The Structures of Unstructured Decisions

NameTypeImprovement
1 - Decision Recognition
Phase 1 - Identification
Encourage people to seek, talk and write down ideas / opportunities. Spread ideas. To know we're not hitting expectation requires good monitoring (a data culture), an ability to talk about company problems, great goal setting and an internal motivation to accept degradation or stagnation and turn the vector instead towards growth.
2 - Diagnosis
Phase 1 - Identification
Instill cultural value of diagnosing well first, as it'll save time and resource for later. Share diagnosises with teams. Bring in analysts who can share learnings, otherwise they'll just be outsourced the job.
3 - Search
Phase 2 - Development
Have search experts. Nodes that know where things are or where to look. I.e. "Librarians". Leave pockets of knowledge around if people have already tried to solve the problem. Encourage early (and novel) problem solving to use off the shelf tooling, and only custom-designs when there is a validated problem-solution fit that needs optimisation.
4 - Design
Phase 2 - Development
Leave design blueprints of previous projects. Make sure to involve cross functional teams for projects that are cross functional. Encourage feedback and iteration design loops.
5 - Screen
Phase 3 - Selection
Had standards of screening. 80% of market solutions looked at, custom solution assessed. What does a good screening process look like? It's a waste to have to come back to this routine when you find another potential candidate you forgot.
6 - Evaluation & Choice
Phase 3 - Selection
Offer standardised criteria to choose from. Choices should be explained rather than using black box thinking - encourages better decisions.
7 - Authorisation
Phase 3 - Selection
Involve at the earliest and appropriate stage, esp if it's a contentious topic. Help those below authority understand how authority thinks and how to work with it. Ensure that authority doesn't become unrealistic, overpowered, biased etc - have checks and balances on authority. Instill cultural values that authority can be challenged. Not at all times, but at constructive times, and that it's safe to do so.
8 - Decision Control
Supporting Routines
Train or document the concepts of designing the decision process. This prevents things happening on the spot, adhoc and unprepared.
9 - Communication
Supporting Routines
Encourage strong, high quality and effective communication, always. Doesn't mean it's insensitive. It's a 5 min catch up when needed, or an hour presentation where everyone is prepared. Might be async or live. Whatever's best to communicate and mobilise action towards global goals.
10 - Politics
Supporting Routines
Is it possible to have people's personal interested align very strongly to company ones? Regardless, let's make it transparent for everyone. We should know what each person wants and make decisions that are more often than not, win-wins. User comes first then it's your team then it's you. Let's instill and encourage this thinking. We shouldn't allow power to be concentrated, and the dispersal of power should be possible under certain circumstances, e.g. for the greater good or against 'evil' actions.
11 - Interrupts
Dynamic Factors
Interrupts will always happen, it's a forming a standardised method of dealing with it. Dealing with it so we can move on swiftly, so that we aren't too destablised (as a person, team or company). Locus of control concepts relate here. In the paper's model Interrupts slow down the core decision process, but forget to mention that it also creates new opportunities / problems / crises. Thus let's build an appreciation for interrupts.
12 - Scheduling Delays
Dynamic Factors
Discourage the use of scheduling delays where possible. If it's important it should be prioritised up and dealt with in a flow manner. If it neccessarily needs chunking up then sure, but let's not let procrastination or in-action develop. This is done by having a culture of accountability and supportive assistance. Champion the behaviour of high quality and high speed executive of important things.
13 - Feedback Delays
Dynamic Factors
Feedback should be placed much higher than other assets in people's minds. Feedback to people, products, projects should be seen as gold. Good feedback takes time, but we mustn't forget that great feedback is also well times. Feedback loses relevance after a certain time period. Let's figure out what the optimal balance may be. Hypothesis: timely good feedback is better than ill-timed great feedback. We will absorb feedback better when we can relate it to recent experiences.
14 - Timing Delays / Speed Ups
Dynamic Factors
Managers should strive to become masters of orchestration. Orchestrating people, resources, decisions etc, but also in the timing of the parts coming together. Speeding up or slowing down a project shouldn't causes stress on those involved, so shouldn't be done in black box. Communicating clearly the reasons shares learnings, aligns motivations and builds team rapport.
15 - Comprehension Cycles
Dynamic Factors
Agree with cyclical thinking and routines, so we need to be willing to elegantly step forward and back, up and down to hone our decision quality. But spinning around in circles is also not progressive. There should be thresholds/ constraints drawn, or an understanding of when we're spinning around too much, and an escape hatch / standard rule to let us eject before we end up wasting all our time and resources.
16 - Failure Recycles
Dynamic Factors
One common problem here would be stubborness and an unwillingness to admit failure / defeat, thus leading to implementation of poor decisions. Accountability to decisions (so you can't run away) are important as well as a supportive environment where we celebrate the courage to stop a whole process to save future turmoil. Whether we back to the drawing board, or modify the goals / constraints or put the project on pause, lets document our learnings and see how we can bring this back into the decision making structure to understand how we could have avoided this detour. Maybe we couldn't, and that's okay.