How organisations are changing.
Just over a decade ago, I published the table in figure 1 to describe how organisations were shifting from a traditional form to a set of next generation behaviours. Whilst the table was welcome in some quarters, it was generally met with derision and comments of “this is just for startups”. Not that I’m bitter … much … gronda gronda.
Figure 1 — Traditional vs Next Generation Company behaviours
I’m glad to say that a decade later, that many have already started their journey to the next generation or at least acknowledge it as their future. So, given that I’m a glutton for punishment, I thought I’d repeat the whole process again and see if we can’t find a new “next generation” or in other words, a next “next generation”. Yes, evolution stands still for no-one and whilst cell based organisational structures, open source as a weapon, chaos engines, continuous deployment and learning from ecosystems sounds cutting edge … it was … a decade ago.
I say find because the future is already here, it’s just not evenly distributed and hence this is an exercise in finding existing differences in corporate populations and not pontificating cause and effects through technology change to “discover” a future. I’m not a fan of crystal balls and the endless line of futurist gurus willing to read your palms if you cross theirs with enough silver. I happen to like things that exist, even if it is only in prototype form.
Starting with a pattern.
Before we can begin our hunt for the next generation (if it exists … spoiler alert, it does), we need to understand a basic pattern from mapping known as co-evolution. This pattern has served me well for the last fifteen years from anticipating the rise of DevOps to gameplay with Canonical. I’ve outlined the basics of this pattern in figure 2.
Figure 2— Co-evolution of Practice with Activities
To summarise, the evolution of a thing across a state (i.e. custom built to product or product to utility) changes the characteristics of the thing which in turn leads to a new set of novel practices that will in turn evolve, becoming first emerging then good and finally best practice for that more evolved thing. The thing that evolves brings us efficiency whilst the practice tends to bring us speed. This is best explained with an example.
When compute evolved from products (called “servers”) to a utility (called “cloud”) in 2006, its characteristics changed. It went from high MTTR (mean time to recovery) where it would take weeks to get a new server to low MTTR where it would take seconds. This led to a new set of novel practices such as design for failure, chaos engines, distributed systems and continuous deployment under a banner of “infrastructure as code”. Eventually this was given the flag of DevOps — see figure 3.
Figure 3— The rise of DevOps
There are a few comments to be made on figure 3 above : -
- The evolving technology is the cause of the change.
- The practices are different competencies but they do share a common meaning ie. DevOps and ITIL were both architectural practices, just one is more “Next Gen”.
- We have inertia to this change caused by past success and pre-existing capital including pre-existing practices. Being good at the past turns out to make adoption of the future that bit more difficult.
- There can be forcing functions to change i.e. today, that is the isolation economy caused by covid.
- This change will also allow for new needs to be met and and hence new opportunities for value to be created.
I’ve summarised this all in figure 4
Figure 4— Co-evolution impacts
I italicised the term “Next Gen” in the text above because there is no reason why practices can’t evolve themselves. ITIL could have adapted to the new world but alas we have that tricky issue of inertia and factional mechanics to consider. Any new faction (a collective) always desires to be different from other collectives. They will often paint the previous practices as old fashioned and create their own flag, symbols, stories and rituals. This is exactly what DevOps did and it’s normal behaviour. NB to all us mappers out there, when someone finds a better way of mapping then we’ve got to be acutely aware of our own inertia and try to evolve before we also get painted as “old fashioned”.
Back in 2010, when I started the research which resulted in the above traditional vs next generation table, the cause was the shift of compute from product to utility. It was fairly easy to target this space. Today, we have many areas of technology that are industrialising and even some that had already industrialised but we have just resisted adopting due to existing inertia. The isolation economy caused by covid isn’t so much creating the new but forcing us to overcome our inertia and adopt what already exists — zoom, telemedicine, virtual environments were all there, pre-covid.
Given this pattern of co-evolution, our first problem in the hunt for the next next generation is to find the target spaces they are operating in i.e. we don’t have that obvious technology change like “cloud” but many technology changes to contend with. In this maelstrom we are going to have to find those changes of practices, those different competencies. Given these competencies have the same meaning, we need to start by finding the changes in meaning first.
Finding changes in meaning.
To help find the target spaces that the next generation were operating in, I needed to create a human sensor network. In May 2020, I put together a team of 70 volunteers, organised into 11 different groups covering subjects such as defense to healthcare to robotics to immersion. Using regular zoom calls and maps, we discussed the changes that we were seeing in the industry. Whilst the teams discussed the changes, I captured any “changes in meaning” that seemed to be highlighted across the groups. For example, if multiple groups were discussing how learning was changing then I would record that.
The process is almost like “Derrida in reverse”. In deconstruction theory, Jacques Derrida created an approach to understand the relationship between text and meaning. The central idea is that the meaning of the words are tied to the context. There is no “truth” to the words, simply a reflection of our context. As the context changes, so does our meaning.
In our approach therefore, what we sought to identify recognisable changes of meaning and work backwards to find the change of context. In total, 43 changes of meanings were identified. I then tested these against various other groups to see if they agreed or disagreed. Some of those changes were more strongly identified as sources of ongoing change than others — see figure 5.
Figure 5— Example list of change of meaning and strength of signal of such change occurring.
Once we had stabilised on a common list of meanings, I then asked each of the research groups to provide a list of prototype competencies for each meaning and characterise them as traditional vs next generation. For example, leadership in traditional was associated with powerful and charismatic individuals whereas leadership in next generation was seen as more distributed.
To avoid us going into the world of science fiction, I also asked for examples of next generation practice, even if it was only an experiment being done or a paper written on the subject. For many of the meanings we ended up with multiple different descriptions of the change —not surprising given eleven research groups. Overall, we ended up with a list of over 600 traditional vs next generation practice prototypes.
Narrowing the space.
Talk about a maelstrom of noise, I was in the middle of it. The research at this point had identified 43 changes of meaning and over 600 prototypes of practice change. We had to reduce this somehow.
To narrow down the target space, I asked each of the groups to map out the area that was changing to attempt identify the underlying technology that was causing it. In other words, we were moving from understanding a change of meaning to examining the context itself. An example of such a map is given in figure 6.
Figure 6— Examining the underlying technology causing a challenge of orthodoxy in the medical profession.
In the map above, the key technology causes were identified as sensors, collaboration tools and access to data. By amalgamating these maps, a list of 19 common technology causes were identified. However, there are two issues to be mindful of here — timing and impact.
On timing
Throughout history, the most significant cause of change is the industrialisation of technology (from product to more commodity / utility). We had our candidate 19 technology areas but the question was — “Is this technology industrialising now or is it likely to do so over the next decade?”
To get a sense of this, I ran a number of tests online with various groups to see when the technology area was considered likely to industrialise — see figure 7.
Figure 7— Example time frame test for industrialisation of technology with 101 people.
Hence from the test above, we can see the group considers computing, network, GPS, sensors and even satellite imagery to either be industrialising or already industrialised. However, augmented and virtual reality are seen to be further away. Not by much but by enough.
On impact.
We know that certain meanings (i.e. leadership) are changing, we can identify the change of practice, we have a list of technology that is causing this, the maps to connect it all and even timing on when this is occurring but how much impact does a technology change have on the practice?
In the case of DevOps then cloud was fundamental. To test this with our 43 meanings, and 19 changes of technology, I created a number of matrix voting structures and asked people to vote on the impact of technology on meaning.
From both the impact matrix (and the clustering of 1,285 votes) and the timing tests, an overall picture of change was created in figure 8.
Figure 8— Changes in Meaning over time and technology causes
What the above figure is telling us, is that if there is a significant difference in companies then it’s likely to be found in the change of meanings highlighted in bold i.e. swarming of people, learning, principles and intent, sustainability, incentives etc.
The key “selecting” questions
Figure 8 is certainly interesting and gives us our target space — an intersection of technology, changes of meaning and time — in which we can go hunting for the next generation of company. Within this target space we already have our examples of traditional vs next generation practice developed by the research group. But how do we test this?
Our hypothesis was that the future is not evenly distributed and hence there should be two distinct populations within corporations i.e. two extremes of traditional and next generation in the wild. The majority of companies would probably be evolving between the extremes. But again, how do we test this?
Well, we could just ask? Alas, no. We can’t simply collect data by asking companies about those traditional vs next generation practices and then go find some correlation because we can always find correlations if we look hard enough in any data set. We needed to first find key identifying questions for the different populations e.g. if you have a population of 1,000 animals, a key set of identifying questions might be … “does it bark?” or “does it go moo?” … after which we can test those populations to see if they are actually different.
During this research process I had discussed these change with numerous companies, the research groups and also run online interviews. Two distinct characteristics had emerged during this time and been identified in our target space. The first was the distinction between being the company being driven by procedures versus the use of guiding principles to run a company. The second was a tendency towards believing the future of work was more remote first or office first. After testing this with several interviews, the following hypothesis was formed.
Companies that were strongly procedure driven and show a bias towards a future of office work would have significantly different characteristics to companies that were driven by guiding principles and show a bias towards a future of remote work.
This I can test. We were now ready to run the survey.
The survey.
We now had the hypothesis including the key questions, the target space, the traditional versus next generation questions and the underlying pattern of co-evolution that explained it. We now simply had to test to see whether the hypothesis held true or not?
Survey questions
The survey conisted of :-
a) Two selecting questions — procedure vs principle, remote vs office — which divided responders into three categories — the traditional, the next generation and the “inbetween”.
b) Twenty questions that included a choice between “Next generation” and “Traditional” characteristics as defined by the research group. These were to be tested to see if phenotypic differences existed between the populations i.e. one population doesn’t look like the other.
c) One directional question based upon views of the future.
d) One re-affirming questions to check for anomalies.
e) Two general questions on company size and industry.
Survey bias
The survey would not be a random population but those connected to my twitter steams — the channel through which the survey would be pushed. Whilst this wouldn’t effect the phenotypic characteristics of two different populations (if they existed), it would certainly effect the volume of each of those populations responding. It was assumed that I might get more people responding from “next generation” style companies over “traditional” than typically found in a random selection. There wasn’t much behind that assumption, it could have just been general arrogance on my part. However, it’s enough to say that I don’t think the population as a whole would be representative of the wider population. Hence, the overall survey cannot be taken as representative of wider populations but instead a biased population.
In other words, whilst the survey could be used to identify if different populations exist, the average results from the entire survey cannot be assumed to be representative of the entire population. Based upon the questions, we guestimated that out of 1,000 responses we might hit 50 companies from next generation, 20 companies from traditional and the vast majority of respondents would fall into the “in between” category.
We now had the survey tests to be run
a) Does the survey show two distinct populations?
b) Do the “in between” companies exist between the average results of the two extremes?
c) Can a “direction” of evolution be shown?
The result
The survey was conducted. 1,000 responses were collected and separated into populations based upon the selecting questions of “procedure vs principle” and “remote vs office”. There were 31 in traditional, 92 in next generation and 867 in the wider “in-between” population. The two populations of next generation and traditional showed significant differences across all the characteristics, for example :-
Figure 9 —Leaderless Leadership vs “Heroic” Leaders
Figure 10 — Top down vs Flash Mob
Figure 11— Outcome vs Output driven
The significance is enough that we can state that these are distinct populations.
The characteristics of the next generation population matched those next generation practices defined by the research group. The averages of the wider “in-between” population fell between the two extremes of next generation and traditional for all the characteristics. We can therefore reasonably state that there are two distinct populations and the in-between companies are evolving from one state to another.
The problem is direction. Are they evolving from next generation to traditional or from traditional to next generation? Whilst the research team had defined these characteristics based upon the view that one set was more future focused, we needed something more to support this. Figure 12 shows the different populations view of the future. The traditional view their future as in decline and the next generation view their future as growth.
Figure 12 — Views of the Future
Given the characteristics aligned with the research groups definition of traditional vs next generation, the population differences are significant, the wider population falls between these two extremes and that one group has a positive view (the other sees decline) then we can reasonably say that the overall population of companies are evolving to become more next generation like.
I have summarised all these characteristics in figure 13 which represents the behaviours of this next “Next” generation versus the behaviours of the traditional.
Figure 13— the Next “Next” Generation.
In narrative form.
I’m a huge proponent of using maps and this research project was conducted using them for communication, learning and challenge. However, I am aware that many are not familiar with the mechanics of mapping and hence in this case it is often useful to revert to story form.
In story form :-
Traditional Company
The traditional company is currently seeking a return to the office. It may talk of hybrid models of working but it has a bias in one direction. It is a procedurally driven beast with executives that consider themselves in the role of heroic leaders (even if they don’t openly say so). Symbols of the powerful matter, the top floor office, the hierarchy, the stories of heroic leaders and top down direction. Principles are an idea that are rarely stated or enforced. What motivates people in this environment is money. Sustainability is a cost to operations but it has to be done for marketing reasons. Marketing research itself is used to justify executive decisions not to question them. The focus of the company is always on the output, it might talk about community but it’s all about the product or whatever project it is doing. External comms is driven by mass influence i.e. getting other to buy the product. Ethics are an add-on. Awareness of the market is considered a function of leadership and the company views that supply chains are a way of shifting responsibility onto others. As a consequence, it poorly understands its own supply chains. To train its people, the company used expert tuition preferring face to face office lectures. In terms of future technology, the company views that AI will replace some jobs and some functions currently undertaken by humans. It also considers the future of the company to be currently one of decline with difficult times ahead.
Next Generation
The next generation company is not seeking to return to the office but adapting to a more remote world. This form of remote working — in many cases enforced by the isolation economy — is now seen as the new norm. The company is driven by guiding principles which are often stated and enforced in both recruitment and promotion. Distribution of power to where it is needed matters. Teams will often swarm around problems, leadership is transient in nature and leaders will arise to fit the problem. In this world, hierarchy is unimportant and few care about the top floor office or the status symbols of power. What motivates people are customer and societal outcomes. Outcome not output matters. The projects undertaken always consider the wider community and sustainability is not a buzz word but a core belief. In support of this, a deep understanding of supply chains are essential, these tend to be modelled as the company views that it is responsible for its entire supply chain. Ethics also matter a lot, it drives external communication, it is not an add-on. Awareness of the market also matters, it is systemic (throughout the organisation) and not the function of a sole leader but instead everyone. To train people, the company used scenarios and gameplay, usually online. The idea of Eve online being a training tool is not an alien concept. In terms of future technology, the company views that AI will replace some tasks and augment some functions currently undertaken by humans. It also consdiers the future of the company to be currently one of growth with positive times ahead.
Most of the readers will exist in companies between these extremes.
Notes on the work.
First, I’d like to say a huge thank you to all those involved in the research groups and the numerous targeting tests, interviews and the survey. As promised to all involved, I have published the results above.
There are a couple of additional notes that I would like to add.
Large company test
The past exercise in 2011 was often dismissed for being “just startups” hence in this population study I added a question on company size. The same phenotypic differences and views on the future were found in larger companies — see figure 14. If anything those large organisation that exhibited next generation characteristics tended to be even more positive about their future than the startups in their population group. Large organisations that exhibited traditional characteristics also tended to be less negative than startups in their population group. There was no significant differences in the population groups related to company size or in other words, this is not a “startup” thing … well, not according to the survey.
Figure 14 — Selecting for Large Organisations alone.
Doctrine
The next generation vs traditional table provides a list of behaviours exhibited by companies. However, behaviours themselves are a function of values (i.e. beliefs) and underlying principles (the corpus of which is known as doctrine) through complicated interactions such as the collective in question, competition with others, enablement systems and power structures — see figure 15. For a more detailed discussion on mapping culture, see the “off the beaten track” posts.
Figure 15 — Culture Map
I am currently mapping out the behaviours to see if any new doctrine (i.e. universally useful principles) can be established. Currently the doctrine table stands as described in figure 16. This might change with the addition of new principles as a result of this work.
Figure 16 — Doctrine (list of universally useful principles).
Limits on
As noted above, behaviours are a function of values and underlying doctrine. As such, some of these behaviours might be destructive if the right values or doctrine are not already established i.e. “leaderless leadership” might be destructive if an organisation does not have values (beliefs) of honesty or trust. We are yet to establish contra indications for these behaviours but would remind the reader to think carefully about the context in which in they are operating in.
Leaderless leadership
One of the more controversial aspects of the new behaviour is leaderless leadership. Prototypes of this form of thinking can be found in Holacracy, Burning Man, GameStop/WSB, Amazon, Haier, Buurtzorg and many other organisations. It should be noted that all these behaviours are emerging practices in much the same way that DevOps was an emerging practice in 2010 (when the previous study was in full swing). Hence, there are no “answers” as to how to do this … yet. Of course, by the time there are answers then the practice will have evolved to the point that any differential competitive advantage will have almost vanished.
The problem with leaderless leadership is that it is likely to suffer significant sources of inertia because of existing power structures. When we talk about powerful and charismatic leaders that is often their perception of themselves. The idea that power can be gained by giving power to others may not sit comfortably with those used to hoarding power especially when those symbols are used to hide personal insecurities.
We have already seen aspects of this with the changes caused by the isolation economy. Whilst many have attempted to adapt physical practices of learning, collaboration and inclusion to virtual practices there has been a constant pushback from issues related to status and power. In some cases this has manifested itself as attempts to recreate status in a virtual world with designed avatars, plush home studios, ideas around NFTs, control of zoom room meeting functions and other devices. The loss of symbols such as location (top floor office) to the fancy office (furniture etc) are powerful sources of inertia to change. Do not underestimate the amount of political capital wrapped up in physical things and the willingness of executives that are unable to adapt to push the company back into the office.
On the future
We are aware that some companies have not yet fully implemented (and in cases, not even started) the earlier behavioural changes from 2011. Some are just starting their journey to devops and cloud today. Others do not even have those most basic of principles in place, for example a focus on user needs or understanding the details. Many of these companies exist in protected spaces (e.g. protected by regulation) and have yet to face the full force of competition. Unfortunately, competition waits for no-one and Van Valen’s Red Queen effect means that in a competitive ecosystem, you have to evolve simply to survive. Many are falling so far behind that once those protective barriers are overcome then they will find it a struggle to adapt or even to survive.
For reference, I have provided the full list of behaviours that your firm should ideally exhibit in this amalgamated table (figure 17) which includes behaviours from 2011 and 2021.
Figure 17 — the combined behaviours.
I will come back and re-examine this subject in 2031 and see if we can’t find a next next “next generation”. I’m going to have to do something about this labelling.
For those wishing to say “this is just for startups” … well, this is the second time I’ve run such a population study and maybe you’ll be right this time. I doubt it. We will find out in 2031. And, that’s the point.
The problem with such a study is that whilst it can be proved to be wrong (e.g. there are no distinct populations), it can never be conclusively proved to be right. It is a model and all models are wrong. Circumstances can change i.e. in the extreme, a Carrington event could wipe out our communication systems or a horde of rampaging Zombies that hunt small groups of humans could arise forcing us to huddle in larger groups … the future is an uncertainty barrier we can peek around but not through. There are no crystal balls.
If you really don’t like the future outlined in this article then you can just bang the “prove it to me” drum and feel comfortable staying where you are. For others, there are examples of companies out there who are forging this path and so now is the time to be listening and experimenting.
I will talk more about this subject and the evolution of companies at Map Camp.
Addendum
A number of questions have arisen from this article. I will try and answer the most common.
“Do you have more details on those next generation practices and how to implement them?”
No-one does. They are emerging which means there will be many experiments that fail. You’re not going to be able to avoid a bit of experimentation here.
“How do we get started on this?”
If you’re looking where to start, I would use the doctrine table provided in figure 16. It will be updated as new principles are found but it’s as good a place as any. The best way to start is to discuss the doctrine within the company. I normally put up a miro board, add the table and ask people to add post-it notes with Blue for “good” and Orange for “poor”. It can get quite messy but I always find the discussion worthwhile. An example for you.
NB, I also use doctrine when assessing M&A (it gives me an idea of adaptability) and also when examining competitors (it gives me a good idea as to what sort of gameplay I can use). One final thing on the miro board … don’t use it as a source of power. I’ve seen people use miro boards where only execs could add post-it notes or the consultants running it. This is quite literally the daftest thing you could do unless you want to disenfranchise everyone. It’s up there with insisting that everyone keeps their camera on in a zoom meeting.
“I work for a company that looks traditional and is trying to get everyone back into the office”.
There are a lot of executive power issues wrapped up in the whole back to office demands. Don’t underestimate the importance of status symbols such as top floor offices etc. A company can also quite happily survive in a more “traditional” form if it is competing against others that look just like it or the industry is somehow protected from outside competition. The main motivation in these companies often tends to be financial i.e. you work there for the money. If you’re not being well rewarded then … you might want to consider other options especially as you no longer live in a world where changing jobs means changing home and schools. Obviously, this is not possible for everyone, it depends upon the industry.
“Can people stop these changes?”
They will try if it impacts existing power structures that are beneficial to them. The problem is the technological causes behind these changes are social media, collaboration and access to data. Whilst you can limit the impact of these technological causes by legislation (for example) those same technological causes are an essential part of solving our sustainability issues.