Agile in Large Complex Organizations | by Chris Combe (he / him) | Apr, 2022

Is there room for innovation in agile?

Picture by drmakete lab on Unsplash

The title might sound extra ominous than meant. I want to unpack a few of my learnings and ideas. In complicated programs, giant scaling frameworks create rigidity and can over-constrain a company.

It’s fairly simple to throw rocks at scaling frameworks and blame them for all types of dysfunctions, the fact is there’s a shared accountability of the folks adopting them both on their very own or by way of the help of a consultancy.

Paying your strategy to enterprise agility is sort of tempting for giant complicated organizations in search of a packaged resolution.

It’s a little bit like shopping for a home totally furnished considering you’ll get every little thing you want. Then you definately notice it was constructed for a household of a smaller dimension and also you are actually pressured to share the identical bed room, absolutely the 2 pool tables the home has are extra essential than your loved ones, having beds.


For the few of you not acquainted with Geoffrey Moore’s guide ‘Crossing the Chasm’ and the well-known diffusion of the innovation curve, you’ve seen it earlier than even when you don’t know the title/origin. It properly plots the unfold of innovation time beyond regulation on a curve and the way every market has differing must be satisfied.

It holds true for any innovation and one of many ignored elements, is the inertia required to cross from the early adopter stage to the early majority — that is typically the place prospects and markets have totally different wants and take distinct kinds of proof to get their enterprise.

Very often rising and novel practices fairly make it over the chasm and stay restricted to fewer folks.

Chris Matts talks about this in his speak ‘Agile — the damaged studying machine’ e.g. quoting Function Injection as one such instance.

The place would you say agile actually is on this curve? I’ve heard many practitioners declare that agile is now on the level the place it’s someplace between the late majority or the laggards (relying in your perspective).

In case you took an identical graph and plotted agile on a Wardley Map, you’d see comparable progress, though Simon differentiates Scrum and XP. The purpose is that they’ve reached their novelty and are now not differentiating, as an alternative, they’re taken with no consideration.

I don’t totally agree that on Simon’s depiction of the place agile works greatest, when it comes to the X-axis, I’ll save that for one more time. It’s honest to say that the place you’ve much less certainty in your work and also you need to scale back your dangers, it’s best to use shorter suggestions loops to allow quicker studying.

Supply: Simon Wardley —

There are quite a few movies and talks with the identical title, I do not suppose agile is lifeless, however I feel agile has turn into one thing far totally different than was initially positioned some 20 years in the past within the Agile Manifesto. It was clearly meant for software program growth and constructing merchandise.

Now it’s used to clarify a set of extra widespread practices that any group/crew can think about using of their methods of working. I don’t have any considerations with adapting strategies to your context and getting worth out of them. What does are inclined to create concern, is when agile will get turned it a business / packaged providing, which is concentrated on giving everybody a task with a top-down command and management construction that appears nothing like agile. Except the system is there to allow and act in an incremental method, then everybody on the high will nonetheless expect massive bang supply.

I’m a fan of the method set out from Klaus Leopold round Flight Levels. It’s an attention-grabbing method of specializing in technique, coordination, and supply of the work such that everybody is considering limiting the work-in-progress and never simply the groups delivering. I like to recommend his speak on how agile groups don’t create enterprise agility on their very own.

When agile is applied in giant organizations as of late, it tends to be accomplished so in a push method vs a pull method. Dan Mezick has talked about this extensively and has an incredible sequence of books and an method he gives referred to as Openspace Agility which I extremely advocate trying into.

The thought is to have folks flip up on their very own volition and volunteer to be concerned within the change quite than top-down chief choice. This has opposed results if folks have agile thrust upon them. Persons are much more more likely to undertake one thing new when they’re collaborating and desirous to contribute.

There are numerous examples, books, and talks in regards to the reversion of behaviors after an Agile Transformation the place folks went by way of a sheep dipping train, they get assigned roles and a few weeks coaching after which pushed into the deep finish. Then 3 months later they’re again to their earlier methods of working.

Look out for the place management holds the ‘agile coach’ liable for different folks’s dedication and habits. They could be gifted, however it’s a complete crew sport and can’t individually be liable for the success of the crew.

Sure, there may be nonetheless room for brand new patterns to develop, particularly when it isn’t at all times clear tips on how to deal with complicated dependency chains that span a company. That is more likely to be a standard sufficient set of issues that different organizations can share their practices and method too in order that the broader neighborhood nonetheless has one thing to study. Not each firm is a tech firm, in truth, most of them are usually not. Most corporations begin with a messy present state which incorporates many dependencies and important quantities of technical and organizational debt.

There’s a number of room to scale back the necessity for giant complicated packages/initiatives over time. The place as an alternative organizations and supply constructions are arrange to have the ability to ship incrementally and extra independently over time in order that issues like function flags can be utilized, in addition to creating extra belief / confidence in testing such {that a} full entrance to backtesting effort isn’t required both as typically or in any respect.

Meaning prioritizing efforts to deal with technical debt, not just for obsolescent know-how however for architectural patterns that allow higher isolation and modularity in relation to testing. Many of those ideas and practices are usually not new, the problem is that those that know the practices are usually not prevalent within the Monetary Companies business.

Apart from the technical and architectural challenges, there may be extra to discover with out resorting to giant scaling frameworks when complicated coordination is required. I consider that we don’t have to resort to ‘the stick’ to get issues accomplished. You don’t must ship ‘on time’ when you ship early and sometimes. Creating these types of capabilities is way extra resilient than ensuring folks hit the supply date, particularly if meaning paying down technical debt as you go.

An attention-grabbing sample to discover is hexagonal architecture (popularised by Alistair Cockburn) which helps you to layer your code/structure in a method that abstracts ports and adaptors in order that issues could be accomplished independently between the within and the surface of your system however with free coupling!

So quite than speaking that Agile will ‘remedy all issues’, I desire to phrase it as agile helps spotlight the issues and permits for the experimentation of practices to discover! Experiment, study, and share!

More Posts