Postemi growth: from posting to results
This skill is the plan; postemi-writing is how each post is written,
postemi-templates has ready drafts, and the main postemi skill has the
tools and commands. If the writing skills are not installed, write the posts
with the rules of the main skill and offer postemi skills add writing (ask
the user first). Reference files:
- reference/30-angles.md: 30 ways to post about one topic, with a worked example.
- reference/review.md: the post log and the weekly review, step by step.
Principles
- Focus. One or two networks done well beat seven done badly. Add more once something works.
- Every post is a test. Growth early on is learning what this audience wants. Post, read the numbers, keep what works, drop what does not.
- Consistency beats bursts. A steady rhythm the user can keep for months wins over a busy week followed by silence. Long gaps cost reach.
- Big changes, not tiny tweaks. With a small audience, a new topic, angle or format teaches more than changing one word.
- Real engagement only. No bought followers, no engagement swaps, no bots. Comments, shares, saves and replies from real readers are the signal.
- What Postemi does and does not do. Postemi writes nothing on its own, schedules and publishes, and shows numbers. It does not reply to comments, follow, connect, like or send messages for the user. Those are the user's part, and they matter.
1. Choose where to start
Ask who the user wants to reach and what they sell or share, then suggest one main network and at most one second one:
| The user | Start with |
|---|---|
| Sells to businesses, consultant, founder, job market | LinkedIn (personal profile), then X |
| Ideas, tech, writing, building in public | X or Threads, then LinkedIn; Bluesky for tech crowds |
| Visual products, food, fashion, beauty, travel, places | Instagram (Reels and carousels), then TikTok |
| Local business (shop, cafe, salon, trades) | Facebook and Instagram |
| Teaching with video | YouTube (Shorts too) and TikTok |
| An existing community | Telegram, Discord or Slack, next to one public network |
Postemi can send the same idea to several networks at once later (section 6), so starting narrow costs nothing.
2. The first niche and the 30-day test
A niche is not for life; it is the first topic the user goes deep on. The skill of building an audience carries over when the topic changes. Find it with data, not by thinking harder:
- Pick one network (usually X or LinkedIn for text, Instagram or TikTok for visuals).
- Post at least once a day for 30 days. Any idea from the user's
experience: use the idea builder and the look-back test in
postemi-writing. Vary topic, angle (how-to, analysis, inspiration, human nature) and format on purpose, so the data has something to compare. - Schedule it with Postemi a week at a time. Write seven posts with the
user, preview them, and create them in one go (one
socialPostentry per day increate_post, or a--jsonfile with the CLI). The user only has to show up for the replies. - Log every post (topic, angle, format, hook, time): see reference/review.md.
- After 30 days, read the numbers (section 3) and pick the topic and the angle with the strongest early signal.
- Go deep on that topic for the next weeks. Writing about the same topic again and again is not repetition; it is why people follow someone. The 30 angles (section 5) make it easy.
Small numbers still count. If 20 posts got no likes and 10 got one, those 10 did infinitely better: that is the signal. The method is the same at 10 followers and at 10,000; only the zeros change.
3. Read the numbers
Where the numbers are:
- Postemi Insights, in the web app (postemi.com/app/insights): reach, engagement and followers per channel over time.
- The CLI:
postemi analytics:platform <integration-id> -d 30for a channel,postemi analytics:post <post-id> -d 7for one post. If a post answers{"missing": true}, link it first (posts:missing, thenposts:connect; see the main skill). - The MCP tools have no analytics tool. Ask the user to open Insights, or
to paste or screenshot the numbers, and work from that.
list_postsgives the post ids, dates and channels to match them to. - The network's own analytics, when the user has it open.
What to look at, in this order:
- Conversations: comments and replies (and how long they are). The strongest sign a post landed.
- Shares, reposts and saves: people found it worth passing on or keeping.
- Follows and profile visits after a post: it made people want more.
- Engagement rate: engagements divided by impressions. Compares posts fairly when reach differs.
- Reach and impressions: how far it went. Useful, but a post can travel far and change nothing.
- Clicks and sign-ups from the profile link, when the user tracks them.
Likes alone are the weakest signal. Compare posts with each other, not with other people's accounts.
4. How often, and when
Starting points, not laws. The best cadence is the one the user can keep.
| Network | Suggested rhythm | Notes |
|---|---|---|
| X | At least 2 short posts a day, 5 days a week; to grow faster, 3 to 5 a day, 7 days a week, plus 2 or more threads or long posts a week. Text with an image or video does best. | Every post and thread entry counts toward the account's monthly X allowance in Postemi, so tell the user how many are left when the plan needs more. |
| 1 post a day at most, 3 to 5 a week is plenty. | Posting more does not grow faster here. Mornings in the audience's timezone work well. Do not go weeks without posting. | |
| 3 to 5 feed posts or Reels a week, Stories in between. | Reels for reach, carousels for saves. | |
| TikTok | Up to 1 or 2 a day. | Postemi allows about 15 posts per account in 24 hours. |
| Threads, Bluesky, Mastodon | 1 to 3 a day. | Conversation matters more than volume. |
| 3 to 5 a week. | Local and community content. | |
| YouTube | 1 a week; Shorts more often. | Consistency of the day helps viewers. |
| Blog channels | 1 to 4 a month. | The long versions of the best posts. |
Timing: start with morning and early evening in the audience's own timezone
(take the user's timezone from Postemi, and ask where the audience is), then
compare times in the log. postemi posts:find-slot <id> suggests the next
free slot on a channel's schedule.
5. One topic, 30 posts
Once the user has a topic, turn it into a month of posts with the 30 angles in reference/30-angles.md: tips, mistakes, tools, steps, stories, lessons, myths, questions and more. Each angle becomes a question, each answer becomes a post.
Mix the 30 across three to five content buckets (themes the audience expects from the user, for example "how-to", "behind the scenes", "customer results", "opinions") and refuse ideas that fit no bucket. That keeps the account recognisable.
Turning the plan into a Postemi schedule
- Agree the network(s), the number of posts per week and the times.
- Write the first week in full (with
postemi-writing), show it, adjust. - Get the clock and timezone (
get_workspace_context, or the CLI's saved timezone), and spread the posts over the days at the agreed local times. - Preview all of them in one
preview_postcall (onesocialPostentry per post), or show the list from the CLI; fix what the preview flags. - On the user's yes, create them in one call with an
idempotencyKey. Usetype: "draft"(CLI-t draft) if the user wants to review them in the Postemi calendar before they go live. - Repeat weekly, folding in what the numbers say. Writing a month ahead is fine, but leave room for timely posts and for what you learn.
6. Repurpose: one idea, every network
A post that worked is the best raw material for the next ones:
- Across networks, adapted, not copied. The same idea as an X short post,
an X thread, a LinkedIn long post with an image, an Instagram carousel, a
Threads post, and later a blog article. Each gets its own hook, length and
format (see
postemi-writing). In Postemi each channel gets its own entry with its own text in the same request. - Never the same post on two X accounts: one X channel per post.
- Up and down in size. A thread becomes an article; an article becomes five short posts; a popular reply or comment becomes its own post.
- Again, later. Re-share a winner after a few months with a new hook and a fresh example; most followers did not see it the first time.
- Collaborations. Instagram posts can name collaborators (a setting in Postemi); a post together with a partner brand or creator reaches both audiences.
7. Profile, bio and the path to a customer
Posts bring people to the profile; the profile decides whether they follow and click. Help the user write it:
- Bio: "I help [who] [get what]" or "I write about [topic]", plus one line of proof: "Baker for 12 years. 3 shops." Clear beats clever, as with hooks. No list of ten roles.
- One link, to one next step: an e-mail list, a free guide or checklist, a booking page, a product page. On X the bio is the place for links, since Postemi removes links from X posts. On Instagram point to the link in the bio. On LinkedIn, feature the free resource or product in the Featured section.
- Name and keywords: put the topic word people search for in the display name or headline (Instagram, LinkedIn) so the profile can be found.
- The path: post, profile, link, e-mail or sign-up, offer. Most posts just give value; about one in five invites people to the next step (a free resource, a waitlist, the product), with a clear reason to click. On LinkedIn and Facebook that link can go in the first comment.
- Pinned post: the user's best post, or the one that explains what they do and where to start.
8. The user's part every day
Share this with the user plainly; Postemi cannot do it for them:
- Reply to every comment in the first hours after a post goes out; replies bring more reach and turn readers into followers.
- Leave two or three thoughtful comments a day on posts by bigger accounts in the same field. A good comment is a small post of its own, and a comment that does well is an idea for a full post.
- Thank people who share.
9. Review and iterate
Once a week (15 minutes), with the user:
- Pull the numbers for the week (section 3) into the log.
- Name the top three and the bottom three posts, and what they had in common (topic, angle, hook pattern, format, time).
- Decide one bigger change for next week: more of the winning angle, a new format, a different time, dropping a topic that never lands.
- Write and schedule next week with that change.
Once a month: is the main network working? Is the niche still the right first one? Is the profile link getting clicks? Keep, adjust, or try the second network. Details and a log template: reference/review.md.
