Tim: Crash and Burn

When is it ok to break the rules? Michael and I wanted an early example of breaking the rules for good reasons, but in a way that then causes the next crisis.

As a friend loves to say, "Every choice has tradeoffs." Sometimes we change the rules to fit our team. Care for the people we lead is highly admirable. I love seeing that.

But solving one problem always causes new problems or makes other problems worse. Which things are okay to make worse, and which changes are causing our troubles? Agile isn’t sacrosanct, but the rules are there for a reason. As Angela advises in the book, if we’ve never done the exercise “by the book”, we should do that first so we can experience it as intended. After that we can see the difference when we customize it.

---

An excerpt from "Shift: from Product to People", Chapter 2, "Launch Missiles!":

Tim fired up his computer for the Monday morning video call. It had been a great weekend. He was even able to take the boat out and do some fishing with his son. He closed his Appreciation Journal, took a pleasant, calming, deep breath, and connected to his first video call. “Good morning, Angela! Thanks for fitting me in. I promise to keep it to five minutes.”

“Good morning, Tim!” enthused Angela from Tim’s screen. “No worries, and thank you for recording the status call.”

“So, what’s the verdict?” Tim responded, jumping the gun.

“Can I start with a piece of advice for next time?” she asked.

“Sure,” Tim agreed.

“You… took some liberties with the Fist of Five there,” she began.

“What do you mean?” challenged Tim. “I know I skipped the countdown before throwing up the fingers, but this team doesn’t have any problem with being swayed by seeing the votes of others. Trust me.”

Angela guffawed. “I can see! Ok, that’s a minor thing. Does… what’s his name?... Don? ...know how to do Fist of Five?”

Now it was Tim’s turn to guffaw. “Ah! Don is a character, isn’t he? He’s been told, but he’s also a bit of an iconoclast. He loves going against the grain. He always adheres to ‘the spirit of the law’, so I’ve decided to just go with his flow.”

“Ok,” conceded Angela. “Then I loved that you encouraged Linda to talk more! But I was concerned that Roy didn’t get to talk at all.”

“Roy,” started Tim. “Roy’s a handful. He’s never shy about expressing his feelings, and I try to head off unnecessary fights when I can. Trust me, we already heard his opinion, and we didn’t miss anything.”

“Ok,” said Angela cautiously. “So my main issue is that the Fist of Five is not supposed to be a vote. Nobody gets outvoted in a Fist of Five. You have to move into risk mitigation mode and create contingency plans to address the concerns of those voting two and below until you can get them to come up to at least a three. Fist of Five is a risk management technique, not a voting technique.”

“Roy will not give up until he gets what he wants,” answered Tim. “I’ve tried. He won’t compromise.”

“It makes me uneasy,” Angela insisted. “I hear him using good words, like ‘hypothesis,’ and I’d like to hear more about the data he wants to collect. I don’t want to say it, but without a risk plan to address Roy’s valid points, I’m afraid that’s going to blow up in your face.”

“Wait,” Tim interrupted. “I’m getting a third text from Evah. I better switch over to the team call. Can I get the rest of your thoughts later?”

“Ok,” agreed Angela. “Good luck!”

“Thanks!” Tim ended the call with Angela, pulled up his calendar and connected to the Monday morning video call. The audio connected just in time for him to hear Naomi yelling, “Our requests are down 20% in just one weekend?! Someone is at fault here! Where was the mistake made?! I need answers now!”

Ali jumped in, “We’re still researching.”

Roy quickly added, “I told you to get the data first. See, Tim? This is what I was talking about.”

“Hi, Tim,” they all chimed in simultaneously.

“Welcome to the show,” added Evah.

“What’s the latest?” Tim queried.

Naomi obviously couldn’t contain herself, “Roy, getting your stupid data doesn’t matter here. Your insistence that your data is going to give us all the answers is… and I’m being charitable here... ‘an exaggeration’. We’ll get you your data, but first we need to get our user base back and growing.”

Tim was a little stunned by the harshness of Naomi’s words. She had always been tough, yes, but in a cool, calm, and collected way. Not like this.

Tim jumped in quickly, “Roy, I want you to know that we recognize your viewpoint, and that you have a valid perspective. We recognize the power of building smart analytics to understand and harness the needs of our special clientele. However, we have to get the basics down now,

and then we can use data to refine our software.”

Roy muttered, “It’s the other way around,” but kept silent after that.

Tim continued, “Let’s figure out the root cause as to why the release backfired. Was any part of the release untested?”

Evah shook her head no as she took over the screen, “Nope. Every part of the code that was changed had active tests. None were disabled. And all of them are green.”

Tim tried the next thing in his mental checklist, “Was there a missing library?”

“Nope,” responded Ali, “we would have caught that in the build report, and with our robust continuous delivery pipeline, that would never make it to production. Also,” continued Ali, tackling the next item on Tim’s list, “we’re following the ‘Brent Protocols’ we learned in The Phoenix Project: no production machine can be directly touched, so there isn’t the risk of

a production defect.“

Tim prodded, “What else?”

Jayson added the tech perspective, “Let’s see. We have universal error handlers, plus static and dynamic code scans, and they all have passed. As Evah will attest, all code goes through the same scripts when promoted from our QA to Staging to Production, and we have an extra

smoke test in production.”

“Yep,” confirmed Evah, “our smoke testers have tried many different scenarios and found nothing technically wrong.”

Jayson continued, “Okay, here’s a compromise. Why don’t we revert back to our previous version, add some basic metrics to help us discover the root causes, fix those causes, and re-deploy?”

Naomi began, quieter this time, but firmly, “No. We aren’t going to back out of this release.” Then she scrunched her face in a strange way and her eyes got wide, “I know! Let’s have a full-fledged User Acceptance Testing and we will quickly get to the bottom of this issue! Evah, I’ll need your help!”

Evah responded, slightly wide-eyed, “You got it, boss! I’ll round up my best testers to start on this today!”

Tim chimed in, “Great idea, Naomi! Don, Linda, please bring in people to help out with this UAT. Hey, we’re Agile. We know how to pivot, right? But,” he added, pausing briefly, “while this seems reasonable, I can only give you two weeks. If we don’t have the root cause, and the results aren’t improved, we will revert back and follow Roy’s plan. Two weeks. That’s all I can give you.”

As the team switched to planning out the UAT, their voices faded into white noise for Tim. His memory took him back to the day when he first had the idea. That was a great day. He was on the golf course, four under par, with two holes in one. The ideas were flowing. His golfing buddies were on board. They all agreed that it would be a ‘breeze’ to own the market space. That’s even how the app got its original name, “Breeze Transfer”. He longed for those days when it felt like it would be a breeze.