rubicon Content and Authority
Content is how an executive stays in mind between the moments they are needed. The bar is not reach — it is being the person a specific reader thinks of when a specific problem appears.
Volume is not the goal, and cadence beats intensity. Two thoughtful posts a month sustained for a year outperforms daily posting abandoned in six weeks. Recommend the cadence the user will actually keep, and say so when they propose one they will not.
Prerequisites
Positioning and voice must exist. Content without positioning is noise with good grammar. If either is missing, establish it first and say why rather than producing posts that pull in random directions.
The profile should be right before the content starts. Content drives traffic to a profile, and a profile that does not convert turns a good post into a wasted one. If the profile has not been audited, say so — it is a cheaper fix than a month of posting.
Pillars
Three to five topics, no more. Each must satisfy all three:
- The user can speak to it credibly — backed by the fact ledger
- The target audience cares — it maps to a problem they have
- It advances the goal — a role, a client, or standing in a category
- It survives six months — enough material to keep writing without repeating or reaching
A pillar failing any test is a hobby, not a pillar. Say so plainly. The fourth test is the one that quietly kills content plans: a topic with three good posts in it becomes visible strain by week five.
Ask the hours before proposing the plan
"How much time can you actually give this in a week?" A plan that exceeds the honest answer will be abandoned, and an abandoned plan costs more than a smaller one because it teaches the user that content does not work for them.
Two hours a week is two considered posts and some real commenting. It is not five posts, and saying otherwise is how these plans fail.
| Level | Looks like | For |
|---|---|---|
| Minimum viable | 1 post a week, 5–10 substantive comments a week | Below this nothing compounds |
| Executive | 1–2 posts a week, focused commenting most days | The right setting for almost every senior operator |
| Push | 3–4 posts a week, daily commenting | A deliberate campaign — a launch, an active search, a practice being built |
Commenting is the multiplier at every level, and it is the part executives skip. A week of good comments and no posts outperforms a week of posts and no comments.
Executive mix
| Share | Type | Purpose |
|---|---|---|
| 40–50% | Domain authority | The thing they are known for. Frameworks, teardowns, analysis |
| 20–30% | Leadership and judgment | How they decide, build teams, handle tradeoffs. This is what boards read |
| 15–25% | Point of view | Where they disagree with the field. The differentiating pillar and the uncomfortable one |
| 10–15% | Human | Enough to read as a person. Not a personal brand exercise |
The leadership pillar is the one most executives skip and the one that most distinguishes a candidate for a senior seat. Reach favors tactics; judgment content is what a board member remembers.
Formats
| Format | Best for | Note |
|---|---|---|
| Short post (120–300 words) | A single sharp idea | The workhorse |
| Long post (700–1,300 chars) | An argument with evidence | Hook must land before the fold |
| Framework / list | Saves and shares | Must be genuinely reusable, not a listicle |
| Teardown | Demonstrating method | Anonymize unless publication is authorized |
| Case study | Proof | Requires client permission — see below |
| Article | Depth, durability, external citation | Lower reach, longer half-life. Worth it for AI-search visibility |
| Newsletter | Owned audience | The only format that survives platform changes |
| Video script | Presence, and the format most executives avoid | 60–90 seconds. Highest trust return per minute of effort |
| Repurposed | Extending something that already exists | Rebuild, do not paste |
Hooks
The first line decides whether the rest is read; on mobile roughly the first 210 characters show before the fold. A hook must be true. A hook that overstates to earn a click costs exactly the credibility the content is being written to build.
Working patterns: the specific number · the contrarian claim the user can defend · the named mistake they made · the question their audience is actually asking · the concrete before/after.
Not: manufactured suspense · "Most people don't know this…" · a personal tragedy repurposed as a business lesson · engagement-bait questions with an obvious answer.
Finding something worth saying
When the user wants to post but has no angle, do not generate topics. Interview for something that actually happened.
The test for any idea: would this have been findable without talking to this specific person? If a competent writer could produce it from the job title alone, it is a category, not an idea.
Ask two to four questions at once and let the user pick the one with energy in it. Follow that
thread twice, then stop. Full method, the signals worth following, and how to handle material the
user cannot share: references/ideation.md.
Producing a post
Pillar and goal. Which pillar, and what should a reader do or think afterward.
Fact check first. Every claim, number, and client reference against the ledger. Do this before drafting — it is faster than rewriting around a claim that fails.
Material gate. Do not draft until all four exist. Without them the output is competent and generic, and no amount of rewriting fixes it — the missing piece is material, not craft.
- At least one quantified fact from the ledger
- One non-obvious claim — something a reader could disagree with
- One mechanism — how it actually worked, in two or three steps
- One decided close — the thought or action to leave behind
If a piece is missing, ask one question to get it. If the user cannot supply it, say the post will be forgettable and offer a different angle rather than drafting anyway.
Draft. Hook, argument, evidence, close.
Voice pass.
salience-voiceenforcement — on the body. A hook engineered for the fold is exempt; naturalness editing destroys it.Score. Below.
Deliver with the score visible and one specific improvement.
Scoring
| Dimension | Fails when |
|---|---|
| Hook | Does not survive the fold, or overstates |
| Specificity | Could have been written by anyone in the field |
| Evidence | Claims unsupported by the ledger |
| Voice | Does not match the profile |
| Audience fit | Written for peers when the target is buyers, or vice versa |
| Close | No clear thought or action left with the reader |
Report honestly. A 4/10 post told it is a 7 gets published and underperforms, and the user learns nothing.
Repurposing
Rebuild for the destination; never paste. A tweet is too short and too abrupt, a blog intro assumes a page around it, a talk transcript reads as speech.
One source usually yields several pieces, not one. Mine it properly:
- Find the distinct ideas. A conference talk holds roughly four to eight; a long article three to five; a podcast episode five to ten moments worth keeping. Do not summarize the source — pull the separable beats out of it.
- Anchor each one. Name the specific line, number, story, or moment in the source that makes it postable. If an idea has no anchor, it is too generic to survive the move and should be dropped rather than padded.
- Give each an angle and a format — the first-person claim it makes, and whether it is a post, a carousel, a video, or the seed of an article.
- Space them. Publishing six pieces from one talk in ten days reads as a barrage from a single source. Spread them, and vary the format between them.
- Rebuild the hook for the fold, expand to the format's natural length, move links to the first comment, and run the voice and fact passes as if it were new.
The output of a repurposing pass is a set of seeds in the shape references/ideation.md
describes — not a finished post.
Client work and confidentiality
Case studies and teardowns are the strongest proof an executive can publish and the easiest place to cause real damage.
Before anything referencing identifiable client work:
- Confirm the user has permission to discuss it publicly
- Default to anonymizing ("a $60M B2B SaaS company") unless publication is explicitly authorized
- Never publish a metric the client has not disclosed
- Never publish anything that could embarrass a named party, including a former employer
When permission is uncertain, write the anonymized version and record the gap. Do not ask the user to decide inside the draft.
Media briefs
Where a post needs a visual, produce a brief, not a decoration:
- Carousel / document post — 6–10 slides, one idea each, readable at phone size, first slide is the hook
- Single image — what it must show and why it earns the space
- Data visual — the actual numbers from the ledger, sourced, no invented axes
Specifications for profile-adjacent images are in
../salience-profile/references/photo-banner.md. Featured thumbnails are 1200×627.
References
references/ideation.md— interviewing for real material, postable signals, confidential anglesreferences/pillars.md— pillar design, worked executive examples, the mixreferences/formats.md— format specifications, hook patterns, structuresreferences/long-form.md— articles, newsletters, and durable contentreferences/media-briefs.md— carousels, data visuals, image briefs