| name | meta-algorithm-guide |
| description | Generates a consolidated, per-platform algorithm ranking factors reference and pre-publication checklist for content producers and social media managers. Covers Facebook, Instagram, TikTok, YouTube, LinkedIn, and X/Twitter with EA-specific notes on data costs, WhatsApp substitution, and EAT peak activity windows. Invoke this skill when a client or team member needs a daily operational reference to maximise organic reach, when onboarding a new content producer, or when a client's organic reach has dropped and a diagnostic checklist is needed. |
Platform Algorithm Guide — Organic Reach Reference
Scope: This is an operational reference document, not a strategy document. It covers ranking signals, penalised behaviours, favoured formats, and posting benchmarks for the six primary platforms used in the East African market. Use the pre-publication checklist (Section 8) before every post. For platform-specific strategy, use the relevant platform-* skill.
Use when
- Generates a consolidated, per-platform algorithm ranking factors reference and pre-publication checklist for content producers and social media managers. Covers Facebook, Instagram, TikTok, YouTube, LinkedIn, and X/Twitter with EA-specific notes on data costs, WhatsApp substitution, and EAT peak activity windows. Invoke this skill when a client or team member needs a daily operational reference to maximise organic reach, when onboarding a new content producer, or when a client's organic reach has dropped and a diagnostic checklist is needed.
- Use this skill when it is the closest match to the requested deliverable or workflow.
Do not use when
- Do not use this skill for graphic design, video production, software development, or legal advice beyond the repository's stated scope.
- Do not use it when another skill in this repository is clearly more specific to the requested deliverable.
Workflow
- Collect the required inputs or source material before drafting, unless this skill explicitly generates the intake itself.
- Follow the section order and decision rules in this
SKILL.md; do not skip mandatory steps or required fields.
- Review the draft against the quality criteria, then deliver the final output in markdown unless the skill specifies another format.
Anti-Patterns
- Do not invent client facts, performance data, budgets, or approvals that were not provided or clearly inferred from evidence.
- Do not skip required inputs, mandatory sections, or quality checks just to make the output shorter.
- Do not drift into out-of-scope work such as code implementation, design production, or unsupported legal conclusions.
Outputs
- A structured audit, report, model, or analytical framework in markdown, with decisions and recommendations tied to evidence.
References
- Use the inline instructions in this skill now. If a
references/ directory is added later, treat its files as the deeper source material and keep this SKILL.md execution-focused.
Required Input
Before generating this guide for a specific client or team, collect the following:
- Client business name — trading name as it appears on social profiles
- Industry — sector and sub-sector (e.g. retail: fashion boutique)
- Country / city — default: Uganda / Kampala
- Primary goal — organic reach growth / engagement rate improvement / content team training / reach recovery after algorithm drop
- Active platforms — list every platform the client currently posts on
- Current posting frequency per platform — how many times per week on each
- Primary content format — video / photo / text / mixed
- Team size — solo operator, small team (2–4), or agency setting
1. How Social Media Algorithms Work — Core Principles
Every major platform uses a machine-learning ranking system that predicts whether a given user will engage with a given piece of content. The algorithm is not a fixed rulebook — it is a probability engine trained on billions of interactions. However, consistent signals influence rankings across all platforms.
Universal ranking signals (all platforms):
| Signal | What it measures |
|---|
| Completion rate | Did users watch/read to the end? |
| Saves / bookmarks | High-intent signal — user wants to return |
| Shares / re-shares | Distribution signal — user vouches for the content |
| Comments (meaningful) | Conversation signal — not emoji-only |
| Likes / reactions | Weakest signal but still counted |
| Profile click-throughs | Interest in the creator beyond the post |
| Dwell time | How long a user paused on the post |
Principle: Platforms serve content that keeps users on the platform. Any format or behaviour that achieves this is rewarded. Any format or behaviour that drives users away is penalised.
2. Per-Platform Algorithm Ranking Signals
2.1 Facebook
Facebook uses a feed ranking model that weighs four primary factors: inventory (all eligible posts), signals (engagement data), predictions (likelihood a user engages), and relevance score.
Key ranking signals:
| Signal | Weight | Notes |
|---|
| Meaningful interactions | High | Comments and shares outweigh likes |
| Video completion | High | Native video watched past 60 seconds strongly rewarded |
| Post type relevance | High | Facebook learns what format each user engages with |
| Recency | Medium | Fresh content boosted within first 2–3 hours |
| Profile relationship | High | Pages with consistent interaction history ranked higher |
| Link posts | Low | Outbound links reduce reach — Facebook penalises posts sending users off-platform |
Facebook-specific notes:
- Facebook Groups posts reach further than Page posts in most niches — consider a community group as a supplementary owned channel.
- Stories do not appear in the main feed algorithm but reset daily and keep the page avatar at the front of the Stories bar.
- Reels on Facebook draw reach from both Facebook and Instagram inventory when cross-posted natively (not via third-party tools).
2.2 Instagram
Instagram's algorithm operates separately across Feed, Reels, Stories, and Explore. Each surface has its own ranking logic.
Key ranking signals:
| Signal | Feed | Reels | Explore |
|---|
| Completion rate | High | Very High | High |
| Saves | Very High | High | High |
| Shares (DM shares) | High | Very High | Medium |
| Comments | High | Medium | High |
| Recency | Medium | Medium | Low |
| Account history with user | High | Medium | Low |
| Hashtag relevance | Low | Medium | High |
The engagement window:
Instagram's algorithm evaluates a post's early performance in the first 30 minutes after publication. High engagement in this window signals the algorithm to push the post to a wider audience. Low early engagement suppresses reach. Post when the target audience is most active (see Section 6).
Instagram-specific notes:
- Relevance, Recency, and Resonance (the "3 Rs") are Instagram's own published ranking criteria for Feed.
- Reels between 15–30 seconds outperform longer formats for completion rate in bandwidth-constrained markets.
- Carousels generate repeat views (users swipe back) — this dwell time signals quality.
- Do not post a Reel and then post a static image within 2 hours — Instagram favours spacing posts to prevent self-competition.
2.3 TikTok
TikTok uses the most transparent ranking model of the major platforms. Its primary signal is completion rate — the percentage of viewers who watch a video from beginning to end. Everything else is secondary.
Key ranking signals (in order of weight):
- Completion rate — the single most important signal
- Re-watches — users who replay the video
- Shares — especially off-platform shares (WhatsApp, SMS) which signal viral pull
- Comments — open-ended videos that invite response perform best
- Likes — weakest signal but still counted
- Follows from the video — indicates the content converted a non-follower
TikTok distribution model:
TikTok shows every video to a small test cohort (typically 300–500 users). If completion rate exceeds threshold, it is shown to progressively larger cohorts. This means a new account with zero followers can achieve massive reach on a single video — follower count is not a gating factor.
TikTok-specific notes:
- Keep videos under 45 seconds for EA audiences — data costs mean users rarely watch long-form TikTok on mobile data.
- Use on-screen captions. Many users watch without sound in public spaces (matatus, offices, queues).
- Hook in the first 2 seconds. The algorithm measures whether users skip past the first frame.
- Trending audio accelerates distribution — use audio from the Creative Library or trending local sounds.
- Posting at low-data-cost times (early morning before commute, evening after 8 PM EAT when users are on Wi-Fi) improves completion rates.
2.4 YouTube
YouTube's algorithm optimises for watch time and session time — how long a viewer watches a video and how many subsequent videos they watch in the same session.
Key ranking signals:
| Signal | Notes |
|---|
| Click-through rate (CTR) | Percentage of impressions that resulted in a click — thumbnail and title are critical |
| Average view duration | Absolute minutes watched, not percentage |
| Watch percentage | Percentage of the video watched |
| Likes and dislikes | Sentiment signal |
| Comments | Engagement depth signal |
| Saves to playlist | Strong long-term signal |
| Subscribers from video | Conversion signal |
Series playlists as algorithm signals:
Organising videos into series playlists signals topical authority to YouTube's algorithm. When a viewer finishes one video in a playlist, YouTube auto-queues the next, increasing session time and directly rewarding the channel. Create a playlist for every content series, campaign, or topic cluster.
YouTube-specific notes:
- Thumbnails drive CTR — test two thumbnail variants where possible.
- The first 30 seconds determine whether a viewer stays. Front-load the value.
- Low-bandwidth EA audiences often watch YouTube on mobile data — keep videos under 10 minutes for educational content; longer only if the audience is Wi-Fi-dominant (corporate, campus).
- Shorts (under 60 seconds) are ranked separately and drive channel discovery — use them to attract new subscribers who then watch long-form content.
- Consistent upload schedule (same day and time each week) trains YouTube's recommendation engine to push content at predicted release times.
2.5 LinkedIn
LinkedIn uses three layered graphs to rank content: the identity graph (who you are), the interest graph (what topics you follow), and the knowledge graph (content quality signals). A post is shown first to your direct connections, then to second-degree connections if it performs well.
Key ranking signals:
| Signal | Notes |
|---|
| Early engagement velocity | Engagement in the first 60 minutes is critical |
| Comment depth | Replies-to-comments (threads) score higher than top-level comments |
| Dwell time | LinkedIn tracks how long users pause on a post |
| Content format | Native posts > articles > documents > links > polls |
| Hashtag relevance | Use 3–5 relevant hashtags — more than 5 suppresses reach |
| Creator mode | Accounts with Creator Mode active get broader distribution |
| Native documents (PDFs) | Carousel-style document posts generate high dwell time |
LinkedIn-specific notes:
- LinkedIn penalises posts with outbound links in the main text. Move the link to the first comment and reference it in the post ("Link in the first comment").
- Native video autoplay on LinkedIn is silent — include captions.
- LinkedIn's algorithm favours personal profiles over Company Pages for organic reach. Encourage team members to share or comment — employee advocacy multiplies reach without ad spend.
- The professional context of LinkedIn means posts that teach a skill, share a genuine lesson, or offer a professional insight outperform promotional posts.
2.6 X / Twitter
X uses a ranked feed (For You) and a chronological feed (Following). The For You algorithm is the primary distribution engine.
Key ranking signals:
| Signal | Notes |
|---|
| Engagement velocity | The first 2–3 hours after posting are the critical window |
| Replies | Conversations started by the post signal quality |
| Reposts (RTs) | Distribution signal |
| Likes | Standard engagement signal |
| Bookmarks | High-intent saves signal |
| Profile authority | Verified accounts and accounts with high follower-to-engagement ratios ranked higher |
| Media inclusion | Posts with images or short video outperform text-only |
X-specific notes:
- Threads perform well for educational or narrative content — post the thread as a sequence and reply to the first tweet to extend reach.
- Do not include links in the first tweet of a thread — X suppresses reach for posts with outbound links. Add the link in a reply.
- The For You feed now surfaces content from accounts the user does not follow — this is the primary discovery mechanism for new audience growth.
- X is the dominant platform for journalists, politicians, and public discourse in Uganda — useful for PR, reputation management, and media relations rather than direct commerce.
3. Engagement Window Reference
| Platform | Critical engagement window | What happens after |
|---|
| Instagram | First 30 minutes | Algorithm decides wider distribution based on early signal |
| Facebook | First 2–3 hours | Feed ranking locks in based on initial engagement rate |
| LinkedIn | First 60 minutes | First-degree network determines whether second-degree sees it |
| X / Twitter | First 2–3 hours | For You algorithm scores and distributes based on velocity |
| TikTok | First cohort (300–500 views) | Completion rate in first cohort determines next cohort size |
| YouTube | First 48 hours | CTR and watch time in this period shape long-term search ranking |
Implication: Post when the target audience is online. Do not post and immediately go offline — respond to early comments to signal activity to the algorithm.
4. Favoured Content Formats by Platform
| Platform | Most favoured | Second | Least favoured |
|---|
| Facebook | Native video / Reels | Stories, Carousels | Link posts |
| Instagram | Reels | Carousels | Static single image |
| TikTok | Short vertical video | Stitches / Duets | Static images (limited) |
| YouTube | Long-form video | Shorts | Community posts |
| LinkedIn | Native text posts, Documents | Native video | External links |
| X / Twitter | Threads with media | Short posts with image/video | Link-only posts |
Universal rule: Native content (uploaded directly to the platform) outperforms content linked from external sources on every platform. Never post a YouTube link on Facebook if the goal is reach — upload the video natively.
5. Posting Frequency Benchmarks
| Platform | Minimum | Recommended | Maximum before quality drops |
|---|
| Facebook | 3× per week | 5× per week | 1–2× per day |
| Instagram | 3× per week | 4–5× per week (inc. Stories daily) | 2× per day |
| TikTok | 3× per week | 5–7× per week | 3× per day |
| YouTube | 1× per week | 2× per week | 1× per day |
| LinkedIn | 2× per week | 3–4× per week | 1× per day |
| X / Twitter | 3× per week | 5× per week | Multiple per day (threads count as one) |
Note: Consistency beats volume. Posting 3 times per week every week outperforms posting 7 times in one week and going silent for two weeks.
6. EA-Specific Considerations
6.1 Peak Activity Times (East Africa Time — EAT, UTC+3)
| Platform | Weekday peaks | Weekend peaks |
|---|
| Facebook | 07:00–09:00, 12:00–13:00, 19:00–21:00 | 10:00–12:00, 19:00–21:00 |
| Instagram | 07:00–08:30, 12:00–13:30, 20:00–22:00 | 11:00–13:00, 20:00–22:00 |
| TikTok | 06:30–08:00, 12:30–14:00, 20:00–23:00 | 10:00–14:00, 19:00–23:00 |
| YouTube | 19:00–23:00 (Wi-Fi hours) | 10:00–23:00 |
| LinkedIn | 07:00–09:00, 12:00–13:00, 17:00–18:30 | Low activity |
| X / Twitter | 07:00–09:00, 12:00–14:00, 20:00–22:00 | 10:00–12:00 |
6.2 Data Costs and Video Consumption
Mobile data costs in Uganda remain a significant barrier to video consumption. 1 GB of data costs approximately UGX 3,000–5,000 on most networks (2026 rates).
Implications for content:
- Keep TikTok and Instagram Reels under 45 seconds — longer videos consume more data and are abandoned before completion, hurting the completion rate signal.
- For YouTube, keep educational content under 10 minutes unless the target audience is Wi-Fi-dominant (universities, corporates, Kampala CBD offices).
- Add on-screen captions to all video — users frequently watch without sound to avoid data drain.
- Compress video files before upload — a 720p video performs equivalently to 1080p on mobile screens and loads faster.
- Schedule video posts for evening hours (after 19:00 EAT) when users are more likely to be on Wi-Fi at home.
6.3 WhatsApp as a Near-Zero-Cost Distribution Channel
WhatsApp operates on negligible data relative to video platforms. For many EA businesses, WhatsApp is the primary customer communication and content distribution channel.
- Share links to published content (blog posts, YouTube videos, Facebook posts) via WhatsApp Business broadcast lists.
- WhatsApp Status (24-hour Stories equivalent) reaches opted-in contacts at near-zero data cost.
- WhatsApp does not have an algorithm — delivery is chronological and universal to the contact list. Use this to guarantee reach to the highest-value contacts (existing customers, warm leads).
- WhatsApp engagement (replies, reactions) does not count towards social platform algorithm signals. Use WhatsApp to drive traffic to the platform post, not as a replacement for it.
7. Algorithm-Penalised Behaviours
Avoid the following. Each behaviour either directly suppresses reach or trains the algorithm to associate the account with low-quality content.
| Behaviour | Platforms affected | Why it is penalised |
|---|
| Engagement bait ("Like if you agree", "Tag 3 friends to win") | Facebook, Instagram | Facebook explicitly demotes engagement bait posts |
| Outbound links in post body | Facebook, LinkedIn, X | Sends users off-platform — algorithms suppress |
| Cross-posting identical content | All | Algorithms detect duplicate content; native upload always preferred |
| Clickbait headlines | YouTube, X, LinkedIn | CTR drops when viewer does not stay — algorithm penalises mismatch |
| Buying followers or likes | All | Inflates metrics without engagement; signals low-quality content to algorithm |
| Posting in bursts then going silent | All | Inconsistency reduces algorithmic trust and scheduling priority |
| Hashtag stuffing (30 tags on every post) | Instagram, LinkedIn | Instagram reduced hashtag weight significantly in 2024; LinkedIn caps effective reach above 5 tags |
| Reposting competitors' viral content without adding value | TikTok, X | Flagged as low-effort duplication |
| Ignoring comments | All | Non-response to comments signals low engagement depth; algorithm deprioritises |
8. Pre-Publication Checklist
Use this checklist before every post goes live. Adapt to platform as indicated.
BEFORE YOU POST — QUICK REFERENCE CHECKLIST
Content fundamentals
Format and upload
Algorithm signals
Penalised behaviours — confirm none are present
Post-publication (first 30–60 minutes)
Quality Criteria
Output produced by this skill meets the standard when:
- Platform specificity — every platform section identifies its unique primary signal (completion rate for TikTok, saves for Instagram, watch time for YouTube) rather than giving generic advice applicable to all platforms.
- EA market grounding — data cost constraints, WhatsApp distribution logic, and EAT peak times are integrated into the operational guidance, not appended as an afterthought.
- Checklist usability — the pre-publication checklist in Section 8 is formatted so a content producer can print it or pin it to their screen and use it without reading the full document.
- Penalised behaviours are specific — each penalised behaviour names the platforms affected and explains the mechanism of penalisation, not just that it "hurts reach".
- Engagement windows are actionable — the timing guidance in Sections 3 and 6.1 gives specific EAT times, not vague recommendations like "post when your audience is active".
- Formats table is decision-ready — a content producer can look at Section 4 and immediately know which format to prioritise for their platform without further research.
- No strategy bleed — this document does not recommend content pillars, audience personas, or brand voice. Those belong in the relevant strategy skills. This skill stays operational.
- British English throughout — organisation, behaviour, favoured, recognise, analyse — no American spellings anywhere in the document.
