to routinize these processes which complements the use of social networking to collectively create the templates. So, out of the six principles, ACM incorporates these ideas just as effectively.

Principles Two and Three are the reasons why lean knowledge work won’t work. Staat and Upton essentially ignore the nature of tacit knowledge and assert that if you can’t make tacit knowledge explicit then you are not trying hard enough. This flies in the face of numerous research studies and the experience of numerous companies as they struggle with the issues of knowledge transfer and best practices. Given the dynamic and novel nature of knowledge work, I believe that ACM has a better method of transmitting tacit knowledge through social networking rather than forcing tacit knowledge to become explicit through traditional business process documentation.

Principle Three, structuring communication, exemplifies the inherent danger of Lean and Six Sigma. What Wipro did was create a table of who should communicate with who, when, where, and how. They even used matrix algebra to determine the optimum communication paths. When I first read this, I was horrified. How does Wipro know that their structured communication was any better than the normal communication that arises during a project?

They don’t. Look at their stats: lean projects were essentially no better in results than traditional projects. In fact, I am surprised that constraining communications as Wipro did wouldn’t cause even worse issues with the lean projects. In every study on project failure, poor communication is at least one of the top three reasons the project failed. I would love to see a study that demonstrated even one project that failed because there was overcommunication. Again, the advantage of ACM is that there is constant communication among all parties involved in the knowledge work which allows for better transmission of tacit knowledge and more effective process discovery as the case unfolds.

This also points to a major complaint I have about Six Sigma and Lean projects. They often force workers into a process that differs from how people actually perform their work. Software solutions are created that hinder the work as employees are compelled to “feed the beast” with data rather then being supported by the technology. I know practitioners don’t start this way but the very practices of Six Sigma and Lean often lead to these constraining practices.

Government needs to be more effective and efficient in its processes. Lean and Six Sigma look like the silver bullets but they only work when you have well-defined and repeatable processes. That is not the nature of government work. Government workers deal with cases that requires intensive knowledge work and that is why Adaptive Case Management is the better solution.

Staats, B.R., & Upton, D.M. (October 2011). Lean Knowledge Work. Harvard Business Review.

Leave a Comment


Leave a Reply

Profile Photo John_Conley

Bill –

You obviously know this topic well and have great passion for it. I’m sure that I’d enjoy reading future posts on the topic.

However, have to quibble with the current post — 1). The connections here to government are tenuous at best, 2). I don’t think that your counter-arguments are as well supported as they could be, and 3). we jump into the topic (on the Agile Development board) in chapter 11 of the story. Just to clarify – if someone has been following your thinking elsewhere, this might all fit together; however, for this audience, we’re missing a lot of context (e.g., how about a primer on why Agilists should care about ACM?).

Just my two cents anyways.

Profile Photo Bill Brantley

@John: Thank you. You might also enjoy my previous posts including the post on ACM that I wrote just before this one. That will provide the context for this posting and especially why Agilists should care about ACM. It would be a good idea to write a summary post in the future.

The connection to government is that I am seeing an increasing effort by consultants to sell government Six Sigma and Lean solutions. Six Sigma and Lean hasn’t caught on as it did in industry but I fear it will. As history has shown, Six Sigma and Lean has had mixed results in industry and I am concerned it will do the same for government.

Also, I am very wary of applying business techniques to government. There are some great business practices that government can benefit from but I feel that many advocates just want to transplant a business practice without first understanding the unique nature of government work. Knowledge transfer is rarely a direct transfer of best practices. There has to be some adaptation.

As to your criticism on my counter-arguments: could you be specific? I will be glad to further support my arguments. Thank you for taking the time to read and respond to my posting.

Bill Brantley

Profile Photo Bill Brantley

@Mark – that is a great article and I fully agree with its conclusions. I do have to admit that I am puzzled on how you saw the connection between that article and my posting. Could you explain? Thanks.

Bill Brantley

Profile Photo Preston G. Baker

I believe your overall statement is much to broad and while your premise may hold true in Human Resource Mgmt there are too many other areas, at least in municipal government where this is not the case. You state “Lean and Six Sigma look like the silver bullets but they only work when you have well-defined and repeatable processes. That is not the nature of government work.” The City of New York has a myriad of ,well defined processes that are repeated on a daily basis. I assert that it truly is the nature of government work for many areas of responsibility in government agencies.

Profile Photo Bill Brantley

@Preston: Thank you for commenting. I appreciate the viewpoint of state and local government workers. I was a state worker (Kentucky) back in the early 90’s and I am certain much has changed since then.

I do not doubt that New York City has a number of processes although I may quibble with “well-defined” when you view it from the customer perspective. For exampe, The Maryland Motor Vehicle Administration also has a number of well-defined processes that work well for the agency but is extremely frustrating for the customer. You may be familiar with Lipsky’s work on the street-level bureaucracy where citizens are transformed into “clients” through a series of processes designed to minimize the amount of time and personal attention given to each individual. This leads to greater citizen dissatisfaction, lower morale in the public service agencies, and a growing disconnection between citizens and their governments.

Surely, you are not arguing that the processes of the New York City government do not need improvement. What are the rates of customer satisfaction with government services? How effective are governmental processes in delivering services? Does the government of New York City provide good customer service?

My contention is that governmental processes need constant improvement to better serve the public. I was just advocating adaptive case management over the traditional lean six sigma approach but whichever approach you choose, I do hope there is a consistent effort to improve governmental services.

I am interested in how municipal governments could benefit from ACM and, if you like, I can send you some case studies on how other city governments are using ACM.