Agile is post industrial. Why ? (and why should you care …)
I am an Agilist (also). A key success factor to implement agile is to understand this is a post industrial concept ( versus industrial). Especially , especially its relation to autonomy in work organisation.
Or else you will most probably implement agile governance … the industrial way … out of habit…
And this is not what you really want, in my opinion…
The Challenge is triple here: 1) try and understand what Post-industrial really means, 2) what it may signify in an IT Agile context, and 3) what it would imply as you redefine your work organisation to go agile.
1 According to Wikipedia, Post-Industrial is when most of the wealth does not come from industry , but from services. Currently, there is roughly 70% of employees in Europe or USA working in services, so we can safely consider we are in a post-industrial era.
2 If we would summarize the core distinctions developped herefater , we could say:
Industrial was central thinking and deciding, using Rules, predefined right or wrong, fostered conformant behavior and provided for the basic material needs thanks to mass production.
Post Industrial is local thinking and deciding, providing coherence via principles that are adjusted to local reality; this requires autonomy, and responsibility, including sharing back the emerging adjustment to these principles; to keep the coherence evolving with the fast changing reality. It provides for higher needs that are less material, more informational in nature and more diverse in variety, possibly unique and ephemeral.
Let’s details this. Strictly, we will get out of the box of IT Agile, and wander in Sociology, Psychology, Economy and Enterprise Architecture. Agile readers are usually curious, so keep on the effort, we try and stay light (and provide pointers) . To make the point clearer, let’s forget nuances and subtleties for a while, push up the contrast and imagine two poles:
- for industrial : an early 20th century factory on one side,
- for post industrial: an Agile IT company on the other, say a mid size 500 people, running a portfolio of products. (to go beyond the archetypal one team, one product startup. )
We will elaborate a sharply contrasted picture to enhance distinctions between industrial and post-industrial, within growing logical levels.
Who Think
“Some think, some do “ could summarize the industrial mindset. It means «Some thinks, the other executes » . Owners think the “vision”, all the rest execute the plan and served mainly as ‘muscles’.
(Note that this is still the dominant vision — execution paradigm of most consulting companies.)
How can we contrast this in the post-industrial paradigm ? Lets try “Everyone think, most execute” . In IT, not only the “doing” is “thinking” but “organizing” is thinking too, like tooling, planning, visioning the “project” or the “product”. Working is no more muscle, but thoughts of action.
Distribution of decision
However, Autonomy is an important factor in agile team. And this old “think and do” paradigm does not capture the essence of autonomy : decision. Lets use the “think -> decide -> do” paradigme instead.
The industrial side summary becomes then “Some think , Some decide , Most do”.
Thinkers are the owners and C Level boards, so are deciders plus the engineers implementing. On the other hand, the assembly chain, and most of the workers just operate the machines (do). There is not much to decide for a worker (in a pre-Lean factory). They are required to “Please, just fit in and be conformant”.
The post industrial paradigm for semi-autonomous team would become :
- Everyone thinks and decide.
- Some decide also on « strategies »
- Most do.
Hence the trick is to find the balance between
- « everyone decide for everything » and
- « no one decides, really » and
- « a few decide for everyone ».
The first one does not scale beyond a handful of people, the second leads to chaos and possibly disintegration, and the last one revert to « central planning », aka the industrial mindset. Autonomy works when the traps are clearly seen, and avoided.
One of the factor to distinguish between a set of independent startups (i.e. Full independence), and a set of reasonably co-operating teams, (I.e semi-autonomous), is the local-central distinctions. Let’s try a formulation:
Everyone thinks and decides.
- All decide locally , for them, between them.
- And some decide (for the common good), in central , on principles.
Most do, locally,
- Adhering to these principles when relevant and applicable,
- And (possibly) emerging new principles when not.
Central and Local
To illustrate what is meant by local , The local decisions are the one a team takes when deciding on the project they build.
And hopefully they reuse an existing architecture, and an existing infrastructure, so they do not reinvent (fully) the wheel. This can be a framework, or an architecture; it can be just some «commodity » open source stack, or something more sophisticated like a scalable, fault-tolerant, self healing GAFA architecture that another project designed previously. Something you cannot « buy » on the market, yet , ready to use. This is something « central » to grow, some « ideas » turned into code, that become part of the « common good » of the company. These « common goods » are managed by a « club »[^2] of interested users , like an architect team and a « community » . Note that central does not mean any more « higher in the pyramid » and the community does not « report » to the club, as the industrial mindset did.
Rules versus Principles
Another important distinctions is Rules versus Principles. The industrial mindset was full of Rules. Do this, don’t do that …
A rule tells “How it is ”and how « it should be », the way How things are « done here ». There is ( ideally) no exceptions and variations, to make it simpler.
The rule is prescriptive [^3]. « Everything that is not authorized is forbidden ». This is THE way of the Business Process. It stems from the fact that an assembly chain is meant to be operated ONE way, and any disturbance or unexpected variation causes disruption upward and downward the chain. Exception is the enemy. Conformance is key. People are confounded with Roles and should be expendable.
Prediction and Predicting strategies are keys in an industrial mindset. But Predicting Strategies are based on assumptions, and this requires some stability in the environment. Predicting Strategies are useless and worthless in a fast changing world, and then Adaptation becomes key. Agile is Adaptation as a Strategy.
In the Post Industrial Mindset , to cope with Adaptation, the principle cannot be prescriptive, [^4], so « Everything that is not forbidden is authorized.»
So the principles states an intention (and possibly interdictions), and the implementation of this principles will be « local », to take care of the local context.
Moral, Freedom and Responsibility
The Industrial way was prescriptive, like Moral that tells you how to live your life “well” , by distinguishing « good » from « bad »; that leads to some uniformity, there is some conformance to a social « process »
But when there are many options possible to go , the binary right or wrong scale is no longer good enough. The opposite of wrong is not necessarily right. The appropriate right, for you, now. Choosing which option is better requires experimentation. Premasticated choices from the past may no longer be valid because. things may have changed since then.
This situation has been called “ambiguous” or “complex”, as opposed to “complicated”, for the industrial age. Building a factory is complicated, raising a child is complex.
However, The fact you have more options means you have increased. freedom. Which is nice. Unless you are uncomfortable with the fact to look for answers by yourself. Post Industrial is freer but less certain. You may need to increase your ability to find response by yourself. The Response-ability. Post Industrial brings more freedom with more responsibility.
Mass Production Versus Long Tail
Industrial was aimed at mass production. That meant producing a lot of copy of the original prototype, with no or few variations.
This is illustrated by the gauss curve, where 80 % of the coverage is within 20 % variation of the median. The other part of the curve, the tail was just ignored. This was the way to make profitable business.
Until eBay came. If you consider eBay as a store, it has the largest number of reference you could ever found. And this is possible because. it has none in stock. To compare with Walmart or Carrefour, these have a physical stock with much less references. This was the way mass distribution worked. When all eBay does is connecting people.
This is making some sense because when you buy fresh vegetable or water at Walmart, you expect to have them not too far from you. Because you cannot live 3 days without drinking. But you can wait 3 weeks for you electronic gizmo to be sent from china via eBay.
Instead of focusing on the middle of the curve, this focus on the other part, the tail. This long tail business model was conceptualised in 2004 and later it was demonstrated that there is more business potential in this thin long tail than in the conformant middle curve. Diversity over Conformance. Connection over Production.
Material versus Informational
If you consider the Digital Age, Production (in the sense of replication) is no longer the hard stuff. Nor distribution. Music CD had to be fabricated and distributed for me to have music. My son brought me Spotify, and now I have all this , plus it gave me the opportunity to discover new music from my son, my friends and connections, and it even create a “list of the week”, unique for me, with suggestions I never heard before, based on my preferences. Unique and ephemeral object. Is it even still an “object” in fact ? A “Product” ? Something I cannot touch and that does not lasts… hmmm.. “Product-ion” is no longer what matter in the post-industrial age.
Needs
Why do we work for, by the way ? To fulfill our needs… If you Consider the pyramid of needs, Maslow-like, you could see that industrial age permitted to develop survival and security needs for a larger part of humanity. Food, shelter, material confort… Now look at the next layer, relational and autonomy. Are these needs still material ? Or rather more informational in nature ?
Take the cd versus spotify metaphor. Both brought me “music” (entertainment, art), but spotify is bringing me autonomy and connections (with my relations) on top. And with the “ weekly list”, it explores for me.
Could it be that the post industrial age is aiming at higher needs, less material, more informational ?
So does it makes sense to work in an post industrial organisation that is mainly informational , provides for higher needs of its ecosystem, tailor its “products” to the point of unicity, and at the same time, is organised with the principles of an industrial factory 19th style ? Possibly not… But how would I know the differences if I cannot differentiate industrial from post industrial ?
So Let’s Summarize (again) :
Industrial was central thinking and deciding, using Rules, predefined right or wrong, fostered conformant behavior and provided for the basic material needs thanks to mass production.
Post Industrial is local thinking and deciding, providing coherence via principles that are adjusted to local reality; this requires autonomy, and responsibility, including sharing back the emerging adjustment to these principles, to keep the coherence evolving with the fast changing reality. It provides for higher needs that are less material, more informational in nature and more diverse in variety, possibly unique and ephemeral.
3 There is a lot of change to cope with in this stack of models , and a lot of new concepts to digest … A lot of chance to miss one also …
At the same time, it makes sense for an enterprise to be aligned to the needs of its clients, even if these are changing deeply in nature, as the society changes.
But ultimately, we could now understand that Agile is not requiring autonomy as some ideologic revolution against centralism, but as a necessity to adapt locally, because of fast changing needs.
It stemmed from a change in the level of needs fulfilled , the less material nature of the needs, and its variety . The nature of the product changes too, and de-materialises. It ensues a change in business model , from Mass Production to Long Tail , that itself command a change in the Operating model, from Central Planning to Semi-autonomous entities.
The risk is that with so many changes, and with a lot of them coming gradually, they stay implicit, not said , not seen. Most organisations are just structured “normally”, i.e. around “classical concepts”. Or to put it explicitely : the Industrial way. And one of the reason could be: there was no other options, no distinctions, explicitly.
And now, you have.
So, to conclude this first part with humor
The next time you hear your clients uttering the motto to « let’s industrialise agile » [5]:, you may realize there is some confusion at play there, a lack of distinctions :
Agile is Post Industrial.
And possibly you may have found a few arguments more. I would be delighted to hear yours of course.
— -
Next Post will details why it matters to understand these distinctions, and how to turn this understanding in practice. Let me know if you are interested to read this .
— —
— —
[^1]: This is still the dominant vision – execution paradigm of most consulting companies
[^2]: https://en.wikipedia.org/wiki/Public_good#Definition_matrix
[^3]: interestingly , Merrian webster show that synonyms of prescriptive are « conventional, traditional, customary »
[^4]: as the future is no more « traditional » , the past cannot help predict the future, as « things changes ( fast) »
[5]:this was the official motto I heard in 5 of the banks I worked in since 2008, when trying to “scale agile”.
[^ 0]:
Your beliefs become your thoughts,
Your thoughts become your words,
Your words become your actions,
Your actions become your habits,
Your habits become your values,
Your values become your destiny.― Mahatma Gandhi
– –
Thanks for they kind review and comments. : Sinan Si Ahir, Tom Graves, Dov Tsal Sela , Damien Thouvenin, Marion Fontaine, Jean Pierre Schmitt, Stephane Badach.