Notes from a talk on Agile management and leadership culture at the Digital Project Managers meetup on 9 March by Chris Davies. Video available.
A large number of the reported causes for failed agile projects are management-related.
This is down to management’s way of thinking.
Why do managers think differently?
Managers still often follow the scientific management principles of Frederick Winslow Taylor:
- It is possible to know all you need to know in order to plan what to do.
- “Planners” and “doers” should be separated.
- There is only one right way to do things.
The harmful divide between “Planners and “Doers”
This distinction between “planners” and “doers” causes an uneven power dynamic.
The powerful (planners) are focused on ambition, politics, mistrust, greed and fear.
The powerless (doers) are focused on resentment and resignation.
This approach is manifested in management creating plans for resources to follow, in milestones, steering groups (the idea that these people set direction and the doers just follow along), progress reports, measuring individual performance, annual budgeting, organisation silos and timesheets.
The split between “planners” and “doers” may have made sense in the early c20th factory system, where you didn’t have an educated workforce. But it doesn’t make sense now that university education is widespread – and it particularly doesn’t make sense for knowledge work, where the “planners” probably won’t be expert in the fast-changing specialist domains of their subordinates.
As Steve Jobs observed:
“It doesn’t make sense to hire smart people and then tell them what to do; we hire smart people so they can tell us what to do”
Why do we tell people how to do their jobs?
We tell people how to do their jobs if the outcomes we want aren’t materialising.
We set objectives, then make plans to meet these objectives, and then carry out the necessary actions.
But we might not get the outcomes we want.
This can happen because of problems in the flow between these steps:
- Knowledge gap: a difference between what we think we know and what we actually know. Assumptions.
- Alignment gap: a difference between what we want people to do and the actions people take.
- Effects gap: a difference between what our effects are and what we want them to achieve.
How does scientific management approach problems?
- Knowledge gap: give me more detailed information
- Alignment gap: I need to give you more detailed instructions
- Effects gap: need more detailed controls
This disempowers people in the organisation.
How else might we approach problems?
Scientific management isn’t the only approach. Prussian General Von Moltke in 1869 advocated a different way.
- Knowledge gap: “Do not command more than is necessary or plan beyond the circumstances you can foresee”
- Alignment gap: “Communicate to every unit as much of the higher intent as is necessary to achieve the purpose” Intent more important than how to achieve it in practice.
- Effects gap: “Everyone retains freedom of decision and action within bounds” everyone decides how to achieve that purpose.
So a superior management approach is to establish alignment on intent and give autonomy on actions.
- Define and communicate intent
- Allow each level to define how they will achieve the intent of the next level up and ‘backbrief’
- Give individuals freedom to adjust their actions in line with intent.
Some examples of this approach
- A BA check-in staff member was presented with a gold customer running late for a flight that was taking off in 20 minutes. All the formal rules suggested that nothing could be done, and she couldn’t get in touch with her manager or with customer services. So she took the initiative to hold up the flight and personally escort a gold customer to the gate, allowing him to make his flight. This was against formal rules, but in alignment with the company’s values “To ensure that BA is the customer’s first choice through the delivery of an unbeatable travel experience.”
- Netflix avoid top-down decision-making. They focus on ‘context, not control’.
- Submarine Captain David Marquet (author of Turn the ship around) divested control to his subordinates. Instead of giving orders, he’d ask people to tell him what they intended to do. He’d then potentially ask a few questions. He’d then give assent. Subordinates would internalise the required dilligence, and grow in their own competence and professionalism. His focus was on providing clarity of purpose.
- Buurtzorg – a nursing company predicated on empowered, independent teams of nurses. Teams hire their own people, and decide how to operate. Patients love it because nurses spend more time with them. And yet the need for care is 40% less than it is in other organisations. The US would save $49 billion a year if it had this system.
- Favi – brass foundry. Legendary for on-time elivery, having not missed a deadline in 28 years. Staff empowered to do what it takes to get results, including delivery by helicopter if that’s what’s needed. This builds trust in delivery and the brand, far beyond the costs.
Changing culture results in bigger gains than changing processes
Simply adopting agile practices will generally give about a 20% benefit.
Adopting an agile culture gives about a 300% benefit. This is much more powerful.
— Richard Atherton (@RathertonRich) March 9, 2017
How people rewarded or punished in an organisation determines your values. Management sets boundaries by how it treats failure. Any cultural change needs to address this.
You need to evolve into theory Y management to realise benefits from agile. Change from theory X management to theory Y management.
Put in place supporting structures, processes and practices.
Role model these behaviors by people with moral authority in the organisation.
Recognise that work is accomplished by teams not individuals. Monitor and value groups.
Divest control within teams. Give teams autonomy and boundaries to work unimpeded.
Encourage people to explore and challenge their personal beliefs. They’ll leave if they don’t like it.
Command and and control versus team-based approaches to work: