The cover version
I’ve had some good experiences using the Modern Agile model for explaining agile thinking to people. For reasons laid out below I use a somewhat adjusted version of the model (words in italics have been altered from the original), like this:
- Make users successful
- Create a setting for great teamwork
- Deliver value continuously
- Experiment and learn rapidly
If you’re interested in why I feel that Modern Agile is important and why I have adjusted it for my own purposes, please read on.
Introduction to Modern Agile
If you’re familiar with the Modern Agile model you may wish to skip to the next section.
Modern Agile (or MA, for short here) is a thinking tool/model by Joshua Kerievsky of Industrial Logic. The model is a modern take on agile thinking that has gained a lot of interest over the last year.
Modern Agile consists of four principles or values, see image below.
The yellow principle, Make people awesome, informs us our job should about creating successful people through our products and services. Who cares if we delivered three new features last week if the customers didn’t find our features useful?
The blue principle, Make safety a prerequisite, points out that nobody performs well unless they feel safe, including psychological safety. And we’re not really safe at work, it turns out. Most employees are just accustomed to taking individual risks or being mistreated. This also includes technical practices that help us share and reduce risks.
The green principle, Deliver value continuously, says that we should create a continuous stream of value to our customers. We know today that stopping and starting work is wasteful. In software we include practices that make us reduce lead times for delivery; improve collaboration, build quality in, break down work into small pieces and how move them quickly and safely into production. Even more ambitious, we need to do this in a sustainable way.
The red principle, Experiment and learn rapidly, concerns our ability to conduct small, safe-to-fail experiments and learn from them. We win by “outlearning” our competition. This concerns both the product dimension (the outcome of the work) and the process dimension (how we work).
For more details, see the Modern Agile homepage.
Why I like it?
I think there are several advantages with Modern Agile compared to the original agile manifesto (or AM, for short here). The AM was, and still is, a fantastic summary of some important values and principles in software development. Although now a historic document, it’s still highly relevant in most corporate settings. For example, many people still believe documents are somehow more professional than “just talking”. Not to mention the importance of “having a good plan”.
But let’s face it, the group that created the AM was desperately low on diversity. Apart from the glaring fact that they were all white, middle-aged men, they were all software developers. There were no interaction designers, no UX specialists, no software operation experts in the Snowbird cabin. Hence, there was a distinct lack of knowledge around important parts of the software development value stream, which, of course, lead to their exclusion from the manifesto.
Just to name a few examples, the manifesto tells us to collaborate with customers, but where are the users? It tells us that the software should be “working”, but not usable, useful or valuable? It’s hard to avoid the conclusion that the model the signatories had in mind was collaboration with one or a few persons that knew all that was needed about what problem we were trying to solve, the users and their context, the intended use of the program, and relative priorities. Today, we know that model to be flawed. Understanding user and customer needs is a journey of discovery.
Even more obvious today, organisational functions outside of software development were not considered in the AM. This is no criticism, just an inevitable fact. I mean, how could they possibly have known that agile thinking would become so successful to spread to departments outside of IT/Development?
The Modern Agile model incorporates the missing parts of the value stream and many learnings we’ve made since the writing of the manifesto in 2001. Just think about what we didn’t have back then; no Kanban method, no Lean Software Development, no DevOps, no Lean Startup, no cloud, just to name a few. We know today that the most effective development model is to flow small pieces of value as often as possible to the users/customers and tighten all feedback loops. We know today (and we actually did already back in 2001) that we cannot simply listen to what users tell us – we need to research them to understand their needs and pains. We know now, thanks to Kathy Sierra, that our objective should not be to create a great product – but a product that makes users great. We know today we can automate, not only the building and deploying of software, but also the infrastructure, configuration, and data on which is depends. We know today that it’s actually possible to work in a way that incorporates improvement on a daily basis, not just reflecting for an hour every two weeks.
And as Agile thinking is spreading outside of the IT department, agile advocates need a common model to explain what we mean without each having to adjust ye ol’ manifesto individually – or leave it up to the listeners (not recommended). I believe Modern Agile is a nice effort to generalise the agile mindset. Having a model that can be used without adjustments when speaking to finance, HR, marketing, sales, executive teams or whatever else is wonderful.
But there are some challenges…
Just one guy
The original manifesto was created by a group of software and methodology experts (albeit a low-diversity one), many of whom were already well-known pioneers and thought leaders. To me, that spoke volumes at the time. I was already into Extreme Programming when the manifesto was published, but when all these really smart people, even heroes to me and my friends, all said the same thing that had to count for something. It showed me I was on to something that would not be ignored.
However, behind Modern Agile there is only one person and no matter how wise and experienced Joshua is, we cannot escape the fact that there’s no community agreement that Modern Agile is really what we mean by Agile in 2017.
“People” is too broad
I find the word “people” in Make people awesome to be too generic a word for my taste. This principle is deliberately inclusive to everyone in the game; customers, users, managers, coworkers, families, suppliers, etc, which is nice. MA asserts we should strive to make them all great. This fits well with Jurgen Appelo’s assertion along the same line, that we can’t focus on one or two stakeholder groups, to the detriment of the rest.
However, one could argue that by including everyone, you are also not excluding anyone. You’re not having an opinion. It’s like a product prioritization where “everything is important”. To me, that makes it weak and not very useful.
Also, in my book there are three groups with more skin in the game than the rest: The software users (using the product or service), the customers (the people paying for the product or service, often the same as the users) and the developers (everyone involved in creating the product or service). The aim of the Make safety a prerequisite principle is primarily developers so they’re already covered, in a way, but we need to specifically address the users and customers, because losing focus of your user/customers is deadly. Perhaps less important I find the word “awesome” a bit childish and exaggerated.
Currently I lean towards using “users” instead of “customers”. My thinking here is that, when users and customers are not the same, which is typical in an internal IT setting, if we provide software that fulfil users’ deep needs and make them succeed in their job that will, in turn, create better organisation and satisfied customers as a side effect. Using “customer” on the other hand could lead us to focus too much on the person paying for it and too little on the people having to use it. If the users are also the customers, it doesn’t matter, the words are interchangeable.
Summing up, I believe the original formulation of this principle was Make Users Awesome and I believe going back to something along those lines makes it stronger.
“Safety” is too narrow
On the other hand, I find the word “safety” in Make safety a prerequisite to be too narrow a word to well describe all the intricate, interlocking mechanisms and conditions needed to release people’s builtin capacity for great teamwork and joy at work. Yes, it is indeed very important, according to project Aristotle, perhaps even the most important, but there are other important factors around teams that I find hard to ignore.
There have been many attempts to capture important factors in models. Dan Pink’s book Drive popularised the Autonomy – Mastery – Purpose model. In Jurgen Appelo’s book Management 3.0 he identifies ten “intrinsic desires”; factors that contribute to motivation. Also, we have Patrick Lencioni’s popular book Five dysfunctions of a team, naming trust, constructive conflict, shared commitment, mutual accountability, and focus on results as vital for teams. To continue, one could add W. Edwards Deming’s and Russell Ackoff’s teachings around quality and systems thinking, pointing out the vital importance of setting up the system (the context) properly. Individuals or teams don’t stand a chance in a broken system.
Another aspect of safety is what one might call technical safety. By that I mean technical practices that help us share and reduce risks. Common examples of this are pairing, mobbing, version control, test driven development, continuous integration, deploy pipelines, etc. I perceive these to contribute a lot to safety, and consequently equally much to teamwork. Of course, they also contribute to the continuous delivery principle.
Moving on, also I prefer “teamwork” over “teams”. Among agile thinkers there’s a lot of focus on the team. To me, however, a team is a vehicle. What we really want is great teamwork. Chances of teamwork are probably higher in a team, but having a team is certainly no guarantee of fantastic teamwork. Also, we want to work well in the team of teams surrounding us. And of course, if we can establish great teamwork with our users and customers, that’s even better.
Summing up, in my opinion the wording around this principle should mirror that we know something about enabling people’s intrinsic motivation setting the stage for great teamwork, but that we have no definite list yet.
If you’ve read this far you know how I feel about Modern Agile and how I use it. I hope it was useful. I want to extend a big thank you to Joshua Kerievsky for his contribution to our field. Comments or suggestions are highly appreciated.