Errata (corrections to the book) are here.
Here are some of the bigger changes to our book The agile Manager (small a) (version 7.6), as we move towards a possible second edition sometime in 2020 or 2021:
So What, p10
The concepts that this book deals with, in advancing enterprise work and management (especially management), rise from an even deeper philosophical shift which we are not qualified to discuss in depth. But we see the new thinking everywhere, in many models we know of, which all point to this renaissance, this flip, step-change, new age… As well as Graves’ two tiers, there are Game A / Game B (modelling changing social behaviour based on complexity theory), Theory X / Theory Y (two modes of management), Alpha and Beta management, and even The Age of Aquarius – slightly tongue in cheek, but even the hippies of the 60s saw The Times They Are a Changin’.
We describe it as returning to values over value. Friedmanism – the doctrine that the sole purpose of an enterprise is to make money – caused great harm to corporate ethics. It needs to be, and is being, rejected.
Social thinking is crossing a boundary (and just in time, looking at world politics). We are breaking through to new levels of enlightenment. Feminism, diversity, egality, socialism, empowerment, digital, virtual, complexity, networks, … so many powerful forces are at play.
You don’t need to embrace a religion, or worse still the antiscience of the counter-culture, in order to grow into these higher-level concepts. They’re the future of humanity. The future is already here, it is just not evenly distributed. We see it everywhere, it has been arriving for decades.
As well as many models which point to the concept of a flip, moving into a new phase, there is another concept that emerges everywhere: the reunification (in the West) of the transcendentals, Science, Art, and Ethics – of truth, beauty, and goodness – which parted ways in the Enlightenment, when Truth rose above all.
This comes out in
• New Age counter culture: alt med, organic food, anticapitalism, antiscience
• Organic systems, holistic, Thriveability
• Spiral Dynamics from Graves, Cowan, Beck…
• Bring your whole self to work, diversity
• Ethical business
• The Open movement
We are not here to say any or all of these models are accurate. All models are wrong, some are useful. And there is a lot of bullshit peddled in their names, especially antiscience. In fact, it is hard to find material about some of these models, e.g. Integral, that isn’t steeped in nonsense. The point is that it is evident that something big is going on across society, and overall it looks positive. It is an optimistic view of human evolution, which is entirely justified based on historical data: humanity gets better. Across the world, organisations and societies are primarily competitive capitalist green with a few pioneering into teal, and a larger long tail spreading all the way back through orange and red to primal magenta (or grey, or whatever, models aren’t consistent). It is a rising bubble of social progress.
So What, p11
We will always need somebody who accepts a goal from the organisation, and authorisation to use the organisation to achieve it. They will take responsibility for developing people, or resources of the organisation, or functions and processes, or the delivery of work. That’s management, even if it may look wildly different in future. It may not be a job position, or even a role that people adopt, but it is an activity that will happen.
Also, we will always need somebody observing the work, close to it, with time for thinking about better ways, and coming up with solutions to challenges. Right now, that’s the middle managers: they are the often-unsung engine of innovation in most organisations. Maybe it will be performed by coaches and scrum-masters in future, but that could still be called management.
New Ways of Working, p13
The full list is now:
Human: people, humanity, wholeness, culture, sharing, empathy, diversity, inclusiveness, egality, trust, integrity, authenticity, open, transparency, curiosity, learning, mastery, pride, empowerment, freedom, authorisation, servant manager, safety, wellbeing, flourishing, health. [states]
Systems: customer, value, flow, feedback, quality, lean, streams, iteration, networks, complexity, chaos, emergence, antifragile, shift left, teams, organisation, collaboration, ritual, sharing, resilience, human error, holistic, data, science. [artefacts]
Agility: ambiguous, uncertain, iterate, increment, experiment, explore, observe, adjust, fluid, organic, improve, curious, embrace failure, fail fast, small, granular, simplify, flexible, pragmatic. [actions/adjectives]
[…and we could add so many more. Maybe we will.]
New Ways of Managing, p21
Management 3.0 has the “Martie” model:
We are all saying the same things: management of human systems agility is empowering, organic, and exploratory.
Another major trend is the separation of personnel-management from work-management.
This leads to a range of dysfunctional management behaviours, some of which we discuss in this book. We see classic management mistakes being made over and over. We are passionate about preventing the harm.
- Big-bang change
- Change done to people instead improvement done by people
- “Transformation” as a finite project
- Expectation that culture can change quickly
- Treating culture as a simple system not a complex one
- Belief that management know the answers
- Starting with a restructure
- Only doing a restructure
- Most of all: failure to change management and governance.
Added new major section Leadership, p52
[To see it, buy the next edition of the book when it comes out]
People blossom, p54
Almost everybody has a potential contribution. A weed is just a plant in the wrong place. If somebody is not performing, consider that it is a failure of management to recognise what their value is and help them find where to deliver it.
Complex systems are not just complicated. There are fundamental differences:
||With sufficient study, we can link causes to effects.
||There is no predictable link between cause and effect. The same cause can produce differing effects.
||Output is proportional to input.
||Inputs can produce unexpected and disproportionate effects.
||You can understand the system as the sum of its parts.
||The system is the sum of the parts and their interactions. Behaviour is emergent.
||Behaviour can be controlled, and problems solved.
||Behaviour is unpredictable and problems are obtuse.
||Can be constrained and bounded.
||Responds to external environment across fuzzy boundaries.
||Can be fully described and modelled.
||Can only be understood by interacting with it.
||Are changed by external forces.
||Will learn and change spontaneously, internally.
Added new section Influencing complex systems, p
The definition of a complex system is that there is not a causal link between inputs and outputs: either there is a chaotic randomising element, or the system is just impossible for us to model. We cannot predict the outputs for given inputs. All we can do is experiment.
Our Unicorn Management Model has a comprehensive description of the inputs at all levels and aspects of the work system.
A well-known model for what kinds of inputs we can experiment with is Donella Meadows’ twelve leverage points. But she says in her paper:
Magical leverage points are not easily accessible, even if we know where they are and which direction to push on them. There are no cheap tickets to mastery. You have to work hard at it, whether that means rigorously analyzing a system or rigorously casting off your own paradigms and throwing yourself into the humility of Not Knowing. In the end, it seems that mastery has less to do with pushing leverage points than it does with strategically, profoundly, madly letting go.
What if we “can‘t” fail? What if we are building or running something critical? A bridge, a hospital, air traffic control, nuclear power. There is no zero risk. We fail at those things. If we want to catastrophically fail less often, we need to get better at failing earlier and smaller, improving quality to reduce risk, and – most of all – accepting failure and learning from it. These new ways of working and managing are better.
Added a major section: Agile Executive, p108
We almost called this “Agile Leader” but we want to keep the word “leader” for a behaviour not a role. This section is about the behaviours of senior management with the authorisation to head and lead the organisation. This might be another book one day.
[To see it, buy the next edition of the book when it comes out]
Agile planning, p120
A roadmap can have the simple format of Now/Next/Later. You can organise into streams based on outcomes/epics (or some other grouping), and sort by certainty that it will be done. We add the MoSCoW model to the certainty sort. You can also link dependencies, though don’t overdo this. It can look like this:
Agile People Management, p121
Section renamed to “Personnel management”
But have no doubt Agile Personnel Management is a thing, a movement underway around the world .
Increasingly, the People Management function is moving from a remote silo of the organisation managing staff as an amorphous mass, to an accountable management line, directly dealing with staff to provide personnel services: recruitment, onboarding, development, pastoral care, performance management, career, and wellbeing. Organisations use matrix or helix models that separate people management from work management, so that staff work with two different managers for these two different areas of their employment.
Added new section: Personnel management vs work management, p121
Organisations use matrix structures, or helix , or guilds, to provide personnel management (recruitment, development, pastoral care, performance management, HR administration…) distinct from the product and project work. Work managers and personnel managers collaborate to manage teams, but the two function are clearly distinct, so that staff work with two different managers for these two different areas of their employment . Nobody should have to do both: their goals sometimes conflict, people are better at one than the other, and they make the job too complex.
This is an important principle, keeping the two apart. In future, the world will increasingly separate work management from personnel management. This isn’t a new idea. It has been around forever in various forms. We have tried these ideas in the past. Rob worked in a matrixed organisation in 1985-7. What’s new is that workers aren’t owned resources. They’re free authorised adult knowledge workers who are invited to accept work according to their team’s capabilities.
You may have seen this division revisited lately, e.g. the “helix”, articulated in new terms. It makes very good sense and fits exactly with where we are going.
It is not specialisation; it is recognition that they’re different things. An incident manager is not a training manager just because they are both called manager.
At Teal Unicorn, we also call out a third role, team coach (or “scrum-master”).
A work manager manages lumps of work (a product, process, service, programme, or project) focused on value, deliverables and milestones.
A personnel manager manages groups of people (by skillset, division , location…) focused on wellbeing, development, and performance.
A team coach helps teams form and improve, focused on output, quality, and improvement. They serve teams, not individuals.
This is not a matrix of hierarchies. The only one of these three roles that has anything like people “belonging” to them are the personnel managers who are responsible for them. All three roles – work, personnel, team – are servant managers: gardeners not commanders.
Personnel function, p124
We have our own views on the recruiting process:
- There is no way to predict how well an employee will fit in, or perform, until they do. It’s a complex system.
- As a result, recruitment is mostly HR theatre to create the illusion that we can somehow methodically choose candidates.
- So, HR invents all sorts of unscientific methods such as personality tests and problem-solving interviews.
- All hires should be on a 90-day trial basis, for both parties.
- The best candidates are those who have been referred, preferably by a current employee.
- Go with the instincts of staff who have been coached in selecting for diversity. Rob was once recruited by a little old lady in cashmere and pearls, who served tea on fine china and chatted for an hour. She had superb instincts: she hired an eclectic bunch of misfits who worked miracles.
Ensure that you don’t apply the same measures to everyone. Humans are diverse, not standardised resources. Measure them for their personal strengths, not for some common set of metrics. If you measure people on their weaknesses, you set them up to fail and you destroy their morale.
Better still, don’t manage people by numbers; manage by knowing what is going on. Managing people solely or primarily by numbers is a dysfunction – you will make the wrong personnel decisions. You can drive a car without a dashboard.
Individual contribution to knowledge work is invisible. We choose metrics to assess the state and progress of the organisation, not to judge individuals. Numbers don’t achieve that for knowledge work.
New Ways of Working, p137
When you think about new ways of working, don’t just consider how you work, optimising it and finding enhanced methods, but also what work is. As we respond to endless change, we must gradually redefine what the work is: what you do and why you do it. As you work through this section, have three considerations: better work results, better methods and tools for getting there, and better objectives and values for doing it .
Second, automation limits the opportunities for new staff to gain experience. Apprentices (“Shu”) usually work on “Standard” transactional work which is increasingly automated, making a greater proportion of our work more challenging, requiring experience that it is harder to get
Agile on a large scale, p155
One way to do this is to offer new capabilities as new products rather than features of existing products. While Microsoft makes Word, Excel, and Powerpoint ever more complicated, Google releases a stream of stand-alone simple products.
Unsuccessful new products can be gracefully removed from market at dramatically lower cost than features of similar complexity. After all, removing a feature from a product involves the same contention and coherence costs as adding a feature.
New features are tech debt. They entwine in your existing successful product, and add risk to it. The worst that can happen is you ruin everything.
New products are bets with a limited loss. Worst case, shut them down, other products continue apace.
Added new section Collaborative decision making, p157
When we get people together and authorise them to find solutions, decisions need to get made. There are at least four kinds of collaborative decision making : consensus, majority voting, autocratic decisions, and consent. And a variation on the idea: the advice process. Which one we use should be contextual: all have their place.
The ideal collaborative decision mechanism is assumed to be consensus, where everybody gets a say and we discuss until everybody agrees, but obviously this has the potential to be slow, and may never find a unanimous result.
A much faster way for everybody to be heard is to put it to the vote, but this requires reducing the question to a binary yes/no, which loses a lot of information and misses the opportunity to find other solutions.
To move quickly, e.g. in a state of chaos, the fastest decision-making method is to have a designated person authorised to make the decision. This doesn’t have to mean escalating up to a “boss”. The nominated decision maker may be a relevant role for the situation: an incident manager, the master of a skill, the owner of the product, the affected customer… Or we may simply designate somebody: the most senior person in the room, the person representing the affected guild, or somebody whose turn it is.
We talk a number of times in this book about finding consensus, but consensus is not always possible. The larger the group, and the more important the decision, the harder it will be.
If we can’t get consensus, we can try for consent . All must give consent (“I can live with that”) not necessarily consensus (“I agree with that”).
If we follow Fair Process, consent is the outcome.
The advice process is the simple idea that a person (or team) should feel authorised to make the decision themselves, if they’ve sought advice from the right people (those affected, and those who have relevant experience)
Getting to New Ways, p171
Renamed to “Advancing to New Ways”
Advancing Thinking, p189
Added section: Advancement models
There are many ways to describe advancement in thinking
[Will be a discussion of Integral/Laloux colours]
If you prefer a more tangible spectrum of levels, there is an excellent model in Aaron Dignan’s Brave New Work:
Commitment: When those with power or influence commit to moving beyond bureaucracy.
Boundaries: When a liminal space is created and protected.
Priming: When the invitation to think and work differently is offered.
Looping: When change is decentralized and self-management begins.
Criticality: When the system has tipped and there’s no going back.
Continuity: When continuous participatory change is a way of life, and the organization is contributing to the broader community of practice.
Do not mistake these for steps. They’re more akin to thresholds. The order can change. They can be revisited and renewed. They can interact. But when they happen, they help. If you squint, they feel almost like a plan … written in pencil.
How to move your boss, p195
There are times when courage is required with a boss, to call out dysfunctions and leadership mis-steps, or to provide white space / buffering/ air-cover to teams. Choose your battles: a cause worth fighting for and a right time to engage. Be careful how you engage: manage your emotion and words. Stay engaged: consolidate your wins and manage any fallout.
New major section Advancing the Organisation, p197
[To see it, buy the next edition of the book when it comes out]
Advancing management, p197
Added a section: Create a sense of need for change
Kotter calls it a sense of urgency, but we don’t always need a “burning platform”, only make it clear that the current state isn’t working.
Seddon describes the best tactic: spend time initially exploring the system with management. Go to the gemba, stand in the circle*. Map the value streams. Help them understand how things work, help them see with their own eyes.
When we take managers into their organisation to study the “what and why” of current performance as a system, they can’t avoid seeing exactly what targets and, for that matter, all other arbitrary measures sent down a hierarchy, do to performance. They are left in no doubt that they only serve to make it worse. It is impossible for them to ignore the dissonance created when they have seen it with their own eyes. When they discover that their supposed means of control are actually the reverse, they are forced to conclude that the control they afford is an illusion. **
All the training and proof points will fall on deaf ears until we can create a perceived need for change. We need to take management from unconscious incompetence to conscious incompetence. Show them the management system impedes work (It’s a safe bet it does. If not, your organisation has already advanced beyond conventional ways).
*“Stand in the circle” refers to an exercise by Taiishi Ohno of Toyota, who would draw a chalk circle on the ground next to the production line, and tell an exec to stay in it for hours, and note what they see. Read Ohno’s Method – Creating a survival work culture. Nakane and Hall, Association for Manufacturing Excellence, Target Volume 18, 2002
** Beyond Command and Control, John Seddon, Vanguard.method.com, 2019
Recall that we advised to adopt in an iterative incremental manner, with feedback to continually improve what we have incubated. At each step of adoption, there will be a J-curve. So minimise the impact and learn from it.
No matter whether you are adopting a new work process, a software tool, an organisational structure, new policy, or anything at all, try to break it up into increments. For example, you can break it up over two dimensions: scope of the solution (components, modules…), and people (teams, business units, locations…). Then expand out into those areas incrementally over time, starting with one of the lowest risk bits. Learn, feedback, and reflect at each step. Then deploy more, adjust what or how you are deploying, go back to the drawing board, or abandon the idea entirely.
Added a new section: Tools and techniques, p241
At Teal Unicorn, we have hundreds of games and techniques that we draw on to populate our workshops and consulting. Here are some of our favourites for you to try.
[To see them, buy the next edition o the book when it comes out]
Added a third type:
Communities of Product or clubs, bring together people working in a specific area or product, i.e. serving a common customer
Hacking the organisation, p265
In terms of our earlier growth model (p234), you have only now grown out of Explore mode. Your mandate has been increased to look at incubating new ways of working for the organisation. You can bring those interested triads together into a tribe to begin truly incubating an agreed new way of working for all of your organisation.
Once we think we have something workable, we seek a mandate to roll it out across the organisation (remember: iteratively, incrementally, with feedback – not big-bang!). If executive management have been on their own journey of learning, they will be supportive of the wider mandate and we can roll. If not, try to hold the line: keep the new ways alive within a tribe, or fall back to triads Keep the candle burning: no state lasts long in a VUCA world. A new opportunity will come along soon enough.
Added a section:
The Design-Measure-Analyse-Improve-Control cycle is a more complex statistically-driven mechanism from Lean Six Sigma, a variant of Lean which focuses on mathematical analysis of variation in quality. it has passionate adherents, but we view it as Taylorist, and only useful in steady-sate flow, i.e. large scale manufacturing.
Changed diagram to say “Release” instead of “Deploy”. We deploy code any time, we release it to users.
There are mercifully few corrections to v7.6 so far:
“Servant Manager”, page 103, 9th bullet point should read
● Commitment to the growth of people (Let others lead).