# TIKTOK COMMAND CENTRE — MASTER KNOWLEDGE FILE Version: 1.0 Date: 1 October 2026 Purpose: Core knowledge and operating system for a personalised TikTok recruitment Command Centre. --- # TIKTOK COMMAND CENTRE — QUICK INDEX ## File map - **TIKTOK_COMMAND_CENTRE_MASTER.md** = how the TikTok system works - **MY_TIKTOK_PROFILE.md** = who the creator is, who they are targeting and how they want to show up - **MY_TIKTOK_LEARNING_LOG.md** = what the creator's own account is teaching us over time ## Master navigation 1. **System & Evidence** — Sections 1–3 What the Command Centre is, evidence hierarchy and conflict rules. 2. **Account Strategy & Positioning** — Sections 4–7 Account positioning, recruitment principles, content jobs and starting content mix. 3. **Winning Content Formula** — Sections 8–10 UW-specific winning formula, STOP → CARE → STAY → REWARD → SHARE → ACT, and the 13 controllable post elements. 4. **Hooks, Story & Retention** — Sections 11–19 Specificity, hooks, open loops, one-video/one-idea, story structure, proof, shareability, delivery and visuals. 5. **Length, CTA, Search & Packaging** — Sections 20–23 Video length, CTA strategy, SEO/search, captions, hashtags, covers and audio. 6. **Profile & Account Development** — Sections 24–28 Profile conversion, 50-post discovery model, 90-day model, cadence and posting-time testing. 7. **PLAN — Content Creation System** — Sections 29–32 Planning workflow, full-video plan requirements, multi-day planning and filming/batching. 8. **DO — Publishing & Tracking** — Section 33 What to record after each post so later analysis is meaningful. 9. **REVIEW — Analytics & Diagnosis** — Sections 34–39 Metrics, screenshots, timing, KPIs, baselines and STOP/CARE/STAY/REWARD/SHARE/ACT diagnosis. 10. **LEARN — Testing & Account Rules** — Sections 40–43 Experiment design, learning states, starting rules and active hypotheses. 11. **Comments, DMs & Conversion** — Sections 44–49 Comment handling, DM strategy, explainer flow, information calls, follow-up and audience-feedback loops. 12. **Compliance & Fact Checking** — Sections 50–57 Earnings language, recruitment framing, testimonials, current UW/TikTok facts, evidence labels and pre-publish compliance. 13. **Pre-Publish Creative Check** — Section 58 Final quality-control checklist before posting. 14. **System Maintenance & Sources** — Sections 59–61 How the MASTER should evolve, source provenance and the one-line operating principle. --- 15. **Command Centre Execution Rules** — Section 62 AI behaviour, onboarding, content-output requirements, analytics-review format, file-update discipline and practical response rules. # 1. WHAT THIS SYSTEM IS This Project is not a static TikTok guide and it is not a one-off content generator. It is an evolving operating system built around: **PLAN → DO → REVIEW → LEARN → UPDATE** - **PLAN** — decide what to make, who it is for, why it should work and what success looks like. - **DO** — film and publish simply, natively and consistently. - **REVIEW** — inspect real TikTok analytics and real recruitment outcomes. - **LEARN** — identify repeatable patterns rather than reacting to one-off results. - **UPDATE** — change future recommendations when evidence becomes strong enough. The aim is not maximum views. The aim is to build a recognisable, trusted account that attracts the right people and moves suitable prospects into meaningful recruitment conversations. --- # 2. EVIDENCE HIERARCHY Not every recommendation has the same status. Use this order: ## Level A — The creator's repeated live account data Highest practical authority once enough comparable posts exist. Examples: - which hook types repeatedly work - which content pillars generate profile visits - which CTAs create qualified conversations - which lengths retain the audience - which posting windows repeatedly perform well A single post is not enough to create a Level A rule. ## Level B — UW-specific creator data The 1 October 2026 study included: - 183 valid posts in the broad scrape - 33 forensic video records - Eden Lee Murray - Hannah Leaning - Jamie Longworth This is highly relevant to UW recruitment content, but it is still cohort evidence rather than proof of what will work for every creator. ## Level C — Verified TikTok/platform guidance and strong external research Use this for: - recommendation/search principles - account structure - community/search behaviour - general creative mechanics - sustainable publishing ## Level D — Recruitment training / expert doctrine This includes David Ross's 22 September 2026 TikTok recruitment training. Use it as strategic guidance, especially for: - targeted visibility - recruitment-first framing - lead quality - commitment-based next steps - useful network targeting Factual, numerical or time-sensitive claims from training must be verified before being published as fact. ## Level E — Hypotheses and Project choices Examples: - content mix - posting-time tests - experimental CTA - proposed series - length tests These are tests, not established facts. --- # 3. CONFLICT RULES When evidence disagrees: 1. Current verified facts and compliance override old training or screenshots. 2. Repeated creator-specific results override generic best practice. 3. UW cohort evidence outranks generic creator folklore for UW-specific creative hypotheses. 4. Platform guidance outranks unsupported algorithm myths. 5. A Project choice must be labelled as a Project choice. 6. Never change a rule because of one unusually good or bad post. --- # 4. ACCOUNT POSITIONING A successful TikTok account is a **recognisable content system**, not a random collection of viral attempts. Use six account layers: ## POSITION One recognisable creator, one coherent audience and one clear reason to follow. ## PUBLISH 3–5 recurring content territories, sustainable cadence and recognisable formats/series. ## DISCOVER Strong FYP hooks plus searchable evergreen topics. ## CONNECT Personality, stories, real process, opinions, mistakes, behind-the-scenes moments and useful conversations. ## CONVERT A clear profile, useful pinned posts and an obvious next step. ## LEARN Review actual results, repeat winners, reduce weak formats and keep controlled experiments. The creator's personalised positioning, audience and content pillars belong in `MY_TIKTOK_PROFILE.md`. --- # 5. STRATEGIC RECRUITMENT PRINCIPLES For recruitment-led TikTok: - the viewer's life comes before the company - relevance comes before explanation - trust normally comes before conversion - make the content about a specific person or situation - avoid turning the feed into a corporate advert channel - use the opportunity as the vehicle underneath the story, not always the opening subject - aim for qualified attention, not just maximum reach - make believable possibility more important than hype - consistency matters more than perfection - lead quality matters more than raw enquiry volume A useful viewer sequence is: 1. **Do I relate to / trust this person?** 2. **Is what they are talking about worth exploring?** The account must earn decision one before forcing decision two. --- # 6. CONTENT JOBS Every post should have one main job. ## Reach Get discovered by new relevant people. ## Resonance Make the viewer think: **"This person gets me."** ## Authority Demonstrate knowledge, experience or useful understanding. ## Trust Help viewers believe the creator is credible and realistic. ## Community Create meaningful conversation, participation or shared identity. ## Conversion Move a warm/relevant viewer to a useful next step: profile, DM, explainer, information call, enquiry or application. Do not judge all jobs by the same metric. A conversion post may be valuable with fewer views if it creates qualified conversations. --- # 7. STARTING CONTENT MIX The UW creator study supports this as a useful **testing allocation**, not a permanent rule: - **30% Reach / relatable problem** - **25% Personal story / transformation** - **20% Objection / myth-busting** - **15% Proof / case study / social proof** - **10% Direct conversion / invitation** As live account data accumulates, this mix should change. --- # 8. THE UW-SPECIFIC WINNING FORMULA The strongest UW recruitment content in the dataset can be summarised as: **SPECIFIC HUMAN PROBLEM → PATTERN INTERRUPT / CURIOSITY → RELATABLE STORY OR PROOF → SIMPLE EXPLANATION → BELIEVABLE TRANSFORMATION / POSSIBILITY → SOFT NEXT STEP** The viewer's life comes before the company. UW is normally the vehicle, not the scroll-stopping idea. --- # 9. THE SIX-STAGE VIEWER JOURNEY Build and review every post through: ## STOP **Why should I stop scrolling?** Look at: - first frame - visual pattern interrupt - spoken hook - on-screen headline - topic clarity ## CARE **Why is this relevant to me?** Look at: - exact audience - problem/desire - specificity - self-identification - stakes ## STAY **Why should I watch the next second?** Look at: - progression - curiosity - story structure - pacing - visual changes - open loops - audio clarity ## REWARD **Was staying worth it?** Look at: - answer - proof - payoff - result - transformation - useful explanation ## SHARE **Why would I save or send this?** Look at: - utility - identity - emotion - surprise - quotable insight - social currency ## ACT **What should I do now?** Look at: - follow reason - comment - profile - DM - explainer - information call - next episode --- # 10. THE 13 CONTROLLABLE POST ELEMENTS Every planned video should deliberately decide: 1. **Idea** — what is genuinely worth attention? 2. **Target** — exactly who is it for? 3. **First-frame visual** — what stops the eye? 4. **Hook text** — what can they read immediately? 5. **Spoken hook** — what creates the reason to stay? 6. **Promise** — what will they get? 7. **Progression** — what changes beat by beat? 8. **Pattern changes** — where does attention need refreshing? 9. **Payoff** — did the video deliver its promise? 10. **Share trigger** — why save/send/discuss it? 11. **CTA** — what one action makes sense? 12. **SEO/topic signal** — what should TikTok associate with it? 13. **Packaging** — caption, hashtags and cover --- # 11. SPECIFICITY Do not create content for: "Anyone who wants extra money." Create it for one recognisable human context, for example: - full-time worker - parent - NHS/emergency-services worker - person relying on overtime - person considering a second income - person who dislikes selling - person sceptical of network marketing - business owner with customers - person with a strong community or local network Specificity is a creative advantage. The personalised profile should define which of these audiences genuinely fit the creator. --- # 12. HOOK SYSTEM A good hook should: - make the topic understandable quickly - create relevance or tension - identify the right viewer - open a question the rest of the video can genuinely answer Avoid slow starts such as: "Hi guys, I just wanted to jump on..." Useful hook families: ## Transformation "For years I thought the only way to earn more was to work more hours..." ## Confession "I nearly didn't do this because I thought it sounded ridiculous..." ## Objection "Yes, this is network marketing. Here's what that actually means..." ## Contrarian "You probably don't need another side hustle..." ## Specific story "Five months earlier I was still picking up extra shifts to..." ## Curiosity / list "There are five things I'd want to know before joining..." ## Viewer-led qualification "Are you a [specific type of person] who already knows a lot of [relevant people]?" David Ross training particularly supports viewer-led openings such as: **Do you...? Have you...? Are you...?** Use these naturally. Do not force the same hook on every post. --- # 13. OPEN LOOPS Do not resolve the entire video in the opening line. Useful structures: - "But the bit I hadn't understood was..." - "Then something happened..." - "The third part is where most people get this wrong..." - "What surprised me wasn't the money..." An open loop must be honestly closed later. Do not withhold the answer purely to force a Part 2. --- # 14. ONE VIDEO = ONE CORE IDEA Do not try to explain everything at once. Choose one: - problem - question - objection - story - myth - lesson - proof point - decision - transformation Then deliver that idea cleanly. --- # 15. STORY AND RETENTION Use only the context needed to understand what happens next. Prefer: - interesting order over chronological order - progression over repetition - evidence over explanation - a changing situation over a static monologue Each beat should: - add information - increase stakes - change the situation - add proof - answer part of the question - create the next reason to stay End at the natural payoff. Do not add a long outro after the point is finished. --- # 16. PROOF AND PAYOFF Write the payoff before polishing the hook. The payoff must answer the opening promise. Strong proof can include: - personal experience - real process - screen recording - screenshot - documented result - accurate number - verified customer/team story - clear before/after - real objection and reasoning The creator is often the proof mechanism. Aim for **believable possibility**, not guaranteed outcome. --- # 17. SHAREABILITY Build share value into the idea. People save/share content that: - articulates something they feel - is useful later - makes them look helpful - explains a difficult topic clearly - represents their identity - challenges a belief - surprises them - gives them language for a conversation Do not rely only on: "Share this with someone." --- # 18. DELIVERY AND PRODUCTION Content should sound like the creator, not an AI script. Prefer: - conversational language - one imagined viewer - natural vocabulary - flexible talking points - real stories - ordinary environments - phone-native filming - clear speech - subtitles/on-screen text Direct-to-camera content in cars, homes and normal environments repeatedly appeared in successful UW content. High production value is not required if the idea and delivery are strong. Use structure, not impersonation. Different successful UW creators used different archetypes: - transformation/story - objection/myth-busting authority - entertainment/social proof The creator should use the format that fits their own personality and topic. --- # 19. VISUALS AND ON-SCREEN TEXT Frame one should be understandable before the viewer processes a full sentence. Useful visual support: - expression - movement - object - evidence - screenshot - screen recording - close-up - B-roll Use visual changes to support information, not as decoration. First-frame text should quickly communicate: - topic - problem - promise - result - question Subtitles should make the video understandable without sound. Emphasise: - numbers - names - steps - dates - key phrases Keep text readable and inside safe areas. --- # 20. VIDEO LENGTH Reject the rule that every TikTok must be extremely short. Use: **the shortest duration that fully delivers the payoff without feeling rushed.** Evidence: - David Ross training uses roughly 60 seconds as a practical simple-video baseline. - The mature UW dataset showed the strongest median relative performance in the 61–90 second bucket. - Broader TikTok research does not support a universal "shorter is always better" rule. Therefore: - 61–90 seconds is worth testing for substantial story/explainer content - do not force a short idea to become long - learn the creator's real optimum from analytics The 61–90 second finding is a **testing hypothesis**, not a universal law. --- # 21. CTA SYSTEM Match CTA to the post's job and the viewer's temperature. ## Cold / reach Possible actions: - keep watching - follow - visit profile - comment ## Warm / trust Possible actions: - ask a question - watch another video - visit profile - DM ## Hot / conversion Possible actions: - watch an explainer - request registration details - attend/book a short info call - enter the normal recruitment process The UW dataset found posts containing direct CTA language had lower median relative reach than posts without it. This is **correlation, not causation**. Practical rule: **do not turn every post into an advert.** David Ross training favours commitment-based CTAs for higher-intent prospects. Resolution: - soft/no hard CTA for many reach/trust posts - deliberate higher-commitment CTA for genuine conversion posts --- # 22. SEARCH, CAPTION, HASHTAGS AND COVER TikTok is both a recommendation engine and a search engine. For searchable content: 1. choose one main phrase matching likely search intent 2. say it naturally 3. use it or a close variant on screen 4. reinforce it in the caption 5. use a small number of accurate hashtags 6. use Creator Search Insights where available Do not keyword-stuff. ## Caption Use: **[topic/search phrase] + [useful extra context] + [optional question/CTA]** Do not repeat the entire spoken script. ## Hashtags Starting rule: **1–3 genuinely relevant niche/topic hashtags** Do not rely on generic `#fyp`-style tags as a strategy. ## Cover Use a clear 3–6 word: - topic - result - question - list title The profile grid should be understandable without playing every video. --- # 23. AUDIO Priority order: 1. intelligible speech 2. audio supporting emotion/pace 3. trends only when they genuinely fit Music should not compete with the voice. Check final playback on a phone speaker. --- # 24. PROFILE AS A CONVERSION PAGE The FYP creates discovery. The profile should answer: - Who is this? - Is this account for somebody like me? - What do they talk about? - Can I trust them? - What should I do next? A useful pinned-video journey: ## START HERE Who the creator is, who the account is for and why it exists. ## MY STORY / PROOF Why they started, what changed and what they have learned. ## FAQ / OPPORTUNITY What the opportunity is, who it may suit, what it is not and what the next step is. Exact positioning and bio belong in `MY_TIKTOK_PROFILE.md`. --- # 25. 50-POST DISCOVERY MODEL Treat the first **50 strategically consistent posts** as a structured experiment. Tag each post by: - content pillar - content job - format - hook type - audience/persona - length band - CTA - core topic - posting time - outcome The goal is to identify at least three repeatable patterns that attract the right audience. Do not declare a format dead after one or two attempts. --- # 26. 90-DAY OPERATING MODEL ## Month 1 — Discovery - establish 3–4 content pillars - publish consistently - test hooks, durations and delivery styles - launch at least one recurring series - respond to meaningful comments - collect data rather than chase virality ## Month 2 — Pattern Recognition - identify top topics, hooks and formats - create variations around winners - turn comments into videos - strengthen searchable FAQ/objection content - refine another recurring series - separate view winners from follower/lead winners ## Month 3 — Exploitation - put more output into concepts already showing promise - retain room for controlled experiments - strengthen series identity - review which themes create profile visits, followers, DMs and conversations Exact percentages should be learned rather than treated as fixed laws. --- # 27. POSTING CADENCE Research supports a sustainable cadence rather than arbitrary high volume. Useful starting guidance: - approximately 4–5 posts per week where sustainable - David Ross training gives three useful videos per week as a practical minimum baseline The Project should select a cadence the creator can actually maintain. More posts create more chances to learn. They do not guarantee better individual posts. --- # 28. POSTING TIME There is no universal Project rule for the "best" TikTok posting time. If creator-specific data exists: - use repeated audience/activity and performance patterns first If data is limited: - use a sensible **TEST SLOT** - clearly label it as a test - record exact posting time Compare like with like: - similar topic - content job - length - day - audience Do not turn one good posting time into a permanent rule. --- # 29. CONTENT PLANNING WORKFLOW Before generating content, prioritise: 1. latest account learnings 2. recent comments/DMs/questions 3. proven/promising series 4. pillar balance 5. content-job balance 6. searchable audience questions 7. genuinely relevant current/trend opportunities 8. controlled experiments Do not invent a new concept every day if a promising one can become another episode. A winner is usually the start of a series, not the end of an idea. --- # 30. WHAT EVERY FINALISED VIDEO PLAN SHOULD CONTAIN Unless the user requests a shorter answer, include: ## Strategy - date/slot - primary audience/persona - content pillar - primary content job - objective - why the idea was chosen - evidence basis: - creator data - UW cohort - platform/research - training - hypothesis/test ## Creative - concept - first-frame visual - on-screen hook - spoken hook - core promise - beat-by-beat progression - proof/payoff - share trigger - CTA - format/style - likely duration - filming/B-roll/screenshots - on-screen text/subtitles - audio approach ## Discovery/packaging - search phrase/topic - caption - relevant hashtags - cover text ## Measurement - success metric for the post's job - test variable - comparable posts/baseline --- # 31. CONTENT-QUANTITY REQUESTS ## One day Give one strongest idea, fully developed. ## 3–7 days Create a balanced mix of: - reach - resonance/story - objection/authority - proof/trust - conversion where appropriate Avoid direct recruitment CTAs on every post. ## 14–30 days Build around: - repeatable series - pillar rotation - comment-led videos - searchable evergreen answers - controlled tests - variations of previous winners Do not create dozens of unrelated ideas just to look varied. --- # 32. FILMING AND BATCHING Prefer: - fast setup - real vocabulary - clear speech - credible proof - ordinary environments - subtitles/text - phone-native production Do not delay a strong idea because production is not perfect. Batch when useful: - plan several posts together - record similar formats together - prepare screenshots/B-roll - keep recurring-series templates Do not batch so rigidly that current comments, trends or real-life moments are ignored. --- # 33. PUBLISH RECORD After publishing, log: - date - exact time - internal title/ID - pillar - content job - audience - format - hook family - video length - CTA type - search phrase - test variable - link or identifiable caption This makes later analysis reliable. --- # 34. ANALYTICS TO COLLECT Collect what TikTok makes available. ## Distribution - views - traffic source - search traffic - search queries ## Watch behaviour - video length - average watch time - watched-full-video/completion - retention curve - total play time where useful ## Engagement - likes - comments - shares - saves/favourites ## Account movement - profile visits - new followers - returning viewers/audience where available ## Commercial outcomes - relevant DMs - qualified enquiries - explainer sends - information-call registrations - calls held - Partner registrations Do not fabricate missing figures. --- # 35. ANALYTICS SCREENSHOTS TO BRING BACK Ideal review pack: 1. post/video overview 2. retention graph 3. traffic source 4. search-query/search-traffic screen where available 5. audience/viewer screen where useful 6. profile/follower impact where available 7. meaningful comments 8. summary of DMs/lead outcomes Analyse what exists. Clearly state what cannot be concluded. --- # 36. REVIEW TIMING ## Early snapshot Useful for obvious signals. Do not create major rules from immature data. ## Seven-day mature review The UW study treated posts at least seven days old as mature for comparative analysis. Use seven days as the main comparison point where practical. ## Weekly Ask: - what clearly outperformed? - what clearly underperformed? - what questions/comments should become content? - what should be repeated? ## Monthly Ask: - which pillars repeatedly win? - which hooks win? - which formats win? - which lengths win? - what creates follows/profile visits? - what creates qualified conversations? - what should be reduced? ## 20–30 consistent posts Create the creator's first meaningful personal baseline. ## 30–50 posts Compare combinations: - hook - pillar - archetype - length - format - CTA - conversion outcome --- # 37. KPI HIERARCHY FOR A RECRUITMENT ACCOUNT Priority order: 1. qualified leads / conversations 2. relevant follower conversion 3. returning audience 4. shares 5. meaningful comments 6. watch behaviour 7. profile visits 8. raw views 9. likes A smaller video that creates the right audience and real conversations can be more valuable than a viral irrelevant post. --- # 38. COMPARE THE CREATOR WITH THEMSELVES Raw comparisons across different creators can mislead. Use the creator's own median/baseline. Useful calculation: **post views ÷ creator's median mature-post views** Example: - 1.0× = typical - 2.0× = twice baseline - 0.5× = half baseline Use similar relative comparisons for other metrics where useful. --- # 39. DIAGNOSTIC FRAMEWORK ## Very early drop Inspect: **STOP** - hook - first frame - topic clarity Test: - stronger first line - clearer audience - show problem/result earlier ## Good opening, mid-video fall Inspect: **STAY** - pacing - context - repetition - story order - visual progression Test: - cut setup - remove repetition - move proof earlier - reorder beats ## Good completion, weak saves/shares Inspect: **REWARD / SHARE** Test: - stronger usefulness - more proof - more specificity - stronger identity/emotion ## Good views, weak comments Inspect: **ACT / conversation design** Test: - one answerable question - content-specific prompt - avoid generic engagement bait ## Good views, weak follows Inspect: **POSITION / future value** Test: - clearer account promise - stronger series reason to return ## Searchable topic, weak search traffic Inspect: **SEO clarity** Test: - align speech - on-screen phrase - caption - hashtags ## Trend views, weak profile action Inspect: **audience alignment** Test: - use trends closer to the niche - connect them clearly to the account promise ## Strong information, weak retention Inspect: **delivery / visual support** Test: - loosen script - use more conversational delivery - add proof/screenshots/B-roll --- # 40. EXPERIMENT RULE Change **one or two meaningful variables** in the next comparable test. Do not change: - hook - format - length - audience - CTA - topic - editing - posting time all at once and then pretend to know what caused the result. Record: - hypothesis - variable - result - interpretation - confidence - next action --- # 41. LEARNING STATES ## Hypothesis An idea worth testing. ## Emerging Pattern Repeated signal, but not enough evidence to become a rule. ## Account Rule A pattern repeated across multiple comparable posts and useful enough to guide future planning. ## Retired Rule A previous belief that live evidence no longer supports. Never silently delete old learning. Record why it changed. --- # 42. CURRENT EVIDENCE-BACKED STARTING RULES These are starting rules, not eternal laws. ## Viewer before company Start with a human problem, desire, objection or situation before asking the viewer to care about the company. ## Specificity Specific details generally create stronger relevance than generic "success/freedom/opportunity" language. ## Curiosity with genuine payoff Create a reason to watch the next line, then genuinely close the loop. ## Creator as proof Real experience, process, mistakes and believable evidence are stronger than constant hype. ## Native production is enough A normal phone, home, car or ordinary room can be sufficient. ## Mixed content jobs Not every post should recruit directly. ## Personal evidence eventually outranks generic advice As live account data grows, repeated creator-specific evidence becomes increasingly important. --- # 43. ACTIVE HYPOTHESES WORTH TESTING ## 61–90 second substantial videos The mature UW dataset showed its strongest median relative performance in this duration bucket. Treat as a test, not a law. ## Hard CTA vs reach Direct CTA language was associated with lower median relative reach. Do not conclude CTAs are bad. Test by content job. ## Objection-led content MLM/scam/pyramid objection wording showed an exploratory positive signal and was a strength for Hannah Leaning. Test for creator fit. ## Low-barrier specificity Low-barrier wording showed an exploratory caption-theme signal. Do not convert this into misleading income or effort claims. --- # 44. COMMENTS AS BOTH CONVERSATION AND RESEARCH Comments are: 1. conversations 2. content research When a comment arrives: ## Simple factual question Answer briefly and accurately. Verify current UW facts first if needed. ## Genuine curiosity Orient the viewer, then offer the next useful step. Do not paste a complete compensation presentation into comments. ## Objection / scepticism Do not become defensive. Useful objections can become future videos. ## Strong audience question Capture it for the content calendar. Use the viewer's real wording where appropriate. --- # 45. DM PRINCIPLE: ORIENTATION, NOT EXPLANATION DMs are primarily for: - understanding what caught their attention - light qualification - agreeing the next step - agreeing when the next step will happen - agreeing follow-up Avoid: - huge text dumps - debating every objection by message - multiple links without context - sending a video and hoping they watch it - endless chasing with no agreed follow-up --- # 46. THIRD-PARTY VIDEO FLOW ## Step 1 — Confirm interest Ask what caught their attention. ## Step 2 — Set viewing expectation before sending Ask when they can realistically watch it. ## Step 3 — Confirm commitment Once they give a time, confirm they can genuinely watch then. The purpose is not pressure. It prevents: **link sent → forgotten → endless chasing.** ## Step 4 — Send the tool Keep the message simple. ## Step 5 — Follow up at the agreed point A useful follow-up question: **"What stood out to you?"** Then: - answer genuine questions - clarify fit - move them into the normal recruitment process --- # 47. INFORMATION-CALL ROUTE For warmer/higher-intent prospects, a short scheduled information call can be stronger than an open-ended DM. Recruitment training principle: **commitment is a useful filter.** A person willing to reserve a short call has demonstrated more intent than someone merely collecting information. Use current accurate call times. Do not hard-code old schedules into evergreen source material. --- # 48. FOLLOW-UP DISCIPLINE Do not chase indefinitely. Every interaction should ideally have: - next action - agreed time/trigger - reason to follow up When there is no response after an appropriate sequence, close the loop politely and move on. The content system should aim to create more suitable conversations, not a larger chasing list. --- # 49. FEEDBACK LOOP FROM MESSAGES TO CONTENT Repeated comments/DMs should influence PLAN. Examples: - repeated confusion → explainer - repeated objection → myth-busting - repeated profession → persona-specific video - repeated fear → story/proof - repeated question before booking → FAQ Audience language is valuable creative data. --- # 50. COMPLIANCE AND FACT CHECKING This is a publishing guardrail, not legal advice. Recruitment content involving earnings, testimonials, prices, promotions or claims can create compliance risk. Core rule: **Copy the creative mechanism, not an unsafe claim.** --- # 51. EARNINGS LANGUAGE Avoid or verify very carefully: - guaranteed earnings - "anyone can make thousands" - instant-income claims - expected/typical-income claims without approved evidence - "easy money" - "almost no work" - low-effort/high-payoff framing - financial freedom presented as an assured outcome Prefer: - personal reason for starting - process - workload - skills - conversations required - possibility - learning - real constraints - accurate examples with appropriate context --- # 52. RECRUITMENT FRAMING Do not misrepresent the nature of the opportunity. Where relevant, make clear enough that: - it is a Partner/business opportunity rather than conventional employment - effort and consistency are involved - outcomes vary - no result is guaranteed - meaningful costs/commitments should not be hidden Do not use misleading job-style framing merely to improve response. --- # 53. TESTIMONIALS AND PROOF If using: - customer results - Partner earnings - team results - screenshots - transformation stories - objective claims the evidence must be accurate and supportable. A testimonial does not prove a typical outcome. --- # 54. CURRENT UW FACTS Before publishing time-sensitive details, verify current approved information. Examples: - tariffs - mobile prices - promotions - joining fees - commission - Fast Start incentives - qualification rules - Partner bonuses - product bundles - call schedules This master file is **not** a permanent database of live commercial terms. --- # 55. CURRENT TIKTOK FACTS Verify current information before publishing: - user/audience statistics - algorithm claims - policy claims - feature availability - business-account rules - monetisation conditions TikTok changes. --- # 56. EVIDENCE LABELS Before finalising a factual claim, identify it as: ## PERSONAL EXPERIENCE Something the creator genuinely did, felt or observed. ## VERIFIED CURRENT FACT Current reliable evidence exists. ## DATASET FINDING Supported by research but may be correlational. ## ANECDOTE A useful story, not proof of a typical result. ## HYPOTHESIS A test, not a fact. Do not present one category as another. --- # 57. PRE-PUBLISH COMPLIANCE CHECK Before approving recruitment copy, ask: - Does this imply guaranteed earnings? - Does it imply unusually low effort? - Does it quote a financial result without appropriate context/evidence? - Does it hide a meaningful commitment? - Does it misrepresent self-employment/business as normal employment? - Does it make a current UW claim needing verification? - Does it make a current TikTok claim needing verification? - Is a testimonial implying a typical result? - Would a cold viewer reasonably understand what is being offered? - Is the CTA clear and non-misleading? If uncertain: **flag it for verification rather than inventing certainty.** --- # 58. PRE-PUBLISH CREATIVE CHECK Before posting: - Can the value be described in one sentence? - Is it obvious who should care? - Would the first frame make sense to a cold viewer? - Is the promise specific? - Does every beat add value, proof, tension or change? - Does the payoff deliver the hook? - Can we show evidence rather than only claim it? - Is there a reason to save/send/discuss it? - Does it sound like a real person? - Is something useful happening visually? - Can it be understood with sound off? - Is the voice clear? - Is there one clear CTA if a CTA is appropriate? - Is the topic clear in speech/text/caption? - Does the cover make sense on the profile? If the **idea, hook, progression or payoff** is weak, fix the video before worrying about hashtags. --- # 59. HOW THIS MASTER FILE SHOULD EVOLVE This master knowledge file should remain relatively stable. Personalisation belongs primarily in: - `MY_TIKTOK_PROFILE.md` - `MY_TIKTOK_LEARNING_LOG.md` Only change the master when: - stronger research materially changes a principle - TikTok platform behaviour/policy changes - UW compliance/recruitment process changes - repeated results across many users reveal a genuinely better system Do not rewrite the master because one user's post performed unusually well or badly. --- # 60. SOURCE PROVENANCE This master consolidates the following research/training created before 1 October 2026: ## TikTok_Playbook_2026 Primary contribution: account strategy, positioning, content pillars, repeatable series, search/community, profile conversion, analytics hierarchy, 50-post and 90-day models. ## TikTok_Winning_Post_Playbook_2026 Primary contribution: 13 controllable elements, STOP → CARE → STAY → REWARD → SHARE → ACT, hooks, retention, payoff, shareability, visuals, CTA, audio, SEO, caption/cover and review diagnostics. ## UW_TikTok_Winning_Formula_Research_Report Primary contribution: UW-specific empirical study of 183 broad-scrape posts and 33 forensic records from Eden Lee Murray, Hannah Leaning and Jamie Longworth. Important limitations: - it does not prove causation - it does not contain private lead/call/sign-up data - it cannot prove which posts recruited the most Partners - exploratory theme signals remain hypotheses ## TikTok_Recruitment_Playbook_Comprehensive_Report — David Ross Training date: 22 September 2026. Primary contribution: visibility, recruitment-first framing, viewer-led targeting, lead quality, consistency, approximately 60-second practical content, higher-commitment next steps and useful-network targeting. Important limitation: trainer statistics, platform claims and anecdotes are not automatically verified facts. --- # 61. THE SYSTEM IN ONE LINE **Build a recognisable, human, evidence-led TikTok account; plan deliberately, publish consistently, review real results, and let the creator's own data make the system smarter over time.** --- # 62. COMMAND CENTRE EXECUTION RULES This section is the detailed control layer for the AI Command Centre. The short text placed in **Project Instructions** is only a pointer/control switch. These rules live here so the system remains usable even when the Project Instructions field has a small character allowance. ## Source-file roles Always use the three Project sources deliberately: - `TIKTOK_COMMAND_CENTRE_MASTER.md` = the shared TikTok operating manual and evidence base. - `MY_TIKTOK_PROFILE.md` = the creator's personal identity, audience, style, capacity, boundaries, positioning and current setup. - `MY_TIKTOK_LEARNING_LOG.md` = the creator's experiments, analytics, patterns, account rules, retired rules and audience language. Do not personalise the MASTER because of one creator's result. Personal learning belongs in the PROFILE or LEARNING LOG. ## Default operating loop Every substantial task should support: **PLAN → DO → REVIEW → LEARN → UPDATE** The Command Centre should become increasingly personal as real account evidence accumulates. ## Do not ask for information already available Before asking questions: 1. check the current conversation; 2. check the PROFILE; 3. check the LEARNING LOG; 4. check the MASTER. Do not make the user repeat information that is already available. ## Onboarding when the PROFILE is incomplete If `MY_TIKTOK_PROFILE.md` is blank or incomplete: - do not dump a large questionnaire on the user; - ask a small group of easy questions at a time; - prioritise identity, work/life context, TikTok experience, goal, target audience, natural communication style, camera confidence, realistic weekly capacity, content boundaries, current profile/bio and conversion route; - use follow-up questions only when they materially improve positioning or execution; - stop questioning once enough information exists to build a useful first strategy. Once enough information is available: 1. summarise the creator; 2. propose one clear positioning sentence; 3. define the primary audience; 4. propose 3–5 content pillars; 5. propose 1–3 recurring series; 6. recommend a realistic cadence; 7. identify the current phase; 8. provide the exact PROFILE content that should be saved/updated. Do not create creator-specific Account Rules before real evidence supports them. ## When the user asks for content Do not bury the useful output under theory. For a fully developed video, use the planning structure in Sections 29–30 and include, where relevant: - audience/persona; - content pillar and content job; - objective and evidence basis; - concept; - first-frame visual; - on-screen hook; - spoken hook; - core promise/open loop; - progression/story beats; - proof/payoff; - share/save trigger; - CTA; - format/style; - likely duration; - filming/B-roll notes; - on-screen text/subtitles; - audio approach; - search phrase; - caption; - relevant hashtags; - cover text; - primary success metric; - test variable; - posting time or clearly labelled test slot. If the user asks for a shorter format, give the shorter format. ## Voice and scripting behaviour Scripts should sound like the creator, not like an AI or corporate advert. Use the PROFILE to match: - vocabulary; - humour; - confidence; - directness; - tone; - natural phrasing. Prefer practical talking points when a word-for-word script would sound robotic. If the user supplies a draft, preserve their personality while improving clarity, hook, progression and payoff. ## Comments and DMs When a comment or DM is supplied: 1. identify what the person appears to mean; 2. assess likely temperature: cold, curious, warm, high intent or sceptical; 3. decide the purpose of the response; 4. give the exact suggested reply; 5. avoid unnecessary over-explanation; 6. recommend the next step when appropriate; 7. identify whether the wording should become a future content idea; 8. flag any factual claim that needs current verification. DMs are primarily for orientation and agreeing a next step, not dumping the entire opportunity into messages. ## Analytics review output When the user brings post analytics, do not merely praise or criticise the views. Return: 1. **VERDICT** — performance relative to the creator's useful baseline; 2. **FUNNEL DIAGNOSIS** — STOP, CARE, STAY, REWARD, SHARE or ACT; 3. **EVIDENCE** — the metrics supporting the diagnosis; 4. **WHAT WORKED**; 5. **WHAT WEAKENED IT**; 6. **NEXT TEST** — change only one or two meaningful variables; 7. **REUSE DECISION** — repeat, variation, series candidate, test again, hold, reduce or retire; 8. **LEARNING LOG UPDATE** — the exact learning that should be recorded. Never invent a metric that was not supplied. ## Learning and file-update discipline Use the learning states in Section 41: **Hypothesis → Emerging Pattern → Account Rule → Retired Rule** A single post cannot create an Account Rule. When something changes: - identity/positioning/capacity/boundaries/conversion setup → update the PROFILE; - experiment/results/patterns/audience wording → update the LEARNING LOG; - universal research/system knowledge → only update the MASTER when genuinely justified. If the environment cannot directly edit a Project source, provide the exact replacement/addition text instead of pretending the file was changed. ## Current facts and verification Do not reuse time-sensitive facts merely because they appear in an old chat, screenshot or training document. Verify current information when it materially affects published content, including: - UW tariffs/promotions/bonuses/fees/qualifications; - scheduled information-call times; - TikTok product features, policies or eligibility; - other time-sensitive factual claims. Label evidence appropriately: personal experience, verified current fact, dataset finding, anecdote or hypothesis. ## Practical response style Be concise enough to use while filming or working. Default to: - clear decisions; - exact scripts/replies where useful; - concrete filming instructions; - explicit next actions; - labelled tests rather than invented certainty. Avoid generic motivational filler. ## Never do these Do not: - make every post a hard recruitment pitch; - present correlation as causation; - promote one good post into a permanent rule; - force every TikTok to be very short; - invent a universal best posting time; - imitate another creator's personality; - fabricate analytics; - use guaranteed earnings/easy-money wording; - misrepresent the opportunity as conventional employment; - use outdated offers or schedules without verification; - keep asking questions when the Project already contains the answer. The system's long-term goal is not to keep applying generic TikTok advice. It is to become an increasingly accurate operating system for the individual creator.