Agile Meridian

Your decisions are slow. Adding another team never fixed that. Year after year the engagement survey says the same thing, and the answer is always one more team, one more layer, one more hand-off. Nawaz Butt named this on The Meridian Point.

He has spent years inside some of Canada's largest organizations, and he kept seeing the same complaint land in the top five like a groundhog day: decision-making is slow. Companies chase scale, so they keep adding teams, and every team they add is one more set of hand-offs a decision has to cross before anyone can act.

More teams do not speed the work. They move the decision further from the people doing it. His fix is the opposite of scaling. Push the decision down to the team that owns the work and let them make it without a committee. He told a story about one team he was sent to fix. It had run a full year without a manager.

Self-organized, covering for each other. It did not need him. The move was not to add structure. It was to get out of the way. That only holds if the right people are in the room when the call gets made.

I wrote the measurement version of this on the blog, Why Your Estimates Are Wrong: It's Access, Not Data.

Where does a decision travel furthest from the people who actually do the work in your company?

🎙️ NEW EPISODE: Most Disruption Stories Have a Happy Ending. His Isn't Written Yet.

Jeff Giesen built and led teams of Release Train Engineers, taught SAFe, and drove enterprise-scale transformation. Then the layoffs came, and he was on the other side of the change he used to lead.

He didn't wait for the next role. He went all in on a real estate business he'd built since 2011, moving to ground-up townhome construction.

Then that market softened too.

Two disruptions at once, tech and real estate, both tying up his capital and his next move. Most people tell these stories after they make it out. Jeff is still inside his.

Jeff Giesen joins Kumar Dattatreyan, ICF PCC to explore:

✦ Why an agile transformation that worked still got cut as overhead ✦ What it feels like when two markets soften at the same time
✦ How AI-driven applicant tracking has rewritten the job search
✦ Why networking still beats the algorithm
✦ What the trough teaches you about managing fear before it manages you

"I didn't quit the company. They quit me."

-Jeff Giesen

Listen:
Apple: https://lnkd.in/gb92KnYQ
Spotify: https://lnkd.in/gbR5YzcW
YouTube: https://www.youtube.com/live/mdErMWJCL6Y?si=1F1SvG3HJGSKwH6U

Connect with Jeff: giesenhouse.org Book time with Kumar: https://tidycal.com/coachkumar/30-minute-meeting

I spent years calling meetings to get everyone aligned. Alignment was never the thing stopping us. I was avoiding the one call only I could make: who gets to decide without me in the room. Being the person every decision ran through was my strength.

It was also the ceiling. The team waited on me because I trained them to. Every meeting I called to get aligned was really a meeting to keep the decision with me. The plateau was not caused by what was broken. It was caused by what was working.

My judgment was good, so I kept using it, and the whole thing stayed exactly as fast as one person could think. Self-disruption here is not another framework. It is giving up the thing you are best at. You hand a real decision to the person closest to the work and you let them make it without you. Sometimes they make it worse than you would have. That is cheaper than being the bottleneck for another year.

So I stopped asking for alignment and started naming the decider. Then I got out of the room. The teams got faster the week I stopped attending my own meetings. Who is waiting on you for a decision they are closer to than you are?

Your delivery teams got fast. Your funding cycle didn't. That gap is your ceiling, and no amount of team-level agility raises it.

Ten years of investment went into making teams fast. Stand-ups, backlogs, coaches, tooling. It worked. The teams got fast.

Then the work hits a budget that moves once a year. A governance board that meets monthly. A hiring loop nobody can shorten. The team ships in two weeks and waits months for a decision about what to ship next.

Evan Leybourn runs the Business Agility Institute. He put it in one line on the podcast: an organization can only be as agile as its least agile function, and that's not technology anymore for most organizations.

Not technology anymore. The constraint moved. Everyone kept optimizing the part that was already fixed.

You don't have an agile problem. You have a slowest-function problem.

Find your slowest function. Not your slowest team, your slowest function. Then ask one question about it. How long does a decision wait in there?

Fixing the fast part again is the most expensive way to stay slow.

What's the slowest function in your company, and who's allowed to say so?

The hardest people to coach aren't the resistant ones. They're the ones who turn every piece of feedback into proof they're already on the right track.

Resistance is easy. You can see it. Someone pushes back, you know where they stand, and you have something to work with.

The skilled ones do something else. They listen. They agree. They restate your point better than you made it. Then they fold it into the story they already had. Nothing moves. The meeting felt useful. The belief underneath never got touched.

Ryan Behrman builds feedback systems for teams. He said it plainly on the podcast: every strength at some point will become overplayed. The person who's very determined at some point will become pushy.

Reframing is a strength. A good leader uses it to stay steady when the news is bad. Overplayed, it becomes a filter that lets nothing in. Feedback doesn't bounce off these leaders. They absorb it and neutralize it.

Watch for it in your next one-on-one. Say the hard thing. Then count how fast they agree with you. Speed is the tell.

You can't hand someone self-disruption. They have to be willing to lose an argument with themselves.

A short quiz walks the opening question. Comment and I'll send it.

95% of AI projects fail to deliver meaningful returns. I read that number to Ashwini Kumar on The Meridian Point. He has spent six years building machine learning inside manufacturing and telecom, and he did not flinch at it. Expectations far exceed reality, he said. Then he described what he keeps walking into.

A company decides on five AI solutions and assigns a team to each one. The teams build them without knowing which problem they are solving. The five systems never talk to each other. By then too much money is in the ground for anyone to say it out loud, so the numbers get fudged instead. His frame stuck with me. AI is another solution to the problem, not the solution.

Another tool in the tool set, and the most expensive one in there. Here is the part underneath the failure.

Between the layer that decides what matters and the layer that builds it, someone has to own the flow of work across that line. When nobody does, the build teams never see the decision, so they automate whatever is in front of them. Upstream it looks like a discipline problem. Downstream it looks like a capacity problem. It is neither.

I wrote the long version of this in Stop Automating Broken Work. Your AI pilot did not fail because the model was weak. It failed because you pointed an expensive machine at a process that was already broken, and it broke it faster. So start from the decision, not the tool. Name one decision your company makes badly and often. Measure what that decision costs you today. Then ask whether a model, a plain automation or a deleted step is what actually fixes it.

What is the AI project in your company that everyone knows is dead and nobody has buried?

I've taught product thinking for years. The breakthrough is never the framework. It's the moment someone turns the lens on themselves.

I watched it again with a recent cohort, three weeks into the course. By the end almost every one of them had gone back and reworked the assumption they walked in most sure of. Not because I told them to. Because they finally saw it.

That's self-disruption. It isn't a mindset you pick up in a workshop. It's what happens when you stop defending the version of you that got you here.

Here's what made it click. We built the AI in the room to push, not to answer. Before it gave anyone anything, it made them show their work. What did you actually observe. What did you just assume. It kept asking until the assumption came apart in their hands. You can't argue with your own answer sitting there on the screen.

Once the assumption goes, the product mindset shows up on its own. They stopped starting from their solution. They started from the customer and the outcome. That's the shift. Not a new tool. A new starting point.

The people who grow fastest disrupt themselves before the market does it for them. And the need to is close to constant.

You can run the first move yourself. Write your strategy in one sentence. Then write the belief about yourself that has to die for that sentence to be true. Work the answer that costs you, not the first one that's convenient.

There's a short quiz that walks through the opening question. The link's in the first comment.

I've been an agile coach for a long time, and for years I was pretty dogmatic about it. Scrum, SAFe, whatever the framework was, I knew the rules and I wanted them run by the book. And so a team would bring in the ceremonies, the stand-up, the board, all of it, and I sort of figured that was the work. It wasn't.

You can run every ceremony perfectly and still change nothing about how the work actually gets decided. Same approvals, same gates. And then the results don't show up and everybody's like, well, agile doesn't work. And it's like, no, you didn't do agile, you did ceremonies.

The framework was never the point. The point is changing the system underneath, and building the capability in the team so it sticks. I saw this over and over. You'd get a team really humming, and the second you left, they'd snap right back, because, you know, the organization was never set up to hold the new way of working.

So watch what's happening right now. Everybody's flattening. Cut the middle managers, wire the whole thing straight to AI, call it lean. And it's the same mistake, one layer up, right?

You pulled out a layer, fine, but you still haven't touched how a decision actually gets made. It'll snap right back too. Cutting a layer isn't changing the system. It never was. That's really what line of sight is about, and it's what I'm getting into on the blog this Friday.

So I'm curious, what are you watching get cut this year that you already know won't fix the real problem?