I have more ideas than time. So I give each idea a score for how useful and how quick it is, and do the best ones first. It's the same trick I write about.
This site teaches how to make decisions with clear options, scores and thresholds. It would be odd not to plan the site the same way. Each possible piece of work gets a few scores, a simple rule decides what goes next, and I write down why.
Candidate work items are scored on reader value, learning value, effort and dependency, combined into a priority with a minimum threshold for inclusion in the next cycle. Decisions are recorded with confidence and reviewed quarterly. No commercial criteria apply until 2027.
Why plan this way
The site is built around one idea: that decisions get better when you name the options, score them, set a rule in advance and write down what you chose. If I planned the site by gut feel, I'd be ignoring my own advice. So the plan is a small decision model, applied to the question "what should I work on next?"
It also keeps me honest about purpose. I'm using this site to learn by teaching, which is the Feynman technique. That means the plan should favour work that teaches me something and helps a reader understand something. Commercial considerations don't come into it; I've set a rule for myself that the site carries no offers until 2027.
The options
Each candidate piece of work is one row. Typical rows are a new article, a new method, a fix to something broken, a tool, or a change to how the site works.
The scores
| Score | Question | Scale |
|---|---|---|
| Reader value | Will this help someone understand or decide something? | 1 to 5 |
| Learning value | Will writing it teach me something I don't yet understand? | 1 to 5 |
| Effort | How long will it take, roughly? | 1 (an hour) to 5 (weeks) |
| Dependency | Does other work need this first? | Yes or no |
Priority = (reader value + learning value) ÷ effort, with any "yes" for dependency moved to the top.
The rule
Anything broken comes first, whatever its score, because a broken link or wrong fact costs readers trust. After that, I take the highest-priority rows that fit the time I have. Anything below a priority of 1 waits, because it's either low value or high effort for what it gives.
A worked example
These scores are illustrative, and close to how I'd score some current options.
| Option | Reader | Learning | Effort | Priority |
|---|---|---|---|---|
| Fill empty collections with first articles | 4 | 4 | 3 | 2.7 |
| Build the threshold calculator | 4 | 3 | 3 | 2.3 |
| Write the Decision Diagnostic in full | 4 | 5 | 4 | 2.25 |
| Redesign the home page | 2 | 1 | 3 | 1.0 |
| Add a newsletter | 2 | 1 | 2 | 1.5 |
Filling the empty collections comes first: many collection pages showed no articles at all, which looks broken to a reader. The calculator and the Diagnostic are close; the calculator wins slightly because it's less effort. The home page redesign sits at the threshold and waits.
Recording the choice
Each planning round gets a decision record: what I considered, what I chose, how confident I am that it'll be useful, and when I'll review it. The review happens in the quarterly notes, where I compare what I planned with what shipped.
What I'm still checking
Learning value is the score I'm least sure how to judge in advance. Some topics I expect to be easy turn out to teach me the most, because explaining them exposes a gap. I may start scoring it after writing instead, as a way of calibrating my guesses.
Sources
This article describes how I plan this site. It uses no external facts or figures; the scores are illustrative.