Sep 232016
 

No, even though 2016 marks the 50th anniversary of Star Trek, I’m not talking about James T. Kirk and the starship Enterprise.

zumwalt

DDG-1000, the USS Zumwalt (U.S. Navy)

I’m talking about Captain James A. Kirk and the USS Zumwalt, America’s “largest and most technologically sophisticated destroyer.” The Zumwalt grabbed my attention when it was commissioned earlier this month because not only is it larger, more powerful, and stealthier than current destroyers, it boasts a crew that is half the size of other ships in its class.

The story of this real-life, science-fact-not-science-fiction vessel and her crew is a great read, and not just because of the cool tech.  There are many lessons here for leaders of non-military “enterprises” such as software development, business, scientific research, and non-profit/NGOs.

In this series, I’ll point out just four.

Lesson 1. Technology & Automation Are Required

The Zumwalt doesn’t have gunners’ mates to operate her guns, a separate radio room for communications, or a helmsman to steer the ship. All these tasks can be performed by an officer standing watch in the Ops Center because these routine, manual tasks are handled by automated systems.

The technology of automation is quite a bit more capable now than it was a decade ago.

A good friend of mine used to work in “network operations security.” His team would spend their days in repetitive tasks that resembled manual labor. Inventorying hardware and software. Researching the latest threats online. Manually patching vulnerabilities wherever they could. It was a Sisyphusian and sometimes scary job, always trying to stay one step ahead of the hackers.

Zumwalt Ops Center (Raytheon)

Zumwalt Ops Center (Raytheon)

But today my friend’s team has software to automate the tasks they used to do. The tech hasn’t replaced the people, it’s raised their game. The software still needs to be told what to do in order to accomplish the why of the company’s risk management strategy. And it’s my friend’s team — now called “risk management” — who is responsible for accomplishing that “why.” The tech has enabled his team to deliver better, stronger services to their company.

Proper automation can enable humans to focus on tasks like synthesizing information, making good decisions, setting strategy, and employing their own technical know-how.  Skills that, you know, require a human.

Let the machines handle the repetitive mechanical tasks; they’re better at it. Don’t expect your systems to be able to handle too much variance and unpredictability; that’s what your people are for. Use the tech to raise the game of your peoples’ service delivery. Tech should allow your people to directly implement the overall goals and strategy of your organization.

If you employ technology this way, you can free up your smart, motivated people from the routine tasks of data-gathering and paper-pushing. You can unlock a pool of potential capability for your organization.

In the next article we’ll look at why automation and technology, while necessary, are not sufficient.

All articles in this series:

  1. “Raise the Game” — Automation is Required (this page)
  2. “Same Sheet, Different Data” — Tech is Not Enough
  3. Management Overhead — Reduce the Waste
  4. Team Skills Needed — Diversity, Communication, Mutual Support
Sep 222016
 

In the previous section, we explored how technology can potentially free up capability. On a high-tech destroyer or in cyber-security, tech can “raise the game” of people in the enterprise.

Technology offers tireless, rapid, low-variance behavior. That doesn’t necessarily mean more intelligent behavior.

Lesson 2. Automation Is Necessary, But Not Sufficient

green-bar-paperOne of my first jobs was loading magnetic tapes onto a backup machine and physically delivering printed reports around the office. (Yes, green-bar paper!) Occasionally I’d mis-load a tape or deliver a report to the wrong place and oh boy, would the phones would start ringing. Cool story, bro, but now we’re better than that, right? Today we’d consider this an egregious waste of human talent.

Yet we think nothing of employees spending hours fighting Excel. Sorting, querying, transposing. We haven’t raised the game. We’ve just made it harder to win. Technology gives you more of what you already have, faster.

“S.S.D.D.” used to mean “Same Sh*t, Different Day” but now it means “Same Sheet, Different Data.”

Image: Walt Disney Co.

Mickey’s automation problems in Fantasia (Walt Disney Co.)

If you add automation without engaging your people, you accelerate whatever already exists in your org. And you may accelerate the bad faster than the good.

If your enterprise (starship, sailing ship, or company) is full of reporting, bureaucracy, fear, or butt-covering, adding a “technical solution” may just amplify that.

I’ve seen far too many orgs drowning in data, yet thirsting for insight. Where people have become servants of the systems, not the other way around. And the mission is forgotten.

The crew of the Zumwalt avoided this by designing around the human side of the equation. Every new shipboard computer system was prototyped and adjusted based on feedback from the crew. Then they trained hands-on, side-by-side with the engineers to be ready to sail the new space-age destroyer.

Electrician’s Mate 1st Class Donald Goldsberry and Electrician's Mate 2nd Class Radarwin Adams of the pre-commissioning crew of the future USS Zumwalt (DDG 1000) train to use the common display system console and engineering control system screen navigation at Naval Surface Warfare Center Carderock Division - Ship Systems Engineering Station (NSWCCD-SSES) in Philadelphia (U.S. Navy photo by Public Affairs Specialist Joseph Battista/Released).

Zumwalt crew members cross-train shared tasks on a shared console (U.S. Navy)

So pay close attention when you add technology. Consciously prioritize the human issues above the technological ones. Use technology to free your thought-workers for thinking. Be ever alert to the failure mode of people “feeding the machines.”

As a leader, you need to enable your teams to move from what to why.  You need to enable them to make the same shift that my friend’s “network operations security” team made when they graduated to “risk management.”

Set clear strategy, manage the work, and let people self-organize around it. Or your organization’s relationship with technology might look a bit like Mickey’s above.

In Lesson 3: Management Overhead, we’ll look at how raising the game for your people re-defines the value of “management.”

All articles in this series:

  1. Raise the Game — Automation is Required
  2. “Same Sheet, Different Data” — Tech is Not Enough (this page)
  3. Management Overhead — Reduce the Waste
  4. Team Skills Needed — Diversity, Communication, Mutual Support
 Posted by at 5:17 pm
Oct 162011
 

I recently spent a day helping the leadership of an organization make the leap from “can we become Agile?” to “now we’re becoming Agile!”  The problem was one of mindset: getting people to think less about barriers and more about what actions are possible.  My solution involved hands-on experiential learning and structured group debriefing.

I’ve written about it below, but I’m a visual person so I’ve added a lot of captioned pictures if you want the gist, and the how-to “recipe” for doing this mindset-shift yourself is at the end.

Continue reading »

 Posted by at 6:59 pm
Oct 022010
 

Oh, the farmer and the cowman should be friends.
One man likes to push a plough,
The other likes to chase a cow,
But that’s no reason why they cain’t be friends.

image from http://www.youtube.com/watch?v=XB90b8xXYIkI’ve noticed a disturbing trend at a few different clients now as far as how UX folk and Agile folk get along.  It put me in mind of a song from Rodgers & Hammerstein’s Oklahoma, barely-recalled from my high-school musical days.  Let me explain by way of example:

Steve is a User Experience (UX) expert on an Agile/Scrum team working on the new HipStartup.com website.  He’s excited because he recently returned from an Agile conference and got a lot of great ideas from Jeff Patton’s session about User-Centered Design.  John is the Product Owner (PO) for the site and is juggling multiple voices and requests, but knows that each day without a new site feature delivered is another day that Venture Capital money is burned with no return on investment.  Geri is the ScrumMaster (SM) for the team and she wants everyone to follow the Scrum process because she knows it’s the best way to deliver value incrementally.

Three weeks ago at the meeting with the VC folks, John heard that they would really love to see HipStartup.com do single-sign-on integration with Facebook, Twitter, NetFlix, and the FAA.gov websites — with a unified look and feel.  This morning during the sprint planning meeting, developer Brian cautiously raises the issue of design: does Steve have any wireframes or web composites done for Brian to use in his estimation of story size?  Steve replies that he’s still testing the wireframes with their user focus group and wants to make sure the design for the new single-sign-on will work for them before he releases them to the rest of the team.

John explodes. “What? It’s been three weeks! You’re telling me we still don’t have a thing to demo yet?”  Steve looks helplessly at Geri — surely she will understand the need for a good design?  “John has a point,” Geri says.  “All this user testing you keep talking about smells a lot like Big Up-Front Effort to me.  Can’t you get some wires to Brian ASAP, even if its just for a few stories?”  Steve is appalled.  Wires for just a portion of the site, when the VC folks clearly wanted a unified look and feel?  “Well, sure,” he huffs, “I can give you what I’ve got so far, and you can get right to work on it… if you don’t mind the users thinking it’s crap!

Let us let the curtain of imagination close over this scene for now.

I’d like to say a word fer the farmer
He come out west and made a lot of changes —
That’s right! He come out west and built a lot of fences,
And built ’em right across our cattle ranges!

Just like the farmer and the cowman, User Experience folk and Agilistas/ScrumMasters each provide value in ways that can inherently interfere with each other.  The UX person wants a good experience, and so needs to engage in some sort of holistic design.  This doesn’t have to mean Big Up-Front Effort, but it can look that way to the Agilista who has been fighting waterfall for much of her professional life.

They also each work in ways which tend to stir up the fears of the other.  The Agilista wants a continuous flow of bite-sized chunks of value, with frequent opportunities for inspection and adaptation.  This doesn’t have to mean a Frankenstein’s monster of cobbled-together parts, but to the UX person, even the possibility of creating such a monster is, pardon the phrase, horrifying.

I’d like to teach you all a little sayin’
And learn the words by heart the way you should
I don’t say I’m no better than anybody else,
But I’ll be damned if I ain’t jist as good!

So at the #womeninagile booth at Agile 2010, I had the great fortune to start a conversation with @carologic of Ask a User.  Carol is a UX expert and works through a variety of methods to incorporate actual user research and experience into the design of her client’s products.  While as an Agile coach and empiricism fanatic, I’m always espousing the “ready-fire-aim” and “fail fast” schools of adaptive outcome development.  Cage-match waiting to happen, right?

Wrong.  In 20 minutes of conversation at the booth, we were able to express our respective concerns with each others’ practices and how they are misapplied, and brainstorm at least three harmonious solutions integrating the best of Agile/Scrum and UX, in very specific ways.  Not to tease, but over the next few months, Carol and I will share those ideas with you in our respective blogs.  We will explore:

  • why we shouldn’t call the end-of-iteration demonstration of incremental functionality a “demo”
  • where to integrate user research into the Agile iteration cycle
  • the use of Boundary Objects as discussed in Israel Gat’s Agile 2010 talk

and more.  How did we come to this resolution so quickly?  Simple:  we both approached each others’ domains from a perspective of curiosity and interest — rather than from one of defensiveness.

And when this territory is a state
An’ joins the Union jus’ like all the others
The farmer, and cowman and the merchant
Mus’ all behave theirselves and act like brothers!

image from http://www.youtube.com/watch?v=jBf1Bkk8Gdk

I believe that diversity does not mean “tolerance of differences.”  Rather, diversity means “our differences make us collectively stronger.”  This isn’t compromise, this is learning from each other’s talents.  As Karl Weick said, “fight as if you are right and listen as if you are wrong.”

Let’s us UX and Agilista people listen to one another.  We’re not truly in competition, we both want the same things.  We just have different ideas of how to go about it.  And it’s in the union of the two that true value will be created.

As Oklahoma‘s Aunt Eller says, “ain’t nobody gonna slug out anything — this here’s a party.


 Posted by at 4:22 pm
Sep 222010
 

Rebecca Porterfield asks “can Agile work in a culture of single point accountability?”  The question stumped a panel discussion.  Too bad I wasn’t there, I’d have been able to give the usual pricey consultant answer:  “It depends.”

(cue laugh track)

image from The Village Voice, http://blogs.villagevoice.com/forkintheroad/archives/2009/12/locavorism_gone.php

In all seriousness, it depends on what you mean by “work.”  I personally have very high standards for “working” Agile.  Agile isn’t a way of doing, it’s a way of being.  So do I believe Agile can work in the spirit of its greatest potential, as a means to enable the shift from an deterministic organization to a learning organization (a la Senge), in a culture of single point accountability?

Can a possum stop a car?

Accountability When Agile is a Process

Sure, Agile can work as a process for incrementally creating quality software product when we insist on maintaining single-point accountability, exactly as Rebecca successfully implemented in her project.  You can “wrap” the Delivery Team in Agile and use the Product Owner and ScrumMaster as the “interfaces” for the team.  You can even promote individual accountability in such a scenario:

  • Product Owner (PO) — on the hook for the health of the (software) product
  • Team Coach/ScrumMaster (SM) — on the hook for the health of the Team and their adoption/adaptation of the process
  • Delivery Team — on the hook for demonstrating a working product increment at the end of each iteration

“Hey Derek, not too fast! That last one is team accountability!  So there isn’t individual accountability at the team level!”

Uh, no.  There certainly is.  In the context of the entire Agile Team (e.g. a Scrum Team consisting of a SM, PO, and Folks Who Make the Software Product Exist) the Delivery Team is accountable only so far as they collectively build working product.  Individually, however:

  • Delivery Team Member — on the hook to the rest of their Team for whatever commitments they make each day at standup; for asking for help when needed; and for swarming and providing help as needed.

See ?  If we are talking about a culture of accountability, then accountability depends in what context you view it.  Each person both has individual accountability and also is part of a Team that has team accountability.  In the context of the Delivery Team, Brian the Developer is accountable for the tasks he signs up for and tells his team “by tomorrow I will complete…”  In the context of accountability to the rest of the organization, however, there is no individual accountability at the Team level lest we promote siloed work and blame-shifting and all the other joys of WaterGile.  But in the same context of the rest of the organization, Geri the PO is the single-point of accountability for the “health of the (software) product (that the team delivers every increment).”

But can Geri truly be single-point accountable for the R-O-I of the product?  Because she is also part of the larger team of Product Management, which is team-accountable for making that R-O-I manifest and be something the organization can use.  In the context of the total utility of the software product, one must also consider the teams of marketing, sales, support, training, etc. to extract and deliver full value on the potential of the software product.

So it’s both.  In a single-point accountability culture, the successful Agilist who wishes to use Agile as a process for creating potentially useful software products needs to promote team accountability and individual accountability as appropriate for the context.  Asking which is more important is like asking whether “3” or “blind mice” is more important.

Lowest Common Denominator

Image from http://seoblackhat.com/2008/05/19/to-make-a-fortune-cookie/ and http://www.isaacsunyer.com/dont-be-evil/

Now, this will work for churning out product at a reasonable rate, and Fred Winslow Taylor and members of his Fan Club can clap and sing that Agile will work in a culture of single-point responsibility.  Assuming that your organization has little commitment to the change necessary to achieve the hyper-productivity promises of Agile, assuming you don’t mind flat or near-flat velocity gains, flat team morale, and assuming you are okay with the criminal waste of human potential of churning out product with no real Awesomeness in it — please, be my guest.  Don’t look back, some other companies might be gaining on you.

But if this doesn’t seem to align with your organization’s values, if you actually want people to work better together, to develop their full potential as a team, to realize the power of their diversity and make some truly creative and useful things… you may wish to be cautious with single-point accountability and Agile.

Because you’re robbing people of responsibility.

Responsibility — being a source or cause for results — depends on having the opportunity to be the source for those results.  When you make one person (e.g. the PO, or the SM) on the hook for delivery of the product, people who don’t have the ability to deliver that product, you steal that opportunity away from the Delivery Team.  And that’s the greatest paradox of how I so often see Agile being implemented.

Trying to realize these kinds of gains while preserving single-point accountability for delivery will be about as successful as the possum trying to stop the car.

 Posted by at 8:11 pm
Sep 092010
 

I had sent an “attaboy” email to one of the developers on my team for doing a great job in communicating and being visible to the rest of the team.  (Don’t punish bad behavior, “catch” and encourage good behavior, right?)  Not having previously received encouragement for anything less than heroics, the developer was perplexed upon receiving my email.  Was I upset with him?

No, I explained, I just noticed he was being visible and wanted to encourage him to keep it up.  “Consider that email a round of applause,” I said.  Brian (another developer on the team) overheard this and then told us of his “Praises and Curses” email folder.

Praises-and-Curses

Pardon?

As it turns out, Brian collects emails from people when they thank him for doing a good job, or when they are upset and wish he had done something differently.

But he doesn’t just collect them — he reviews them.  At least once a year prior to performance reviews.  And sometimes more frequently.

Now, just to put this in context, Brian also uses the Pomodoro Technique in his daily work.  So what we have here is an individual — a so-called “technical” person, no less — who:

  1. works in discrete intervals, identifying interruptions and tracking velocity
  2. inspects the feedback received as a result of his work

I gaped.  What an incredibly good idea: building in the means for self-improvement that just sort of “happens.”  No need to depend on the performance review as the only means of formal retrospection, simply collect the information as it comes in, and review at your leisure.

How many of us set up these kinds of mechanisms in our lives — feedback mechanisms which help provide awareness?  What would life be like if more of us did this?

 Posted by at 12:10 pm
Sep 062010
 

So, I’ll be participating in Scrum Beyond Software in Phoenix later this September.

I have a bubbling head full of ideas to share there, and as a collaboration junkie, I’m making them visible for comment.  Suggestions? Criticisms?  Want to join me in Phoenix and collaborate?  Leave a comment!

Topic 1 — Science: A Framework to Aid Scientific Research Teams

Scrum is, at its heart, a simple empirical framework for learning and discovery.  Often it’s used to “discover” the unbuilt-but-needed features of a software product, and so people confuse it with a product development methodology.  But it’s more than that — Scrum is also useful for process improvement or organizational change.  Scrum is even self-modifying: it can be used to “discover” the best way to shape the framework itself to help make the team using it more effective.

When you generalize Scrum this way, it seems pretty obvious to try to apply it to how teams “do science”:

  • backlogs as analogs of hypotheses
  • swarming as part of research
  • team-synchronization via standups and lo-fi tools rather than long publishing processes
  • scrum of scrums to synchronize multiple teams
  • demos and retrospectives as analogs of findings and conclusions

These are fairly naive mappings; I hope to make richer ones in Phoenix.

Topic 2 — Healthcare: A Framework for Training Medical Teams

I first mentioned this in a Twitter post inspired by discussion at the 2010 Conference of the International Association of Medical Education.

It seems to me that much of the thought in medical education views education as a rather linear, deterministic process, using what Bob Marshall calls “analytical models,” as contrasted to empirical, stochastic or even chaordic models.  Example:  a “learning process” depends on:

  • a set of initial “learning outcomes”
  • guided “deliberate practice” where students work either on medical tasks, or simulated scenarios (as a team, no less!)
  • “reflective learning” where students debrief and self-assess.

To a Scrummer like me, this just screams “backlog, whole-team execution/delivery, retrospection, REPEAT” and also embracing student/team learning as a chaordic rather than deterministic process.  Unfortunately I see some of the literature and discussion in the medical education world take a rather “one-shot” approach to training:  Students come in, execute, and now they are “trained.”

Some topics and questions I’d like to raise in Phoenix:

  • How might we frame learning in the context of Scrum?  By individual lesson (sprint), and as an entire curriculum (release)?
  • Comments from people in the medical training/simulation field? How have you used something similar to Scrum?
  • What are the problems with taking an approach driven by feedback loops rather than a stepwise process?

Topic 3 — Healthcare: Patient-Centered Treatment

Inspired by Compassion as a Golden Rule for Healthcare and Real Participatory Healthcare Starts With Assigning the Patient to Your Team (same author)

I need to give more thought on this — or ideally jam with another collaboration junkie — but the nebulous ideas here are:

  • Linda Rising’s talk at Agile 2009 about how people who solve problems together, despite their backgrounds or knowledge, feel more empathy and understanding for each other
  • How this effect brings diverse skillsets on an expert team together, despite egos and fears, and makes the team more effective at problem-solving
  • Mapping the patient, and patient’s family into the role of “user”; mapping the medical staff into the role of “team”, and then:
  • Using the Scrum framework as a structure in which to allow these roles to interact.
  • How this is very different from the standard model of “lead physician interacts with patient/family, and specialists mostly  interact with lead physician”
  • Who should be the ScrumMaster and Product Owner?
 Posted by at 5:57 pm
Mar 082010
 

Edit: This should properly be titled “Dispersion and Implosion in Teams.” See Tobias’ comment.

Exercise at ScrumGathering 2010: how simple internal models (“rules”) can have very different effects on team behavior.

In the first situation, each person has to use their “best friend” to protect themselves from their own “worst enemy.” (In this case, you “protect” yourself by moving so that your “friend” is between you and your “enemy.”)

Notice how the group fragments and disperses.

In the second situation, there’s one little change: each person has to protect their “friend” from their “enemy.”

Bit of a difference! The overall behavior is toward cohesion.

P.S. “Kids, don’t try this at home!” This exercise works because the participants agree to the rules. It does NOT imply that you can simply give people rules to follow and expect to get the desired behavior. Why not? Because people aren’t machines, that’s why not! :) The art of the team is, of course, in coaching and coaxing the teams such that the individuals experience a shift in their own “internal rules.”

Best Friend Worst Enemy

 Posted by at 1:30 pm
May 242007
 

A colleague writes:

Over the past few weeks, I’ve been digging into links, blogs, forums using this term (Post-Agile)… Have you seen this term? Followed it’s discussions? Any comments?

…Pliant Software is another term that seems to be associating itself with Post-Agile…

I think the software world is sort of like a customer who has an idea of what they want or need, but can’t put it into words. But, and after several iterations of “Requirements Gathering”, has come out of the room with more terms and an RD that is better than it was at the start — but not perfect, yet!

(italics mine)

Yes, I have comments!

  1. I agree that the newness of agile has worn off, and the immediate benefits from it have been tapped dry by many.
  2. I’m annoyed at the PliantAlliance site’s claim for “a new way of thinking about developing software” — pliant sounds like adaptive to me, and Scrum has been saying this for a while.
  3. While the immediate benefits of agile have been tapped, the deeper benefits that can be realized by changing the way we collaboratively build product — not just the way developers write code — have a long way to go. Before we decide that “agile didn’t work,” I’d like to see more organizations actually practicing basic agile behaviour, and I’d like to see actual practices in use catch up to the huge body of literature describing deeper agile practices.

Good Software is a craft, like smithing or theatre. There was never a school that turned out expert blacksmiths; novices had to progress from apprentice to journeyman to master. The “Fame” school doesn’t promise its graduates fame, they have to go out and earn it through experience.

Good Software is the result of human thought — not just human labor — and is highly resistant to being made into a detailed process/algorithm, or body of knowledge.

Good Software is the result of collaboration. Social Science and more enlightened ways of interacting can help, but trying to describe them is missing the point.

I understand that post-Agilists see themselves as giving a name to an existing movement, rather than trying to create a new movement. But it smacks of picking a new name for your club because one member of your club is acting silly, and people are now making fun of you: you fragment the group, you tacitly approve of the bad behavior by distancing yourself from it rather than correcting it, and you spend too much time and energy on labels rather than action.

My friend’s point is right on: the SD world is like a user with a poor Requirements Doc, who “knows what they want, they just can’t describe it yet.” Trying to describe it, trying to codify it, is ulitimately a losing game and a waste of time — just like writing big RDs is a waste of time.

Build successful, adaptive teams who can make good, useful software, point to them and say “THAT is what Agile is.”

Apr 162006
 

Facilitators, has this happened to you?

You have a nice four-hour block for your working requirements / product planning / process re-engineering / whatever meeting. You allot a “generous” 30-minute chunk near the beginning for the definition of terms. You put a few domain-specific words up on a flipchart to get ideas flowing, point to the first term, and prompt the team:

“So, everybody, what is the definition of _______ ?”

(Two hours later…)

DefinitionsOne phrase is detailed to a precision that would satisfy most German engineers, the total number of terms to define has tripled — but none of them have been defined — and you have had to break up three near fist-fights.

A common vocabulary is important, but does it have to be this hard? I recently had a delightful experience which suggests that it does not.
Continue reading »