What you should know about change even in post-agile times

Post-agility refers to the state in which agility as a term has lost its clear core meaning because it has become too widespread and therefore too vague. Change in organizations is more successful when several models are combined: Kotter’s eight-step model for major changes, Satir’s Change Curve for individual learning processes and Tribal Leadership for group maturity.

Key Takeaways

• Launching several change initiatives at the same time is counterproductive: if you constantly impose new changes on teams before the previous ones have been integrated, you create frustration instead of progress.
• In practice, the Kotter model usually fails in the first two steps: no real sense of urgency and no coalition of leaders mean that all further steps come to nothing.
• The Satir Change Curve does not describe a crash, but a J-shaped learning curve: Those who understand that the chaos phase is normal learning can consciously decide when the next change is appropriate.
• The Tribal Leadership Model distinguishes five stages of maturity from “Life sucks” to “Life’s great” and gives managers concrete starting points for developing groups one stage further.
• No single change model is sufficient: If you combine Kotter, Satir and Tribal Leadership as a kit, you can use the right tool for each specific change.

What post-agility really means

Post-agility describes the state in which a topic has arrived in widespread practice and has lost its sharpness precisely because of this. The term works in a similar way to postmodernism: not a clear successor, but a signal that people are no longer talking about what the term originally meant.

Michael Mahlberg puts the development into historical perspective. What is now known as agility originated at a conference for lightweight processes. The result was called “Manifesto for Agile software development”, not Manifesto for agile management and not Manifesto for agile organizational structures.

This is precisely the problem for change projects. Agile has become an umbrella term that stands for almost everything. Throwing a term into the room that can mean anything does not give a change project any useful direction.

This leads to a practical point: there have always been separate models for organization, leadership and structure, and they continue to exist. Instead of imposing agility on everything, it is worth looking at existing change models, each of which is intended for a specific level.

Why major changes fail due to a lack of suffering

Major change usually fails at the very first step: there is no real sense of urgency. It is no coincidence that John Kotter’s eight-step model begins with this point. Without a tangible pull to change something, the following seven steps will not work.

The typical pattern is recognizable. A single person thinks something needs to change, but is unable to convince anyone. Anyone who then pushes the change through by force is working against a system without traction.

The second step is often misnamed. It is not about an “alliance of the willing”, but about a coalition of leaders who want to shape change from within. This small internal alliance is the basis for everything else.

The third step, the vision, means something other than a new corporate vision. The question is: How can we recognize that life will be different after the change? Not blue forms instead of yellow ones, but: How do the interactions feel, and how can even an outsider tell that something has changed?

It’s not about, we now need a new vision of what the company wants to do in the world, but really: When we are through with this change, how do we recognize that life is different? Michael Mahlberg

Just reading through these steps reveals gaps. In many initiatives, steps two and three were completely missing. In hindsight, this explains why many things didn’t work.

Sustaining change means two things at the same time

For Kotter, perpetuating change means two directions, not one. On the one hand, it is about securing what has been achieved. On the other hand, it is about keeping changeability itself permanently available.

Securing what has been achieved is more challenging than it sounds. Systems tend to fall back into old ways, especially if the informal old structures are still functioning. As soon as the old ways of thinking and acting are still within reach, behavior easily reverts.

The second direction is aimed at the willingness to try out more and change more. You have to differentiate between the two. Some things you consciously want to consolidate, others you consciously want to keep flexible.

The change curve according to Satir: change is not automatically negative

Virginia Satir’s Change Curve describes how a person goes through change, and it makes it clear that change is not a bad thing per se. The model originates from family therapy and came to IT via Jerry Weinberg as a bridge between soft and hard perspectives.

The process follows a J-shape. At the beginning there is an old status quo. Then an external influence arrives and little happens at first because people have an inertia.

This is followed by a phase called chaos in this model. The name sounds worse than it is. It simply means that you are no longer in your usual state, you no longer do things as a matter of course, you have to think, ask questions, you need time in between.

At some point, a stabilizing element comes along. You see, hear or feel something that helps you to integrate what you previously experienced as chaos, and you settle into a new state. How long this takes is different for everyone.

It is important to distinguish it from Kübler-Ross’ crisis model with Denial, Anger, Bargaining. Both curves look similar, but mean different things. Learning something new can be motivating, especially if you do less of something else for a while.

Why too many simultaneous changes are frustrating

The dose makes the poison. A single change can be fun, a second one too. It becomes problematic when the next change comes while you are still at the bottom of the J-curve, and then the one after that.

If you constantly have to relearn without ever reaching the top, you become frustrated. Michael gives a concrete example from the testing environment: introducing Gherkin today, a Test Automation Server tomorrow, a new Test Runner the day after tomorrow. That’s not wise as long as the first is not integrated.

Task loading from technical diving serves as an image. No single task is a problem: look at the air, make a deco stop, switch the mixture. But if you do 15 things at the same time, you will forget something at some point. It’s the same when someone has to learn 15 things at the same time.

The consequence is the central practical lesson from the Satir model: it is better to take many small turns instead of one big one. Instead of starting a huge change, it is worth asking which small step is going in the right direction and is already bearing fruit.

One change per experiment: the monocausal principle

If you only change one thing at a time, you can clearly attribute the feedback. If you change five things at the same time, you won’t know afterwards what brought the added value.

The same principle applies to testing. You test one thing and see whether it produces an error. If you test everything at once, errors occur whose cause remains unclear.

The appropriate term comes from scientific work: the monocausal experiment. It may sound a bit stilted, but it means something simple: a learning loop with only one single reason for change.

Tribal leadership: what level is the group at?

Before choosing an intervention, it is worth taking a look at the maturity level of the group. The Tribal Leadership model describes how a group looks at the world in five distinctive stages.

The “We suck” level does not appear in the model. “I suck” is directly followed by “I’m great”, i.e. heroism. This attitude characterizes some very successful companies and, according to the model, affects around a third of jobs in larger companies.

The starting points are particularly clear at the lower levels. If a team has a lot of people who think they are great, but the others don’t, as a manager you can point to what they have in common: There’s someone who’s also good sitting right next to you, look what you’ve achieved together. The path from “I’m great” to “We are great” is itself a change.

How the models work together as a kit

No single model solves the problem, the models interlock. If you want to initiate change, you should have several of them in mind at the same time and choose the right one depending on the situation.

The starting point depends on the trigger. If it is a specific pain, it is worth doing a tribal assessment first: what stage is the team at, what intervention options does it have itself? This can often be done without external help, as a team member or as a manager.

The next question is about your own proximity. As a manager, can I initiate this step myself, or am I too far away from the team? Anyone sitting three teams away should give Kotter a thought: Can I even create a Sense of Urgency with someone who is close enough? If not, the basis is missing.

The Satir model runs in parallel. If a team is currently working on collaboration, you should not demand a new test tool and exploratory testing on top of that. The group is already on a change curve.

The decisive factor when transferring this into practice is that even a harmless sentence such as “something has to change” contains many triggers. Whose influence, why, with what vision for the state afterwards? There is one model for each of these triggers. You put them together like Lego.

The most honest tip: You are never in just one change

The most important reflection is: How many changes am I in at the same time? The idea that there is only one change is an illusion.

Change is often not centrally organized, it just happens. One example: In winter, pair programming on a shared screen works perfectly. In summer, the sun is on the monitors and productivity drops without anyone having planned for it.

As soon as one curve is completed after three weeks and the team is more productive, the next unplanned change occurs. This places a clear requirement on management’s powers of observation. If you only look at the figures in your own office, you won’t see these superimposed changes because there are rarely any KPIs.