Annual Planning: Working Specification
Connecting project decisions to evidence and individual contributions
Objective
Give someone joining a project enough context to identify the consequential choices ahead and propose useful work. Keep the user’s three questions at the center: what decisions will we make over the next year, what existing data will inform them, and what additional data should we generate?
Add the information needed to act on those answers: the project’s objective, available options, decision owner, deadline, current position and the evidence that could change that position. Distinguish reanalysis of existing data from collecting new observations.
Out of scope: a complete project schedule, a budget approval process, an individual treatment recommendation, or a requirement to model every decision. The template records decision authority; it does not confer it.
Intended Use
Use one worksheet per project and a rolling twelve-month horizon. An annual planning meeting establishes priorities. Refresh the worksheet when evidence, constraints or the project’s objective changes, and before a consequential commitment. Copy the priority decisions into the one-slide overview for discussion with the team.
The worksheet should also help Andy decide where to spend time across projects. A useful contribution has a named decision, an output someone can use and a delivery date that leaves time for review. Its value can come from clarifying the question or testing the logic before any model is built.
Structure
| Component | Question it answers | Content |
|---|---|---|
| Project context | What are we trying to accomplish? | Intended outcome, current stage, route to success, constraints and assumptions |
| Decision register | What choices are coming? | Options, current position, owner, deadline and linked decisions |
| Evidence plan | What could change those choices? | Existing evidence, limitations, analysis, new data and usable-evidence dates |
| Contribution plan | Where can I help? | Decision framing, critical review, modeling or study design, with an output and collaborator |
| Decision record | What did we decide and why? | Selected option, rationale, residual uncertainty and conditions for reopening |
The slide contains three priority decisions, with brief project context and a specific request for help. Three is a space limit for this template, not a rule about how many decisions a project should have. The worksheet keeps the full inventory and the reasoning behind each row.
Design Rules
- Write a decision as a choice among actions, including continuing, deferring or stopping where those are feasible. An analysis or study is a means of informing a choice.
- Define success and tradeoffs before comparing options. Record disagreement about objectives separately from uncertainty about facts.
- Give each decision one accountable owner and a named forum. Identify the people whose input that owner needs.
- Compare the decision deadline with the date evidence becomes usable after collection, cleaning, analysis and review. A readout date alone is incomplete.
- Write a test of the current position: what plausible result would lead to another action? If no result would change the action, explain the other purpose of the work before making it a priority.
- Keep observation, inference and assumption distinct. More computation does not create an unobserved comparison or resolve a disagreement about values.
- Identify early commitments that constrain later choices. Include work that must start this year for a decision beyond the planning horizon.
- Record confidence in the evidence separately from team agreement. Consensus alone does not establish that an assumption is supported.
Initial Deliverables
| Document | Reader | Kind | Length |
|---|---|---|---|
| Project index | Someone finding the material | How-to | Short |
| Working specification | Someone improving the template | Reference | Enough to define scope and acceptance |
| Planning worksheet | A project contributor preparing a discussion | How-to and copyable form | Prompts, blank tables and one short example |
| Overview page and slide | The team discussing priorities | Reference and presentation | One slide plus brief filling instructions |
| References | Someone tracing the approach | Reference | Short, with source status |
Acceptance and Revision
Test the template on one real project. Ask the decision owner to review the options and dates, and ask an evidence owner to check feasibility. The first test should reveal a concrete choice, the uncertainty affecting it and a useful next action. A slide containing only deliverables and study names needs another pass through the decision questions.
Keep the revision lightweight: remove fields that nobody uses, retain the link from evidence to action, and record where the discussion changed the work plan. The current template is a proposed practice and has not yet been tested on a real project.