Lessons Learned From Five Years of Agile Implementation Failures – AgileDC 2013 Presentation

<

p><img class=”size-thumbnail wp-image-1142″ alt=”Picture of me taking a picture of Sprint Zero of the Wikispeed workshop courtesy Elinor Slomba of Arts Interstices.” src=”http://test.devinhedge.com/wp-content/uploads/2013/10/2013-10-08-09.58.52-150×150 Picture of me taking a picture of Sprint Zero of the Wikispeed workshop courtesy Elinor Slomba of Arts Interstices.

Well, another year of AgileDC is in the can. This year was another winner. Even though the flavor at AgileDC is always biased towards the Federal Government, it was strange that the topics seemed to be diverse and more engaging than those at the Agile Alliance‘s Agile 2013. I confess that the topics at Agile 2013 were so non-interesting that I didn’t even go this year. This is not to say I didn’t miss something. I did. First of all I missed hanging out with old friends. There was also a few sessions that I would have liked to attended. For the money, though, AgileDC was a much better deal. Additionally, we raised some $14,000 for a cause I’m deeply passionate about, the Juvenile Diabetes Research Fund.

Wikispeed Keynote

This year, the keynote was just as engaging as last year. Joe Justice of Wikispeed filled in the gaps between Agile 2012 and today. Always willing to put others before self, Joe brought J.J. Sutherland* of Scrum Inc. to talk about how he used Scrum to manage NPR’s coverage of the Arab Spring in Egypt. According to J.J. the situation was so fluid that rather and unifying the reporters, it had a tendency to put reporters at odds with each other, causing missed deadlines and misinformation. J.J. spoke of being reminded of a technique his father forced him to learn by attending a Certified Scrum Master course: Scrum. J.J. talked about pulling out sticky notes and pulling the reporting team together twice daily. It worked and NPR’s coverage remained some of the most relevant and comprehensive. ( Transparency: I contribute funds to NPR so it is good to hear that my money is being managed well. )

Finished product after three one-hour Sprints.
Finished product after three one-hour Sprints.

Another thing from Wikispeed is the phrase “eXtreme Manufacturing”. I like where Joe and the gang are going with this. It has all the makings of changing the world in the same way that Demming did. Yes… I just went there. Expanding beyond building a car that is posed to reinvent how cars are designed and built, Wikispeed is starting to focus on another one of my passions, solving the problem of involuntary homelessness using eXtreme Manufacturing to build MicroHouses. One application I could see of this refugee camps, displaced peoples from natural disasters, and a way for cities to set up transition programs for those placed in involuntary homelessness situations. (NOTE: I probably should talk sometime about what we are learning about why these programs fail and how Habitat for Humanity has overcome these obstacles to success.) Throughout the day, Joe and the folks at Scrum Inc. used the Wikispeed eXtreme Manufacturing workshop to teach pairing, eXtreme Manufacturing, Scrum and Kanban.

Personal Experience

Ballroom B where my session was at AgileDC.
Ballroom B where my session was at AgileDC.

My session went perfectly. I’ve never had that happen before so I thought I would make note of it. I was presenting on the topic of Agile Failures, something no Agile Coaching account manager or business development person is likely to ever talk about. I expected the session, Lessons Learned From Five Years of Agile Implementation Failures, or… What NOT to Do When Becoming Agile, to have about ten people show up. Ten minutes before the session, it was full. Five minutes into the session, it was standing room only.

Needless to say, I was nervous. This was also my first public appearance under the ResultLinq Associates monicker.  Would the audience get the message? That will still remains to be seen, but the feedback was overwhelmingly that I hit home. The feedback told me exactly what I expected. Everyone liked the format. The opening blew everyone’s mind. A lot of people were stuck in deterministic thinking headspace so they wanted a one-size-fits-all checklist when all that could be had are certain principles. Oh… and I thought the projector/screen combination was terrible, too. It made the smaller text unreadable and I was standing right in front of it.

I have to confess that I lied on one slide. The picture of an Agile adoption coaching plan was actually a release planning session. I couldn’t find my coaching plan pictures and had to substitute with something worked and looked the same as a coaching plan. I openly apologize and ask for forgiveness. I found the pictures today after some creative searching through my Dropbox history. I’ve updated the slide so that you can actually see what my coaching plan wall looks like.

Several folks have asked for the slides. I did one better this time around. Below is a corrected recording of the session and a link to the PDF of the corrected slides.

[vimeo 76835486]

PDF File

Feel free to ask questions and challenge some of my hypothesis and theories.

Enhanced by Zemanta

Overcoming the role of cognitive bias in critical decision making

Folks attending Agile 2013 got a treat this yesterday. Former colleague Manoj Vadakkan and Agilist Bob Payne are going to address how cognitive bias plays a role in Agile Adoption

English: GCOM Cognitive Schemata
English: GCOM Cognitive Schemata (Photo credit: Wikipedia)

. In researching and garnering feedback, Manoj and I were swapping anecdotal stories of cognitive bias. I’m going to share two here. If you happen to be at Agile 2013, try to catch this session as I know it will be great.

The two examples I want to share where I have seen cognitive bias preventing Agile Adoption are what I’ll call Hourly Estimation Bias and Risk Adverse Homosocial Reproduction. It sounds fancy, but your see it is just a situation we take for granted not knowing its adverse effects in catalyzing necessary changes.

HOURLY ESTIMATION BIAS

This bias can be defined as the need for managers who have not fulling embraced “The Agile Way” of Empirical Evidence in planning and tracking. In Agile practices we tend to shift away from hourly estimation of way, relying instead on relative estimation based on NUTS and Throughput Account measures such as Velocity, Cycle-Time and Lead-Time. While it is often dangerous to apply Lean Manufacturing metrics to Product Development due to the vast variation in effort from Product Feature to Product Feature, when sufficiently broken-down into work units that can be completed in 2-3 days, the law of averages in large datasets takes over giving you a nice Guassian Curve.

The anecdotal story I’ll share here is that managers, being unfamiliar with empirical evidence and having relying on vanity metrics so long, they suffer bias towards the vanity metrics even when we have proven that the Paredo Principle applies to these estimations and metrics… they are only correct ~20% of the time. Or stated another way, vanity metrics of Scope, Schedule and Cost in a complex human system are invariably WRONG roughly 80% of the time.

So why would a manager bet their bonus and possibly their job on something that it WRONG 80% of the time? Familiarity. The brain naturally filters and reduces the world around us into simples terms in order to perform sensemaking. This works when you need to know if a Lion is a threat while walking across the Serenghetti. It works poorly in estimation of complex systems and complex human system. Yet, time and again I have managers that ask: “What is your percent complete?” and “How many hows do you have left?” Why? Cognitive Bias towards the familiar. It turns out we favor five-to-one something we are familiar with over something we are unfamiliar with. Paredo’s Principle strikes again (One-Fifth = 20%).

So how can Managers overcome their natural bias towards the familiar? Managers should be required to read and peer-review two books that span a large body of knowledge about neuroscience, critical analysis and decision making. These two books are, “Your Brain at Work: Strategies for Overcoming Distraction, Regaining Focus, and Working Smarter All Day Long” by David Rock, and “The Power of Habit: Why We Do What We Do in Life and Business” by Charles Duhigg. In both books the skills of mindfulness (understanding when your brain is reacting again reason in its need for the familiar), and how to form new habits so that the unfamiliar becomes familiar.

RISK ADVERSE HOMOSOCIAL REPRODUCTION

The second cognitive bias I’ll discuss is a phrase I’m hijacking and twisting, called Risk Adverse Homosocial Reproduction. In a study about diversity in the workplace and the hiring habits of managers, it has been found that managers subconsciously hire people that act and look like themselves (homosocial mirrors), and that they favor candidates that are homosocial mirrors over candidates that are more qualified.

I have seen this behavior present itself in the following more times than I am even aware of. For example, say a manager needs to fill a position with someone that has Agile skills. Together, we look to hire someone internally first because it is seemingly less risky as a person is likely to already understand the political landscape and challenges facing the position. But, what often happens in a command and control environment shifting towards a collaborative environment? We look around the company for a likely candidate and don’t find the right mix of technical skills, domain knowledge and soft skills. We have to hire someone from outside. However, when faced with a marginally competent internal candidate and strong external candidate, managers favor an internal candidate over the outside candidate. Thus, the command and control culture is further cemented into place and change becomes that much more difficult if not impossible.

Soft skills (people skills) in an Agile environment are typically the more important than technical skills and domain knowledge. That is not to say that a person should not have a base competency in the requisite technical skills or domain knowledge; however, we have found that a person with the soft skills of mentoring, servant leadership and lifelong learning can overcome not having strong technical skills or domain knowledge.

So why would a manager purposely choose the lesser qualified candidate?

It’s called Risk Adverse Homosocial Reproduction: a cognitive bias that favors hiring people that are “just like me” in order to make you feel good. This feeling of comfort comes from patterns of familiarity in the Superior Parietal Lobe combined with the fear of the unfamiliar in the Amygdala overriding our ability to reason using logic in the Pre-Frontal Cortex.

Unfortunately, knowing this is mostly useless. Typically, the people that need to know how to steer around this Cognitive Bias the most, are the least likely to know that they need to steer around this Cognitive Bias.

Русский: Cognitive Hazard by Arenamontanus
Русский: Cognitive Hazard by Arenamontanus (Photo credit: Wikipedia)

It is also the single, most-challenging, and most-obvious “problem to everyone but the person suffering this bias” that I face when coaching.

We all naturally suffer from it in some small way because it is a natural defense mechanism built into the brain against wild animal attacks. It is also the root of racism, prejudice, and class (Caste?) politics.

The bias is that strong. There is one way to overcome this bias, but it is quite ugly.

Generally, the only way out of hiring bias is through judicious use of external recruiting, allowing internal wannabes to apply for the position through an external vetting agency. Corporate recruiters hate this strategy because they feel they don’t have control over the vetting process, so a certain “make HR feel part of the process” tactic has to be employed. Properly engaged, HR becomes your strongest ally in the process.

I’ve successfully used this technique for overcoming my own bias. The result was being able to work some of the most brilliant people in the business.

What examples of cognitive bias in decision making have you seen?

Enhanced by Zemanta