Making systems

Table of contents

• Part XIII: Project plan
• Chapter 44: Plan introduction
• Chapter 45: Multihorizon planning
• 45.1 Updating the plan
• 45.2 As the long term changes
• 45.3 Meeting the objectives
• Bibliography

Part XIII: Project plan

• uncertainty
• multi horizon planning
• resource constrained project

• purpose of plan

• uncertainty
  • definition
  • uncertainty-prioritized planning
  • relative role of integration and large structure
  • relative role of unknown technology

• risk
  • definition of risk as opposed to uncertainty
  • risk register
  • risk mitigation and tracking

• hybrid agile-spiral with per-component status

• risks and benefits of prototyping
• risks and benefits of minimum viable product

Chapter 44: Plan introduction

Plans are a record of

• What tasks are expected to be done
• When
• By whom
• With what outcome

Used by

• Managers to track how a project is progressing
• Staff to understand what to do, and how tasks are related to each other

Help the project move toward a goal

• Helps people make decisions that move the project in the direction it needs to go
• Helps people do work efficiently by avoiding stalls, unneeded work
• Avoids a project becoming a random walk, changing direction constantly and not making progress Lets staff know what to work on
• What working on today or this week
• What the future will hold for them Helps managers manage the project
• Analyze what may happen, what resources or time needed
• Plan for coordinated tasks Maintain flexibility
• As needs change, situation changes & Respect the cone of uncertainty:
  • Things farther in the future are less certain
  • As a project moves forward, tasks and design are worked out and things become more certain
  • Always: this week’s tasks are more certain than the ones a year or two out
• Plan needs regular re-evaluation to adjust
• Planning in too much detail too far out is a waste and gives a false sense of security

Chapter 45: Multihorizon planning

Related ideas

• Agile and spiral development methodologies
  • Both advocate working in increments, with defined goals for increments and re-planning before starting the next increment
  • Agile can lead to a random walk if not guided
  • Spiral can lead to a falsely detailed long-term plan

• OKRs
  • Provide longer-term aspirational goals as well as concrete results that should lead to those goals
  • Can provide guidance in developing a plan
  • Often lead to out-of-date objectives and results that people stop taking seriously

Key ideas:

1. Create a plan that has multiple horizons
   • Iteration/sprint/week
   • Quarter
   • Year
   • Long-term
2. Plan current iteration/quarter/year plus drafts of the next one to three periods
   • Current iteration is a solid commitment
   • Drafts of upcoming periods are less certain
3. Define objectives (and possibly key results) at each horizon
4. Re-evaluate the plan each time one gets to a particular horizon

• Details for each iteration or sprint; more general for longer horizons
• Commit to the current iteration/quarter/year, draft plans out beyond
• Long term provides the general direction

45.1 Updating the plan

At the end of each iteration/sprint:

• Review how the iteration went
• Update the draft for the next iteration
• Add one more draft iteration
• Commit to the next iteration

At the end of a quarter or year:

• Similar process; maintain 1-3 draft periods

45.2 As the long term changes

When the long term goal changes, or when the plan needs to be rearranged:

• Change the longest term affected
• Work backward to see how to achieve that
  • Change the quarterly drafts or current quarter
  • Change the iteration drafts or current iteration

There’s no magic in this… just adjust the plan

45.3 Meeting the objectives

Keep moving toward a goal:

• There are long- and shorter-term goals to help make good choices about the plan for each iteration/sprint
• Having draft iterations and quarters promotes longer-term decisions, not just short-term reactivity

Flexibility:

• The drafts are still just drafts and can be changed as needed
• Draft iterations and quarters reflect uncertainty that changes over time

The plan is still a plan

• It expresses schedule and tasking information just as any other kind of plan
• Supports normal schedule analyses (e.g. critical paths, resource loading)
• It expresses task assignments for people
  • Staff have a longer-term view than with pure agile, but a more realistic understanding of uncertainty than a detailed long-term schedule

Bibliography