The ‘Supermarket shelf’ analogy

December 7, 2022 giantknave Brain Farts Autonomy, Autonomy Support, Handelsbanken, John Seddon, Push vs Pull, Service, Taiichi Ohno

I’ve had many constructive conversations with people about a ‘way of thinking’ regarding the relationship between those helping people (value creating) and those enabling this to happen (value enabling).

I wrote this up in a two-part post, which is linked here:

In addition to these two posts, I often use an analogy¹ of a supermarket shelf to assist me in explaining the desirable inter-connectedness between those helping clients, and those (tasked with) enabling this.

Now, whilst every analogy is limited², humanity has derived great value in using them. So this post is my attempt at writing down my ‘supermarket shelf’ analogy for others to ponder and play with.

Here goes…

So, how does a physical supermarket shelf work?

A simple diagram to assist:

In pulling according to their needs, the customer determines:

whether what’s available is suitable for their needs (ref. autonomy);
whether to try a product;
what they think about the product once they’ve used it; and
what’s missing from the selection (ref. not engaging/ going elsewhere)

In observing the customer, the supplier uses this as feedback to determine:

which products are being selected, and therefore need restocking (ref. capacity);
what new products to develop and test (ref. experimentation); and
which products to improve or discontinue (ref. learning)

Note that the customer decides what combination of products goes in their basket (i.e. what selections work for their needs) …no matter how unexpected these may appear to the supplier.

The supplier’s role is to (safely) enable the selection, whilst providing knowledge to the customer as to what they are selecting and how to get the best out of it.

…and what doesn’t happen?!

The supplier DOES NOT decide what products to put into the customer’s basket.

The supplier DOES NOT decide whether the product worked out for the customer.

In short, the supplier DOES NOT control the customer. Rather, they aim to understand and enable them.

The analogy with People-centred services

If a social system wants to truly help people (its clients) then it will achieve this through autonomy support (rather than attempts at control).

The following diagram applies the supermarket analogy to a social system by using the roles defined within Part 1: Autonomy – Autonomy Support – Autonomy Enabling:

THE big point within this analogy is that ‘the centre’ (i.e. those removed from the client relationship) DOES NOT determine what services a given ‘client – helper working alliance’ is provided with/ can access.

Rather, their role is one of enablement: to ‘stock the shelves’ with a full range of useful tools that will assist, to ensure that it is clear and easy as to how to get the best out of these tools, and to constantly monitor and improve what’s available.

The ‘client – helper working alliance’ decides what combinations to pull and when to pull them as they progress on their relational journey to a better client life.

As set out within an earlier post on Vector Measurement, each client is unique, and their situation complex…so the journey will be about iteratively working alongside them according to what emerges.

This would be the opposite of putting people into a tiered model of services…and moving them up these tiers as things deteriorate (see below on this).

Clarification as to why the shelf is required

You might think “Yes! I agree…we need to get the (attempted) centralised control out of the equation!!” As such, you might be wondering why the helper (the ‘front line’ worker) even needs the centre.

Well, they do…but in a different way than has become the norm.

The helper needs to focus on the client – who they are, what’s going on for them, why this is so, what they need to move forwards.

The helper doesn’t have the time, capability, and resources to develop the host of necessary tools to help. Instead, they need to rely on others to enable them. NOT to dictate to them or refer onwards to…but to focus expertise where it is required.

Example: a client might want/need help in getting an appropriate job. Their helper would really like to be able to pull useful tools like:

- - careers advice with respect to a client’s strengths and experiences
  - knowledge as to what’s available to help develop a client in desirable directions (e.g. education courses, work experience opportunities…)
  - knowledge of the local recruitment agencies, their specialisations, and who/ how to contact
  - guidance on how to write a good CV, and some templates to get the client going
  - interview coaching/advice
  - financial assistance with logistics (e.g. transport, clothing…)
  - …

All of these things would be very useful for the client & helper to have available to them, to pick and choose. It would be mad for each helper to have to design their own.

Conversely, none of this is the ‘centre’ dictating specific responses when a client’s attributes fit certain criteria, such as ‘you need to attend this work seminar, take part in this course/ programme, apply for these available roles, reapply for this benefit, …move from service tier x to service tier y’.

We could find lots of other examples (e.g. in health, education, corrections…), and the point would be the same. The client and their helper are on a journey. The service design needs to enable them, not attempt to categorise and control them.

What has become the conventional norm?

i.e. why is it necessary for me to have to write this post?

The current norm in many large service organisations is to adopt a centrally designed tiered model of services. Such a design involves multiple tiers of escalating service support.

A customer typically starts at ‘tier 0’, where the aim (from the organisation’s point of view) is for the customer to self-solve their needs (#digital).

‘Tier 1’ usually involves the need to ‘talk to a person’ but these ‘generalist’ processors are only allowed and/or equipped to answer the very basic needs.

‘Tier 2’ involves escalation to ‘specialists’ (people who may be able to help absorb some aspects of a client’s instance of variety)

…and we may go up to ‘tier n’ according to how specialised we choose to design.

The thinking is that such a model is cost-efficient (ref. front office – back office).

What’s not seen, let alone appreciated, is that such a design causes huge waste (and therefore cost) and significant client distress:

of people steadily progressing up the tiers because the one before couldn’t meet their needs:
- particularly between tiers 0 and 1 (“I tried online but…)
- and again, between tiers 1 and 2 (“sorry, I can’t help with that. I’ll refer you onwards to…)

of the wasted time and effort involved in managing this transferring of people between tiers (including the recording, and then deciphering of the notes in the system)

of people not being able to access obvious and immediate help that could avert things getting worse (“sorry, your situation isn’t bad enough yet to access ‘tier 2’. Go away until you can show that ‘the wheels have really fallen off’ ”)

of people getting ‘stuck in the system’ (“I can’t help with your actual underlying needs…but I can put this ‘sticky plaster’ over it”)

of things getting worse for people as these transfers occur (“My initial problem was x…but now it’s morphed into 3x, y and z!”)

of the confusion between those working within the various tiers (e.g. tier 2 getting frustrated with tier 1: “why did you do that for/to them? They’re my client!”)

of the disengaged workforce who, confined to their allocated service tier, aren’t allowed to develop and use their capability to actually help the clients before them. They aren’t ‘helpers’ (as per the supermarket analogy). They are merely ‘processors’ (according to a centralised design).

If you operate such a system design, you could go and look (#studyyoursystem) …and see all of this. You probably won’t believe it until you do.

In summary

There’s a paradigm shift between attempting to efficiently handle volumes of events (transactional mindset), and effectively helping people move forwards (relational mindset).

The ‘tiered services’ model is focused on events and efficiency…causing yet more events to occur.

The ‘autonomy – autonomy support – autonomy enabling’ model is focused on journeys and effectiveness… leading to stability and independence.

The source of the analogy

I like to make clear how I got to my (current) thinking. Three stimuli for this post:

In the 1950s Taiichi Ohno (Toyota production system) saw photos of American self-service supermarkets showing products arranged on shelves, available to be selected by customers as and when they required them.

He marvelled at how customers chose exactly what they wanted and how much of it, and then at how the products were replenished (Japan didn’t have self-service supermarkets at that time).

This led to Ohno and his colleagues devising the pull system of production (the opposite of the conventional push), and tools such as Kanban cards to convey the necessary information.

In the 1970s Jan Wallander (Handelsbanken) set out a system whereby the bank’s central services worked to define useful banking products (to put on the shelf), and local branches decided (with their customers) which of these products to use (pull from the shelf), and in what combinations.

Wallander removed the ability of the centre to dictate the usage of their products. He turned them into suppliers to the branch. The branch ‘owned’ the customer relationship and determined if/how the available products helped.

Note that Buurtzorg, and its ‘onion model’ of care, is a modern example.

Over the last few decades John Seddon and his colleagues have repeatedly demonstrated the effectiveness of pulling interventions according to customer need in people-centred services (rather than pushing according to centralised, standardised, specialised design). He’s written several books that clearly set this work out.

A valid criticism of this post

I expect that John Seddon would quite rightly ask of me “why are you writing this down for others to rationally read and then merely apply to their current thinking?”…and thus achieve very little change.

I would like to prompt a level of curiosity: to open ‘managements’ eyes⁴ to the simple fact that the conventional ‘tiered model of services’ isn’t the only way (or even a good way!) …

…and I would then want them to (meaningfully) go to the Gemba, to see the evidence of the dysfunctional performance caused by the current system design, looking at this from the customer/ client’s point of view. #Normative

“The starting point for designing a better system is understanding what’s wrong with the current system.” (Seddon)

The most important part of doing such a thing would be to critically reflect on why things are as they are (which, if we ‘dig deep’ we will find to be ‘thinking things’).

If we don’t uncover this then we will likely just ‘rearrange the deckchairs on the Titanic.’ i.e. the same, but different.

Footnotes

1. “Analogy: a comparison between one thing and another, typically for the purpose of explanation and clarification” (Oxford Dictionary)

2. Limitations of the analogy: All analogies are limited and, if you have the desire to do so, you could choose to kick giant-sized holes in this one. A few example holes that I see:

a) Psychology: those of you aware of the world of consumer products marketing may push back at me with something like “but the supplier IS controlling the customer with the psychology around placement of products on the shelves, packaging, advertising….”

The usage of this analogy is about the customer choosing whether to select, what to select and whether it works for them (ref. autonomy). I ‘get’ that the supplier will use whatever it can to influence the client.

This also applies in people-centred services. A supplier of help (enabler) may want people to, say, stop smoking. As such, it is reasonable for them to understand what interventions work and to use psychology to ‘nudge’ the customer in desirable directions. However, attempts at control will likely backfire (ref. disengagement, defiance)

b) The language of ‘products and services’: A big part of the problem within many service organisations is a ‘product/service’ focus rather than a ‘client need’ focus.

I’d note that service organisations will likely need to create services to help clients, but the analogy is suggesting that these services should be the slave to the client’s needs, not the master. Further, it is making clear that any services need to be able to flex to absorb the variety of the clients before them.

c) Instantly: In a physical supermarket we usually get our need met ‘there and then’. In a people-centred world, it can take time and perseverance for help to ‘take effect’. I’m not wishing to trivialise this.

If I butcher George Box’s ‘model’ quote then “All analogies are wrong, but some analogies are useful”. The question I would wish the reader to ask: ‘is this analogy of some use [to them and others around them]?’

4. Autonomy support: This analogy requires the ‘client – helper working alliance’ to be seen as the customer ‘at the shelf’. It’s not saying that a client can do whatever they want. The autonomy support aspect (the helper role) is critical.

5. On opening eyes: Some will be curious. Some need to be re-awakened from their ‘conventional’ slumber. Others require (what I refer to as) ‘professional provocation’.

Part 2: The problem of changing from ‘this’ (Control) to ‘that’ (Autonomy)

July 15, 2022 giantknave Brain Farts Autonomy, Autonomy Support, Handelsbanken, Jan Wallander, Service, transformation

This post discusses a ‘how’ and follows on from a discussion of the ‘what and why’ in Part 1: Autonomy – Autonomy Support – Autonomy Enabling (a Dec ’21 post – it only took me 7 months!)

I’ll assume you know that the ‘how’ I am writing about is with respect to an approach to moving a large people-centred system:

from (attempted) control
to autonomy (and its enablement).

Please (re)read ‘part 1’ if you need to (including what is meant by people-centred)

Oh for the luxury of a ‘Green field’!

You could be fortunate to start in a relatively ‘green field’ situation (i.e. with very little already in place).

This is what Jos de Blok did in 2006 when he founded the community healthcare provider Buurtzorg in the Netherlands. He started with a few like-minded colleagues to form a self-managed team (i.e. an autonomous unit), and when it reached a defined size (which, in their model, is a team of 12), it ‘calved off’ another autonomous unit.

Buurtzorg carried on doing this until, 10 years later, there were 850 highly effective self-managing teams (autonomous units) in towns and villages all over the country.

In doing this, the autonomous units evolved the desire to have some (very limited) support functions, that would enable (and most definitely not attempt to control) them.

Sounds wonderful.

But many (most) of us don’t have this green field scenario.

We are starting with huge organisations, with thousands of workers within an existing set of highly defined (and usually inflexible) structures. The local, regional and (usually large and deeply functionalised) central model exist in the ‘here and now’.

So, the Buurtzoorg example (whilst recognised as a brilliant social system) is limited.

Rather, we would do well to look to an organisation that successfully changed itself after it had become a big control system. And Handelsbanken is, for me, a highly valuable organisation to study in this regard.

Some context on Handelsbanken

I recall writing about Handelsbanken and their forward-thinking CEO Jan Wallander some time back…and, after searching around, found a couple of articles that I wrote five years ago (how time flies!). A reminder if you are interested:

I’ve added some historical context¹ in a footnote at the bottom of this post, but the upshot is that the results have been hugely impressive, such that they have been written into management case studies and books. Wallander successfully transformed the organisation, for the long term. It is now an International bank (across 6 countries) and turned 150 years old in 2021. It continually wins awards for customer satisfaction, financial safety, and sustainability.

I should deal with a likely critique before I go any further:

“But it’s a bank Steve!!!” Over the last few years, my area of interest has become the social sector (rather than ‘for profit’ organisations)…and, if you’ve noticed this, you may be questioning my use of a bank for this ‘Part 2’ post.

I’d respond that much of Jan Wallander’s thinking fits incredibly well for organisational design within the social sector. He saw that the answer was ‘about the long-term client journey’ (people-centred), within a community and NOT about pushing ‘products and services’.

An introduction to the ‘how’

I’ll break up my explanation into:

Some key ‘up-front’ points;
How Wallander achieved the transformation; and
The core of Wallander’s organisational design

Some key ‘up front’ points

“As a rule, it is large and complex systems and structures that have to be changed if a real change is to take place.” [Wallander]

There is no magic ‘let’s do another change programme’ silver bullet. It is a change in organisational paradigm
It will take time and, above all, leadership (in the true meaning of the word)
Most important for the ‘leadership’ bit is that those leading ‘get the why’. Constancy of purpose can only come from this

How Wallander achieved the transformation

Saying and doing are quite different things. I expect that I could have lots of agreeable conversations with people throughout an organisation (and particularly with those at its ‘centre’) about ‘autonomy at the front line, with enablement from the centre’…and yet nothing would change.

The following quote interests me greatly:

“The reason why [such an ‘autonomy and enabling’ system] is so rare and so difficult to achieve is quite simply that people who have power…are unwilling to hand part of it over to anyone else – a very human reaction. So, if one is to succeed, one needs a firm and clear goal and one has to begin at the top…with oneself.” [Wallander]

Which leads on to…

Wallander’s head office problem

“The staff working at the head office saw themselves as smarter, better educated and ‘more up-to-date’ than the great mass of practical men and women out in the field. They also felt superior because they were in close contact with the highest authority [i.e. the members of the senior management team].

A steady stream of instructions…poured out from the head office departments…amounting to several hundred a month. Even if a branch manager was critical of these instructions and recommendations, he did not feel he had any possibility of questioning them.” [Wallander]

Wallander saw things very differently to this:

“The new policy…aimed at turning the pyramid upside down and making the branch office² the primary units.

In the old organisation there was a clear hierarchy: at the bottom of the scale were the branch offices, and on the next step up the [regional] centres. [‘Success’ was] a move up to a post at the head office.

In the new organisation, it was service at the branch offices that would be the primary merit…[because] it was the branch offices that gave the bank its income [i.e. delivered value to clients, towards their purpose]

In the new organisation:

- the branch offices were the buyers of [enabling services]; and
- the [central] departments the sellers who needed to cover their costs.”

It’s one thing saying this. It’s quite another achieving it. So how did Wallander do it?

Put simply, he stopped ‘push to’ the front line and enabled ‘pull by’ the front line.

Initially, he wanted to ‘stop the train’:

“[Head office] Departments were forbidden to send out any more memos to the branch offices apart from those that were necessary for daily work and reports to the authorities [e.g. regulators].

[All the head office] committees and working groups engaged in various development projects…were told to stop their work at once and the secretaries were asked to submit a report on not more than one A4 page describing what the work had resulting in so far.”

Then he wanted to change the direction:

“An important tool in the change process was the creation of a central planning committee that had a majority of members representing the branch offices.

This committee summoned the managers of the various head office departments, who had to report on what they were planning to do… and how much it would cost the branch offices [i.e. what value would be derived].

The committee could then decide [what was of use to their clients] and [where they considered it was not] the departmental managers were sent back to do their homework again.

In short, those delivering to clients (the branch offices) were given ‘right of veto’ over those from elsewhere proposing changes to this.

This had a dramatic effect. If you are a central department and you don’t want to be wasting your time on unwanted stuff (and who does?), then you’d better (regularly) get out to the front line, observe what’s happening, understand what’s required and/or getting in the way, and then collaborate on helping to resolve.

Further, if (after doing this) a central department develops something that the branches don’t want to use, then the next step turns into dialogue (leading to deeper understanding and valuable pivots in direction), not enforcement.

…which would create adult – adult interactions (rather than parent – child).

The core of the organisational design

I’m going to set out a bunch of inter-related points that need to be taken together for their effect to take hold. Here goes:

1. The local team ‘owns’ the client (relationship):

A client isn’t split up into lots of ‘bits’ and referred ‘all over the place’. Rather, the local team owns the client, as a whole person, throughout their journey. This provides: the client with direct access to the people responsible for helping them; and the local team with meaningful work.

“If you want people motivated to do a good job, give them a good job to do.” (Hertzberg)

2. The local team functions as a semi-autonomous self-managing unit, and is the primary unit of importance:

They are fully responsible for their local cohort of clients and, where they need to, pulling expertise to them (rather than pushing them off to other places).

3. The local team manager is a hugely important role:

This is because they are the linchpin (the vital link). They are responsible for:

the meaningful and sustained help provided to their local cohort of clients;
the development and wellbeing of the local team helping clients; AND
ensuring that the necessary support is being pulled from (and being provided by) the enabling centre.

They are NOT the enforcers of the centre’s rules. Rather they are the stewards of their local community (clients and employees). This is a big responsibility…which means turning the ‘local team manager’ into a highly desirable role and promoting highly capable people into it.

In a reversal of thinking:

the old way was that you ran a local site in the hope of one day moving ‘up’ to head office
the new way is that a job at the centre might be a stepping-stone to being appointed as one of the highly respected local team managers

4. The capability of the local team employees is paramount:

If the local team are to truly help clients, then focused, ongoing time and effort needs to go into developing their capability to do this.

They are no longer merely ‘front line order takers’…so they need help to develop and grow (‘on the job’, ‘in the moment’).

5. Local team employees must receive a reasonable salary:

If we want people to develop and grow such that they deliver outstanding help to their clients, then we need to pay them accordingly. This is not ‘unskilled’ work (whatever that means nowadays).

Putting points 1 through to 5 together creates a virtuous circle: meaningful work, self-direction, relatedness, growth, wellbeing…and back around to meaningful work.

We now turn to some enabling principles:

6. The centre’s job is to enable (and NOT control) the local team:

Basically, re-read (if you need to) what I’ve written above about Wallander’s head office problem and how to transform this.

THE core principle here is that the local teams (via representation) have right of veto over central ideas for change. After all, YOU (the centre) are impacting upon THEIR clients.

For those of you ‘in the centre’ who are thinking “but this is Bullsh!t, the problem is that the local team just don’t get why [my brilliant change] is necessary!” then

either they know something you don’t…and so you’d better get learning
and/or you know something they don’t…and so you’d better get into a productive dialogue with them

…and whichever it is, a dose of experiential learning (at the Gemba) is likely to be the route out of any impasse.

In short: If the local team don’t ‘get it’ or don’t ‘want it’ then the centre has more work to do!

7. All employees (local team and centre) need clarity of purpose and a set of guiding principles:

Whilst we want each and every local team ‘thinking for itself’, we need them all to be going in the same direction. For this to occur, they need a simple, clear (client centred) purpose and principles, to use as their anchor for everything they do…and with this they can amaze us!

Which leads to one of my all-time favourite quotes:

“Simple, clear purpose and principles give rise to complex and intelligent behaviour.

Complex rules and regulations give rise to simple and stupid behaviour.” (Dee Hock)

We also need the ‘centre’ (and most definitely senior management) to live and breathe the purpose and principles*. Without this it’s just ‘happy talk’.

* Note: This list of nine points (and the context around them) is a good start on a core set of principles.

8. There needs to be transparency across local teams as to the outcomes being achieved (towards purpose):

A key role for the centre is simply to make transparent that which is being achieved, so that this is clear to all…and then resist dictating ‘answers’.

This means that each local team can see their outcomes and those of all the other local teams…and can therefore gain feedback, which creates curiosity, stimulates collaboration, generates innovation, and produces learning towards purpose…

…we are then on the way to a purpose-seeking organisation.

Once outcomes are transparent, then if the centre is demonstrably able to help, it is highly likely that local teams will pull for it.

Clarification: Transparency of systemic outcomes towards purpose, NOT activities and outputs (i.e. how many ‘widgets’ were shoved from ‘a’ to ‘b’ and how quickly!)

and last, but by no means least…

9. The budget process needs to be replaced with something better:

I’m not going to write anything detailed on this, as I’ve written about this before. You are welcome to explore this point if you wish: ‘The Great Budget God in the Sky’.

In short, the conventional budget process is also a HUGE problem for a system in meeting its client defined purpose.

Right, I’m one point short of ‘The Ten Commandments’…so I’ll stop there. This isn’t a religion 🙂

To close

The above might seem unpalatable, even frightening, to the current ‘Head Office machine’ but there’s still seriously important work for ‘the centre’ to do…but just different to how you do it now.

In fact, if those in the centre really thought about it (and looked at the copious evidence), how much of what the head office machine currently does provides truly meaningful and inspiring work for you? Perhaps not so much.

Sure, we tell ourselves that we’ve ‘delivered’ what we said we would, according to a budget and timescale…but how are those clients going? Are things (really) getting better for them, or are they stuck and perhaps even going backwards?

How would you know? I expect the local team could tell you!

Footnotes:

1. Historical context: Handelsbanken was, back at the end of the 1960s, a large historic Swedish bank that had got itself into an existential crisis

It was run in the typical functionalised, centralised command-and-control manner, according to the dominant management ideas emanating from American management schools (and which still leave a heavy mark on so many organisations today)

The Handelsbanken Board decided to headhunt a new CEO in an attempt to turn things around (i.e. ‘transform’ in the proper use of the word)

Dr. Jan Wallander was an academic who, through years of research and experience, had arrived at a systemic human-centred management philosophy. To test his thinking, he had taken up the role of CEO at a small Swedish bank (which was becoming an increasingly successful competitor of Handelsbanken)

The Handelsbanken Board were struck with the highly successful outcomes that Wallander had achieved, and they wanted him to rescue them from their crisis.

In 1970 he said yes to their request…but with strings attached: They had to allow him the time and space to put his ideas into practice.

2. ‘Local team’, ‘Branch’, ….: Organisations choose various words to refer to their local presence, be it office, site, shop, outlet, etc. What matters most is the decentralised thinking, not the word chosen or even how the ‘localness’ manifests itself.

I’ve used the generic phrase ‘local team’ in this blog because it implies (to me) a distinct group of people looking after their clients. You’ll note that Jan Wallander refers to the ‘Branch’ in this respect (which fits with his banking world). You can substitute any word you wish so long as it retains the point.

3. The detail for this post comes mainly from Jan Wallander’s book ‘Decentralisation – Why and How to make it work: The Handelsbanken Way’. It is an interesting (and relatively short) read.

I found this book a bit hard to get hold of. I got mine shipped from Sweden…though I had a little mishap of accidently procuring the Swedish language version – sadly, not much use to me and my limited linguistic skills. I happily amended my order to the English version.

4. ‘Within a community’: The conventional public sector model is to have a huge number of branches within a given community – one for each central government department, one for each separate NGO, etc. etc. They are (virtually always) at cross purposes with each other, causing vulnerable clients to have to adopt the inappropriate (and unacceptable) position of trying to ‘project manage’ themselves through the malaise.

The need is to replace this controlling central-ness (lots of central Octopuses with overlapping local tentacles) with meaningful local teams. This is a huge subject in itself, and worthy of another post (or a series of).

Part 1: Autonomy – Autonomy Support – Autonomy Enabling

December 3, 2021August 1, 2022 giantknave Brain Farts Autonomy, Autonomy Support, Donella Meadows, Gerrit Broekstra, Helper, Hillary Cottam, Law of requisite variety, Service, service archetypes, W. Ross Ashby

I wrote the 1st of a 2-part post some time ago, but never published it because I didn’t get around to completing part 2. I’m finally publishing part 1 now…because it might ‘make me’ finish the 2^nd part.

I’ve been on a book-reading¹ and work-experiential journey over many years and have reached a place that I’d like to ‘set out clearly in writing’ (for myself, and others) …and yet I’ve struggled to do this so far (a.k.a writer’s block).

I’ve been talking ‘about’ (but not been writing) this post for months!

So here goes…

Some context:

I (and hordes of other people) work in, around, and across large, complex people-centred systems. Think government departments, NGO’s and so on who are trying to help thousands of clients² better live their lives.

Such systems are usually highly bureaucratic, operating in an (attempted) ‘control system’ way.

They can be helped to see the problematic performance that this system design provides to their clients, but they (we) can then struggle with moving to a different way of thinking and being.

The aim of this post (‘Part 1’) is to set out the ‘what and why’ of moving to a different paradigm, where ‘control’ is replaced by ‘autonomy’. ‘Part 2’ will aim at discussing the rather important topic of a ‘how’ to get there.

As the title and opening image suggest, there’s an ‘autonomy’ chain (from the client, all the way through the system)

What follows is broken up into:

Link 1 (of the chain): Autonomy;

Link 2: Autonomy support; and

Link 3: Autonomy enabling.

Link 1: Autonomy – People-centred…what does that mean?

I used the term ‘people-centred systems’³ above and I should set out what is being meant, and its relevance.

(As usual) I am writing in respect of service systems (as distinct from manufacturing).

Service systems themselves have been broken into different archetypes. I’ve read various ways of setting these out⁴, but here’s a simplified version (of 3 archetypes) from me:

Archetype 1: Transactional where, given the nature of a customer’s demand (a.k.a. need), it should be able to be resolved in a single interaction.

Example: If I contact my home entertainment provider because I want to change aspects of my plan (e.g. gain access to a movie channel), then I should be able to achieve this by the end of that one interaction.

Note: this may or may not happen in the one interaction, depending on how they have designed their system to respond! It might turn into a right dog’s dinner!

Archetype 2: Process where, due to the nature of what the customer/ client needs, it could not be delivered in a single interaction. Instead it needs to enter a process.

Obvious example: If I contact a builder because I want an extension to my house, this can’t be delivered within an initial interaction. Instead, a bunch of (often well understood) stuff now needs to happen (and hopefully flow).

I should reach a point at which the process is completed: I’ve got what I wanted (hopefully efficiently and to a high quality) and I can go on my way.

Archetype 3: People-centred (or Relational) where the nature of the need is about the person (client) themselves. This is far more complex than a transaction or a process. All our social systems fit here.

Each person (client) is unique and recognising this is of fundamental importance if we are to help – we can’t just treat it like ‘a transaction’ or ‘put them through a ‘process’.

To help the person, we will need to achieve engagement with them (reaching and sustaining a trusting relationship), and then understand and support them through a (dynamic and evolving) journey in which they have the most important role to play.

The other word used here is relational, because THE important variable for successful change is the relationship between the person and their helper(s).

“At the heart of this new way of working is human connection. When people feel supported by strong human relationships, change happens. And when we design new systems that make this sort of collaboration and connection feel simple and easy, people want to join in. This is not surprising.” (Hillary Cottam, ‘Radical Help’)

The big mistake that many social systems make is that they are designed along transactional and process lines by a (centralised) control system. A relational response can’t be achieved transaction-ally. Sure, such a system might be able to ‘smash out’ transactions…but it will most likely keep its clients dependent upon it, with repeat visits and, sadly, relapses and deterioration.

So, I’ve tried to explain ‘people-centred’ and you might now think ‘er, okay, so what?’

Well, if we are talking about people, and them achieving meaningful and sustained change, then we need to understand the fundamental importance of self-determination (a.k.a autonomy).

People achieve true change when they choose to do so for themselves.

We end up with the (hopefully) rather obvious point that the best way to help another person is via autonomy support (with this concept set out in detail in a previous post), which can only happen via strong human relationships.

Link 2: Autonomy Support – Variety…and how best to handle it

So, we’ve got:

the uniqueness of every client and their contextual and dynamic situation;
each client (a human being) needing to ‘go there for themselves’; and
helpers (also human beings!) that need to travel alongside them, without judgement or attempts at control.

Looking at this, we can see that we have enormous variety in a people-centred system (in clients, in helpers, in ‘client–helper’ relationships). So, what’s the best way to handle this?

Turning to W. Ross Ashby’s Law of requisite variety to assist…this is explored by Gerrit Broekstra in his book ‘Building high-performance, high-trust organisations’:

“Variety is a measure of complexity and refers to the number of possible states of a system.

[There is a] colossal proliferation of variety in dynamic systems.

Applied to organisations in their complex environments, only matching the variety of the organisation can soak up the complexity of the situation with which it is confronted. No hordes of consultants, planning committees, surveys or management training courses can do the trick.

[The most effective way to deal with such variety is to] be organised on the basis of the principle of self-organisation….as exemplified in autonomous groups.”

Or, to put it into a simple statement:

“Only variety absorbs variety.” [Stafford Beer]

I recognise that the above reads as rather theoretical. Here’s a practical example:

Take any large social system. It is usually trying to serve a large geographic area (perhaps even a small country).

A central control system attempts to:

monitor what’s going on across the whole (ref. KPI’s);
judge from afar (ref. targets, comparisons);
design ‘solutions’ (ref. departments/ functions/ specialist roles, products & services, detailed policies and procedures…)

…and then push these out to all (to be ‘standardised’) and then…‘rinse and repeat’.

Not only does this fail to absorb client variety, it often (usually) frustrates it!

The opposite would be where each ‘locality’ functions, by design, as an autonomous unit – a group of peers who (are enabled to) take responsibility for the wellbeing of their local cohort of clients and helping them with their unique needs.

“Self-managed teams are far more productive than any other form of organising. There is a clear correlation between participation and productivity.” (Margaret Wheatley)

This creates that glorious (yet rare) combination of:

each client feeling that someone actually cares about them;
each front-line worker feeling like they have a meaningful job; and
each unique client situation (variety) being understood, with bespoke support provided (variety absorbing variety)

“Idleness, indifference and irresponsibility are healthy responses to absurd work…if you want someone to do a good job, give them a good job to do” (Frederick Herzberg)

We’ve now added the concept of autonomous units to deciding how best to help and support a cohort of autonomous clients.

So, if autonomous units are what’s required, where does that leave the concept of ‘organisation’?

Link 3: Autonomy enabling – Organisation…what should it be for?

This post is not suggesting that such autonomous units can, or even should, work on their own. It is saying that they pull for help from central support functions when they need to.

The (paradigm-shifting-ly) huge difference is that the front-line autonomous unit is the one leading what they need help with, rather than being dictated to.

They also decide whether any central support provided is actually helpful, and adopt or disregard it accordingly.

The onus is on the enabling centre to provide valuable support (as judged by those helping clients).

“The logical consequences of this way of viewing organisations, with its abundance of freedom of choice for its purposeful parts, unequivocally leads:

- on the one hand, to the autonomy of the operating units and the people contained in it as purposeful entities; and
- on the other hand, the remaining ‘upper part’ of the high-variety organisation as an enabling entity, actively involved in furthering the development of the autonomous unit and their people.” (Gerrit Broekstra)

A sense-check on the purpose of ‘organisation’:

“Hierarchical systems evolve from the bottom up. The purpose of the upper layers of the hierarchy is to serve the purposes of the lower layers…

The original purpose of a hierarchy is always to help its originating systems do their jobs better…

This is something …that…a greatly articulated hierarchy can easily forget.

Many systems are not meeting their goals because of malfunctioning hierarchies.” (Donnella Meadows)

Right, so we’ve worked our way through the concepts of self-determining clients (tick), autonomous units (tick) and an enabling (as opposed to controlling) centre (tick), but if this is sooo good, why doesn’t this eventuate?

That’s the subject of Part 2: the problem of changing from ‘this’ to ‘that’ [Link to Part 2]

Addendum: Towards a picture…

If we’d like to ‘draw a picture’ of what the fundamentals of an effective people-centred system might look like, whilst remembering that “all models are wrong, some are useful” (George Box),

Then here’s an attempt from me…

I’ll build it up in three stages:

Picture 1 – The core:

1. The client (in green) is at the centre, along with the principle of self-determination (a.k.a. autonomy). It is about helping them on their journey, not (attempts at) pushing them through ours.

2. Their family/ Whānau and friends (in yellow) surround them, in recognition that:

no one is an island. The client exists within a social structure (for good, and sometimes, not so good)
to effectively support a client, we need to understand, contextualise and appropriately leverage their social network. This includes helping the client do the same. This might be the most powerful thing we can do.

3. The helper(s) (in blue) directly working with the client (‘value creating’). Their aim should be to create, and retain, a ‘working alliance’, in which there is mutual trust and respect. This will be about:

active/ reflective listening (truly seeing the world from the client’s point of view); and then
supporting the client through their journey, including resolving their transactional and process needs along the way

In doing this, we shouldn’t expect that helper(s) need to know everything. We would hope that they can easily ‘pull’ knowledge (from a useful source) as and when they need it. Note the deliberate choice of the word knowledge, rather than ‘policies and procedures’.

Picture 2 – adding support ‘in the work’:

We’ve now added a support ring around the core.

4. Local support (also in blue) that support the value creating workers in helping clients. Their aim is to be ‘in the work’, to observe the work as it happens, to be pulled ‘live’ as and when the value creating worker hits an obstacle.

Such a pull may result in:

- developing the capability of the helper (‘in the moment’ experiential learning); and/or
- capturing evidence of obstacles outside of their control

5. Accessible help (in red) which represent those things that naturally sit outside the capability and/or capacity of local support.

It (deliberately) isn’t worded ‘available products and services’ because this isn’t the nub of what’s required. Sure, there might be a set of services available to be utilised, but it is up to the local team (with the client) to determine what to pull for, and when.

Further, the client isn’t being ‘referred over to’ the help. The help is being pulled into the client and their value creating worker…. who will also consider whether it is working for them or not.

Picture 3 – adding the enablement of those in the work:

6. Central roles (in purple) whose whole purpose is to enable those working with clients. They don’t attempt to control. Rather, they enable when they are pulled for help in removing obstacles in the way. This is a drastic change to the conventional central function.

The locus of control is with those directly helping the client.

An enabling response might include a focus on developing any of:

- the knowledge available;
- the tools (such as financial and technology support);
- the accessible help (such as adding/ linking to new services, making them more flexible,…);
- the capability of those helping clients (and those supporting this).

A note re. ‘Delivery engine’: Many central functions of large organisations are working hard to become ‘Agile’. I’d concur that it is important to work on ‘how’ to deliver (efficiency). However, of more importance is ‘what’ to deliver (effectiveness).

So, whilst we would want an ‘Agile’ delivery engine, we want it to be delivering to enable (from pulls), not to control (via pushes).

Being ‘Agile’ might be necessary….but it’s by no means sufficient.

And this completes an ‘autonomy – autonomy support – autonomy enabling’ model.

Or (if you are a Kiwi who watches the NZ comedy quiz programme ‘7 Days’) …”and this is my picture”

[Link to Part 2]

Footnotes

1. A book reading journey for this post – for those that are interested:

I first came across the idea of different system types and the transformational importance of matching the appropriate organisational model from reading a brilliant Russell Ackoff essay in his book ‘Ackoff’s Best: His Classic Writings on Management’. If you’d like to understand (or remind yourself of) this, then I wrote a post some years back: “Oh, so that’s why command and control doesn’t work very well!” Today’s post picks up where the ‘what might that [social] model look like?’ question posed at the end left off.
Some years back I read Hope and Fraser’s 2003 book ‘Beyond Budgeting’, which gave me an initial introduction to Jan Wallander and his wonderful thinking in respect of how best to run a large service system for the good of clients, and employees. If you’d like a bit of an introduction to Wallander and his different thinking then I wrote about him here: Back to that ‘Profit Sharing’ Nirvana
This book led me to Gerrit Broekstra’s 2014 book titled ‘Building high-performance, high-trust organisations: Decentralisation 2.0’. This gets into some meaty understanding of the ‘what and why’ of a different model. Broekstra refers to the model as an ‘autonomy and enabling’ paradigm.
I was on a parallel path when I read Edward Deci’s ‘Why we do what we do’, which introduced me to self-determination theory and the hugely important concept of autonomy support.
I had original come across Deci from reading Alfie Kohn’s ‘Punished by Rewards’.
I then managed to get hold of Jan Wallander’s 2003 book ‘Decentralisation – why and how to make it work’, which nicely got me into the nuts and bolts of how he successfully led the change of a service system from centrally controlled to autonomous units. [This is the subject of my ‘Part 2’ post]
I also read Hilary Cottam’s 2018 book ‘Radical’ which cemented my thinking in respect of the need to create relational (rather than transactional) systems when we are trying to help people help themselves.

I then read Miller and Rollnick’s book on Motivational Interviewing (3^rd Edition) which also connects extremely well.

And Motivational interviewing fits well with Gerard Egan’s book ‘The Skilled Helper’.

And this all comes together into an ‘Autonomy – Autonomy support – Autonomy enabling’ logic. Now you know why I was struggling to put something into a single post.

2. Clients, not customers: See footnote on my previous post for a comment on the difference

3. Person-centred system: I believe that the root of this phrase is Carl Roger’s person-centred approach.

4. Service system archetypes: Stuart Corrigan’s 2012 little book ‘The need for change’ was the first time I read about service archetypes. This thinking has been refined/ matured by others since then. Also reference Hillary Cottam and her setting out relational systems.

5. Re. people-centred/ relational: Amusingly (to me), many ‘for profit’ consumer businesses are desperate to be relational with us…even though they are, by their nature, transactional or process system archetypes. They want to forge a relationship with us (their customers) and yet we don’t want it! Think about that communications/ entertainment provider – they’ve often invested millions in expensive ‘Customer Relationship Management’ (CRM) technology that remembers your birthday!

6. Systems within systems: A tricky thing about people-centred systems is that they are likely to involve transactional and process sub-sets within. The important bit (I think) is to realise that the people-centred (relational) system is the master, and transactions and processes are its servant…. they aren’t separate.

A social system example: Let’s imagine that I am working relationally with a client on their journey.

Whilst doing so, they might move address – it should (probably) be a single and simple transaction to handle this change in circumstances.
Further, my relational work might lead to the client wanting help to, say, gain a qualification – this should be a smooth process to set up and then help them to achieve.

Both the transaction and the process sit within the relational journey of helping them better lead their lives (whatever that means for them). Here’s a diagram that hopefully visualises these words:

Autonomy Support

November 19, 2019November 19, 2019 giantknave Brain Farts Autonomy, Autonomy Support, Case Management, Competency, Edward Deci, Relatedness

Over my years of reading and experiential learning, I’ve understood that there are three fundamental elements relevant to a human being’s psychological well-being¹:

Competency;
Autonomy; and
Relatedness

Competency (‘sense of efficacy’) refers to our need to produce desirable outcomes; to be effective whilst experiencing an agreeable level of challenge in doing so; and to experience mastery. Basically, we want to be good at stuff, and we like getting better at it.

Autonomy (‘sense of volition’) refers to our need to feel ownership of our actions; to perceive that we have choices and can self-determine what to do and/or how to do it. As per my previous post (The comedy of…), the opposite of autonomy is control. Put simply, autonomy is the freedom to decide for yourself.

Relatedness (‘sense of caring relationships’) refers to our need to feel connected with others; to care about, and be cared about by, others without ulterior motives.

Together, these three are what makes life meaningful. It follows that they are necessary for meaningful work to occur and should therefore be front-and-centre for anyone (and everyone) trying to design an effective management system.

If you get this right, then just wow, what a place to work!
If you get it wrong… no prizes for guessing what that might look like.

I want to focus this post in on the autonomy piece, and how those in management positions can (and should) interact:

What’s so good about this ‘autonomy’ malarkey?

Edward Deci writes that:

“When autonomous, people are fully willing to do what they are doing, and they embrace the activity with a sense of interest and commitment. Their actions emanate from their true sense of self, so they are being authentic.”

“Autonomy fuels growth and health, because it allows people to experience themselves as themselves, as the initiator of their own actions. The person who feels competent and autonomous, who directs his or her own life, is immeasurably better off than a person who does not.”

Now that’s going to be a powerful force if we can create the conditions for it to occur!

“The key to whether people are living autonomously is whether they feel, deep within themselves, that their actions are their own choice…The main thing about meaningful choice is that it engenders willingness.”

I should note that there are whole bodies of research performed over decades that powerfully back these statements up.

Being in a ‘one up’ position

Deci uses the ‘one up’ language to refer to a person having a role ‘one up’ a hierarchy in respect of another. It applies across many situations, including:

a parent/ guardian to a child;
a teacher to a student;
a manager to a direct report; and
a ‘case manager’ to a client

I’m particularly interested in the last scenario. There are heaps of public sector roles in which someone in (what can generically be referred to as) a case management position has the goal of assisting their client to get to, and stay in, a much better place.

Examples include:

a probation/corrections officer with a prisoner
a nurse/medical worker with a patient
social workers and benefits officers with their clients
….and on and on and on

A ‘one up’ working with their client

A conventional approach to case management is often one of control. It starts with a prescription (“I’m the expert here – this is what’s needed”) and then direction (“and therefore this is what you need to do”).

Deci writes that there are two potential responses to control – compliance or defiance.

The problem with compliance is that the person isn’t actually doing it for themselves, they are ‘going through the motions’. Any outcome is most likely to be superficial, short-lived and requiring of continual (and potentially escalating) maintenance.

..and the (rather obvious) problem with defiance is that you’ve got them running in the wrong direction!

“Compliance produces change that is not likely to be maintained, and defiance blocks change in the first place.”

There are many people out there that distinctly dislike attempting to control others and go in the other direction of permissiveness – i.e. letting people ‘get away with [whatever they want]’. This isn’t healthy – for the client, the system or society. There is a need for a degree of structure, with limits and consequences.

The risks² of operating anywhere on the control spectrum include: undesirable and/or ineffective behaviours; a lack of personal responsibility; and the creation of dependency (on the system).

Introducing the concept of ‘Autonomy Support’

Neither control nor permissiveness are what’s needed for people/clients to achieve meaningful and sustained outcomes. Instead, we need to adopt a totally new orientation that is perpendicular to (i.e. removed from) the control spectrum.

Deci introduces:

“Autonomy Support [as] a personal orientation you take toward other people.

It requires being able to take their perspective – being able to see the world as they see it. It thus allows you to understand why they want what they want and why they do what they do.

As an autonomy supportive [Case Manager], you would be building an alliance with your [Client] and you would engage new situations from that perspective.”

Put simply, Autonomy Support starts from the position that we ‘get’ that a person’s autonomy is key, and therefore our role is to support (nurture, enable, and enhance) a client’s ability to ‘work it out and/or do it for themselves’. This might take some up-front time and effort, but it is undeniably worthwhile.

Right, so that sounds lovely, but how do we go about it?!

How to promote autonomy

The following diagram is my summary of the key points from Deci’s work, which I’ll then attempt to explain and open up. It represents the dynamic collaboration between two people (the person and a ‘one-up’):

The two prominent points are that:

The most important role of the ‘one up’ is to fully and continually understand (the situation, the context, the immediate and emergent needs, what really matters and so on); and

most important for the person/client is to feel like they are in control (autonomous), and this will be achieved through the ‘one up’ providing them with real and meaningful choice.

…but, as you can see, there’s more to it than this. Taking each piece in turn:

Understanding: “to relate to others as human beings, as active agents worthy of support, rather than as objects to be manipulated…It begins with listening openly.”

Put another way, if we don’t really understand what’s going on, how on earth can we expect to help them get to where they need to be? If we merely assume, judge and cajole, we will most likely make matters (much) worse.

Knowledge: This point refers to the person’s/client’s knowledge. If we are going to provide them with choice, then it is our responsibility to ensure they have the necessary knowledge to reasonably decide. It would be unfair not to do so!

Choice: This is perhaps the central point but there are some clarifications to be made:

In some situations, we may not be able to allow choice in the ‘what’. If this is the case, we should turn to providing choice about the ‘how’ to go about the what.

Sometimes the offering of choice isn’t appropriate…HOWEVER, we need to be very careful that we haven’t conveniently ‘presumed’ this simply to control.

Deci notes that a usual critique is that ‘[my person/client] doesn’t want the choice – they want to be told what to do’. He goes on to write that:

“If [this is true] it is because people have been pushed to that point by being overly controlled in the past.

If you control people enough, they may begin to act is if they want to be controlled. [This is] a self-protection strategy…

People adapt to being controlled and act as if they don’t want the very thing that is integral to their nature – namely, the opportunity to be autonomous. They probably fear that they will be evaluated – perhaps even punished – if they make the ‘wrong’ choice. And they may well be.

Being autonomy supportive can be very difficult, especially with people who are accustomed to being controlled…we have to be patient…to work with [them] to reawaken what is basic to their nature and what will almost surely lead to more positive results.”

Limits: Remembering that we aren’t about permissiveness, we will need to be clear on societal limits to a person’s choices and actions.

“Supporting autonomy does not mean condoning irresponsibility, nor does it mean allowing people to engage in dangerous or harmful acts. Central to promoting autonomy is encouraging people to understand where their rights end and others’ rights begin. Setting limits is a way of communicating about people’s rights and about constraints that exist in the social world. As such, it helps people to learn to be responsible in making their choices.”

Some considerations in setting limits:

The person/client should (where possible) be involved in setting the limits, rather than having them unilaterally imposed upon them
The reasoning for any limits needs to be well understood by the person/ client, together with the likely consequences of breaking them
We should aim to keep limits as wide as possible, and allow choice within them
Limits should be communicated in factual, rather than controlling, language

Consequences: “Once limits are set and the consequences communicated, it is important to follow through, else one is undermining one’s own credibility.”

Some key points on consequences:

Consequences must be appropriate to any lapses or contraventions. This MUSTN’T be about punishment or attempting control. It should be about encouraging responsibility
The person/ client should be free to choose whether they stay within the set limits…and then experience the (previously communicated) outcomes of whether they do or don’t.
However, if we (the ‘one up’) don’t follow through, we are really back at permissiveness.

“…it is easy to confuse autonomy support with permissiveness [because] people find it hard to admit that they are being permissive, so they mis-portray their permissive behaviour as autonomy support.

Permissiveness is easy, but autonomy support is hard [yet ultimately rewarding] work”

Reset and repeat: and finally, if and when a person/ client chooses a given course direction (whatever direction this may be), we are dynamically back at understanding.

We don’t judge a client for their choices, we work hard at understanding more deeply why they made them, so that we can go round the ‘autonomy support’ loop until we and they succeed.

In Summary:

It’s not about saying ‘No’ (Control) or ‘Yes’ (Permissive)…it’s about moving from an overly simplistic view and on to real understanding and valuable human interactions.

It may be hard (in the short term) but hugely effective (over time). Conversely, we may be surprised at how easy it is – many a client is just stuck, waiting to be understood!

Future Linkages from this post:

I’ve got a couple of related topics that should ‘dangle’ off this post. They are:

‘Motivational Interviewing’: Discussing a method that fits with equipping those in case management roles with the skills to be autonomy supportive; and

‘Decentralisation 2.0’: Discussing how the central and support functions of large service organisations should interact with their sites to enable, rather than (attempt to) control.

I haven’t written them yet…but I hope to find the time to do so.

Footnotes

1. These three elements are central to Deci and Ryan’s ‘Self Determination Theory’

2. These risks apply equally to the person/client and the ‘one up’ person/ case manager!

3. All quotes from Deci in this post come from his highly readable book titled “Why we do what we do”