Voice and proof: show the work

Rung 1, the claim. "I am good at forecasting." An adjective about yourself. It costs nothing to write, so it carries no information; everyone in your field can type the same sentence. Most profiles live entirely on this rung.

Rung 2, the description. "I built a forecasting model." Better: it names a thing that happened. But the reader still has only your word for it, and cannot tell whether the model was good, whether you built it or sat near it, or what you learned.

Rung 3, the artefact a stranger can open. A public teardown, a notebook on public data, a talk recording, a worked example. The reader can open it and judge the reasoning for themselves. This is the first rung that works when you are not there to explain, which is the whole point of a reputation.

Rung 4, the artefact someone else used. Somebody linked it, cited it, reused the method, or asked you about it in public. You cannot build this rung directly; it is what rung 3 earns over time. That is why the build-along aims at rung 3 and lets rung 4 follow.

Most of your best work belongs to an employer or a client, and the professional norm is simple: never publish their data, their numbers, their client names or anything a reader could trace back to them. That includes screenshots with the axis labels cropped, "anonymised" figures that are still recognisable, and the name of the client in a caption. If you would not show it to a competitor, do not show it to the internet.

What is yours is the method: the reasoning, the mistake you caught, the approach that fixed it. Rebuild that on public data or on a small toy example you make up, and say plainly that you did: "rebuilt on public data; the original was internal work". The rebuild is honest, it is shareable, and it often teaches better than the original because you strip it to the part that matters. When in doubt, ask whoever owns the work before you publish, in writing.

The chapter titles, as listed on austinkleon.com. The second half of each line is our reading of what the title asks of a working professional, not Kleon's text.

1. You don't have to be a genius. You do not need to be the leading expert to share; being a careful practitioner who is still learning is enough to start.
2. Think process, not product. The steps, drafts and dead ends are shareable long before the finished thing is.
3. Share something small every day. A habit of small, regular pieces instead of waiting for a big reveal.
4. Open up your cabinet of curiosities. Share what you read and learn from, with credit, not only what you make.
5. Tell good stories. Frame work as problem, attempt, result, so a stranger can follow it.
6. Teach what you know. Explaining a method gives it away and makes you the person who explained it.
7. Don't turn into human spam. Read, reply and help as much as you broadcast; the figure above draws the leak.
8. Learn to take a punch. Public work draws criticism; separate the useful correction from the noise.
9. Sell out. It is fine to say plainly what you offer, once the work has earned attention.
10. Stick around. Reputations are built over years, not launches; session 6 comes back to this one.

The third title is the one that gets lifted out of context. Read with its neighbours, it is advice about a working habit: get your process out of your head in small pieces, often, so showing your work stops being an event. It is not a measurement. The title does not say that posting every day grows an audience, and a chapter title could not show that even if it did.

Whether a higher cadence actually buys more engagement is an empirical question with an answer you can count, not a matter of who said it with more confidence. Session 4 tests it on real public data. Until then, treat "share something small every day" as a way to make sharing routine, and keep the separate question of what cadence earns open.

1 · Who you write as. "I" for what you did, "we" only for what a team did, and never "we" to blur who did what. Readers looking for a person want to know which part was yours.

2 · How you hedge. Hedging everything sounds unsure; hedging nothing sounds like a sales page. Put one hedge exactly where your evidence runs out, and state the rest plainly. A hedge in the right place is a signal of honesty, not weakness.

3 · How much jargon. Decide who you write for and use their words. For grocery planners, "line years up by holiday" beats "temporal alignment". Keep a term of art only when the reader would search for it.

4 · The claims you never make. Write down two or three: no result without a number or a link, no "thought leader", no "world-class". A short list of refusals shapes a voice faster than a long list of aspirations.

5 · How you credit others. By name, specifically, every time the idea was not wholly yours. It costs one clause and it is the single most trusted signal on a page about your own work.

Length and format are choices too; pick a length you can sustain and keep it. Writing craft itself, sentence by sentence, lives in the communication course, and drafting, repurposing and AI writing tools live in the AI content course. Here the point is narrower: decide the five, so that every piece sounds like the same person.

Take your three best pieces of writing, anything public or private you were proud of. For each of the five choices, write what you actually did, not what you wish you did. Where the three disagree, pick one and make it the rule. The result fits on an index card and you check every new piece against it before you publish.

Your card will differ. If a line on yours reads like an adjective ("friendly", "bold"), rewrite it as a decision someone could check.