Desiging for decisions

I've been working with an organization, like many others, that wants to improve its decision making processes. Employees are saying things like:

'There's a lack of understanding around decision criteria. We have a tendency to push decisions up and over-bureaucratize. Currently the senior leaders make most of the decisions, but their role should be more about direction setting, and then setting people up to succeed. They need to step back. There's no definition on what decisions should be made at what level. People ask whether a decision is in their remit. There are too many decision points.'

For them, an improved process would:

  • Make it quicker and easier to make decisions whilst maintaining the right level of controls.
  • Create an organization that is more customer focused, market alert and fast on its feet
  • Give staff the mandate and confidence to make decisions at their level without unnecessary referring up

Probing a bit more on this reveals that what people mean by 'decisions' varies from the high level strategic things like 'Shall we buy this business?' to the day to day operational decisions that frontline staff need to make in their interactions with customers. I had an interesting one of the latter the other day at Doncaster Station in the Pumpkin café there.

I ordered a scone and a cup of tea. I asked if the scone could be warmed in the microwave for a few seconds. My server said 'yes', but her supervisor overheard and said 'no'. (I'm guessing it was the supervisor as my server was prepared to obey her). I asked the supervisor why not since I had just been told 'yes'. She said 'we don't warm scones'. To cut a long story short – and with some practice of my assertiveness skills – I got my scone warmed but was baffled to know why I got a customer oriented 'yes' from one server, and a stony 'no' from her colleague – for a simple 15 second request.

Conventional wisdom has it that 'If you can't make the right decisions quickly and effectively, and execute those decisions consistently, your business will lose ground.' It seems that this statement holds good at all levels of the organization. A surly Pumpkin employee is not going to win business for the company. It's easy enough to identify issues around decision making but much more difficult to work out how to actually get to the point of a) making the right decisions b) making them quickly and effectively c) executing on them consistently.

To begin with what are the 'right decisions'? Generally they are the decisions that are going to add value to the organization, and value can on be a range of metrics. In the warmed scone example, the first assistant decided to enable me to enjoy my Pumpkin experience – she was willing to add value to it. The second assistant who refused my request would have diminished the value of Pumpkin in my eyes (and maybe in the eyes of the others in the queue who were listening to the exchange).

Two Danish researchers, Christensen and Knudsen, look at this sort of interaction in terms of decision error – both errors result in organizational loss of value. The second assistant made what they term a Type 1 error: she rejected the superior alternative of offering me good customer service. If the first assistant had gone along with that decision she would have made a Type 2 error – of accepting an inferior alternative.

In terms of organization design they are of the view that hierarchical organizations are less prone to making Type 2 errors – accepting an inferior alternative – because any decision 'needs to be validated by successive ranks of the hierarchy to be approved.' In flat organizations where managers are making decisions in parallel they observe a tendency to minimize the probability of rejecting a superior alternative (a Type 1 error).

They suggest that 'the essence of designing decision-making organizations is to choose a structure that most effectively reduces Type I and/or Type II errors as required by the organization's task environment.' They then describe a reliable way to do this which, I warn you, is not for the fainthearted or more qualitatively oriented among us as it involves six theorems, and heavy duty statistical analysis. But all this finally emerges into four interesting statements

  • Decision making organizations can be constructed so they maximize decision reliability i.e. avoidance of Type 1 and/or Type 2 errors
  • Appropriate incentive structures can help people make better decisions
  • Some people are better at making good decisions than others so think where these people should be placed in the organization
  • Different structures filter information in different ways so some people may be making decisions on inadequate information.

Also they leave us with one 'important but unanswered' question. To what extent is decision making ability a function of (formal and informal) organization structures as well as talent?

The idea that organizations can be designed for good decision making is explored from a different angle in a Harvard Business Review article the Decision Driven Organization the authors note 'Reorganizations are popular with chief executives, who believe that making big structural changes will lead to better performance. … In reality, a company's structure results in better performance only if it improves the organization's ability to make and execute key decisions better and faster than competitors. If you can sync your organization's structure with its decisions, then the structure will work better and performance will improve.' They suggest six steps to take to organize around decisions:

  1. Be clear about which decisions are most important.
  2. Fi gure out where in the organization those decisions need to be made.
  3. Organize your structure around sources of value.
  4. Fi gure out the level of authority your decision makers need, and give it to them.
  5. Adjust other parts of your organizational system to support decision making and execution.
  6. Equip your managers to make decisions quickly and well.' (I think 'people' would be a better choice of word than 'managers')

Valve, a computer games design company, is designed around these six principles in a rather novel way, and judging by their success seems also to avoid rejecting superior alternatives. The company has no hierarchy and their employee handbook answers the question 'How does Valve decide what to work on?' with the response:

The same way we make other decisions: by waiting for someone to decide that it's the right thing to do, and then letting them recruit other people to work on it with them. We believe in each other to make these decisions, and this faith has proven to be well-founded over and over again. But rather than simply trusting each other to just be smart, we also constantly test our own decisions. Whenever we move into unknown territory, our findings defy our own predictions far more often than we would like to admit. We've found it vitally important to, whenever possible, not operate by using assumptions, unproven theories, or folk wisdom.

The Danish researchers, the HRB authors, and Valve make it clear that although a structure can be designed to support good decision making, the people working in the structure should have the will and the capability to make good decisions. How they get to this is perhaps illuminated by the quote below:

Sir, What is the secret of your success?" a reporter asked a bank president.
"Two words."
"And, sir, what are they?"
"Good decisions."
"And how do you make good decisions?"
"One word."
"And sir, what is that?"
"And how do you get Experience?"
"Two words."
"And, sir, what are they?"
"Bad decisions."

ENDNOTE: A great source of information, models, papers, pointers etc. on organizational decision making is Decision Making Confidence.