Research anyone can copy. Social outliers and AI answers.

Tutorials are 64% of outliers and average 37.5x, and utility-mechanism videos (50% of posts) average 41.5x — the winning formula is a specific, provable hack delivered fast. Bold-claim hooks (56% of posts, mean 40.7x) beat questions and direct address for volume, but curiosity_gap hooks like @growingplantzz's papaya-tree question post the single highest means (58.7x). Posts with no CTA outperform (mean 55.7x) every CTA type including link_in_bio and comment asks, so drop the sell this week. Carrots are the standout topic (mean 209.3x from @growingplantzz and @pharmunique_), and contrarian myth-busts on houseplant/foraging care (@shoptansy spider plant, @naturereliance poison ivy) reliably beat baseline even if their mean is lower than utility. Keep videos speech-led with burned captions and light music; most winners run 15-30s tutorials, though longer contrarian/curiosity videos (60-170s) also hit when the hook promises a payoff.

The top two outliers this window are both extended, speech-only talking-heads: @toby_joshua's 201s 'behaviors to avoid' syllabus-week list (36x, forbidden/warning hook, utility) and @nids.secret's 124s flirting storytime (21x, bold-claim hook, relatable pain). Trend-audio hooks are the most common technique (50% of posts) but average only 11.8x, well below the forbidden-warning and bold-claim hooks, so riding a sound alone isn't enough. Posts with a follow CTA averaged 28.6x versus 11.8x for none, and Apple Notes shows up in half the posts as a background prop. This week, favor a direct, slightly taboo or authority-driven hook, talk for 90s+ into the camera, skip heavy editing, and close by asking for a follow.

This is a thin sample (n=2), so treat conclusions as directional. The one clear win, @tinassee's POV skit about jail rules, hit 8x by using bold-claim hooks plus niche insider knowledge wrapped in relatable storytelling rather than direct advice-giving. @walkingsunlady's talking-head sobriety confessional only hit 0.9x despite also using a bold-claim hook, suggesting the hook technique alone isn't enough without novelty or a scene/character format. If you're posting this week, lean toward dramatized, character-driven scenarios that reveal 'insider' logistics of recovery/institutional life over generic direct-to-camera advice. Both videos used speech-only audio and no CTA, so that part of the formula seems stable across the niche.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

With just one analyzed outlier, the signal is thin but directionally clear: a big-dollar claim tied to a niche affiliate tool (stay22) and a link-in-bio CTA outperformed @girlonazebra6's normal baseline by 2.1x. The winning post is 87.1s, 100% speech, burned captions, and stays in talking-head advice format with no music. Treat this as a hypothesis to test this week, not a proven playbook, since sample size is one video. Lean into specific dollar figures and named tools rather than vague 'make money online' framing, and keep captions burned in since that's the only pattern observed.

With only 3 outliers this window, the standout pattern is extreme-age authority paired with a counterintuitive, list-style claim (@oldsoulwisdom0's 104-year-old 'cut these 5 things' post hit 213x). Talking-head format and bold_claim hooks outperform tutorial/demo on mean multiple, but the demo from @vocic_lift shows product-fit content still reliably clears 5-6x on caregiving pain points like fall recovery. All three outliers used no CTA, so don't force one. Speech-driven, caption-burned talking-head videos in the 20-30s range (or longer if authority-driven like the 180s video) are the safer bet this week; pair a strong bold or novelty hook with a real caregiving pain point.

Sample is thin (n=3) and none of the winners are AI art or character posts, so treat this as a signal on format/hook mechanics to port into the niche, not proof of topic fit. The single highest performer, @ulovelailahakim's storytime, shows a confident story_open hook plus a relationship/boundaries topic pulling 21x with zero CTA. @sebintel's bold_claim listicle on Claude MCPs (14x) proves a quantified-FOMO framing works for AI tool utility content, which is the closest analog to AI art tutorials. @kaidenkilpatrick_'s wordless POV vlog (8.5x) is the weakest of the three, suggesting low-speech, caption-free vlogging underperforms story and listicle formats here. This week: borrow the story_open and bold_claim hook shapes, apply them to AI art/character topics, and lean speech-driven video over silent POV.

Talking-head is the workhorse format (63% of outliers, mean 20.5x) and bold_claim hooks show up in 75% of winners at 20.1x mean. The single biggest hit, @20vc_tok's Dyson toothbrush reaction, shows novelty/absurd-gadget framing can spike 72x even outside core AI tooling. Contrarian tool-rivalry takes (@sabrina_ramonov on Claude vs ChatGPT) and specific-number claims (@ai.with.andrew's $10k Claude Fable websites) both cleared 18x+, while plain utility explainers (@aistartupfren, @nate.b.jones) barely beat baseline at ~1x. No-CTA posts actually out-multiplied comment-CTA posts (24.1x vs 9.9x mean), suggesting the hook and topic are doing the heavy lifting, not the ask. Post this week as a 50-65s talking-head, 100% burned captions, opinionated or number-driven, on a named tool people are already arguing about.

Data is thin (1 outlier, 1 analyzed video), so treat this as a single strong signal rather than a trend. The winning post used a hyper-specific, high-stakes number ($280,000 from 45 scholarships) as the hook, spoken directly to camera with burned-in captions, and asked viewers to comment. Topic cluster was scholarships/financial aid/college tips aimed at students actively searching for money-saving or time-saving tools. Until more outliers surface, the safest bet this week is to replicate the structure — big number hook, talking-head delivery, captions on, comment CTA — while testing it on adjacent study-tool topics (AI homework help, exam prep) rather than assuming it only works for scholarships.

With only 4 outliers this window, two clear plays emerge: bold-claim talking-head/tutorial content about faceless YouTube case studies (comment CTA, mean ~9.5x) and surreal anthropomorphic animal skits in local-language dialect (visual spectacle, mean ~6x but massive absolute views, 750k-1.1M). @wanneracademy's copycat-channel breakdown and @deekay_life's 'watch this' salary-tease tutorial both drove near-double-digit multiples using contrarian/curiosity hooks and a comment CTA. @luxuryanimal345's two goat/animal family skits show the visual-spectacle novelty format can post 7-figure views even without a CTA. This week, pair a proof-driven faceless-channel case study with a comment-bait question, and if you can produce photoreal anthropomorphic characters, test a culturally-specific family skit with zero CTA.

The single analyzed outlier, @wagwanitzella's concert outfit transition, hit only a 1x multiple and centers on fashion/Yeat concert content, not campus safety, personal safety apps, or student life hacks. With just one post and a 1x median multiple, there is no evidence of a working pattern for this niche this period. Treat this report as a placeholder: don't copy the format or topic, since it doesn't match the niche's audience intent. Best move this week is to post normal campus safety content (app demos, dorm safety tips, walk-home routines) and let the system gather real outliers next window. Avoid overfitting to a single off-topic data point.

All 3 outliers this window are @indiegamejoe screen-record showcases of indie games, using a bold, hyperbolic one-line claim as the hook, novelty as the driving mechanism, and music under the footage (no burned captions). Median multiple is 5.3x, with the top post hitting 9x by comparing an indie game to two well-known IPs (Metal Gear + Stray). Speech is optional (1 of 3 used voiceover) but music is universal, and every post closes with link_in_bio. If you're posting this week, lead with a wild, specific claim about the game's mechanic or scale, show pure screen-recorded gameplay around 30-37s, and skip captions.

The strongest results this window come from short (48-51s) showcase edits that open with an intense, out-of-context line of dialogue or a bold biblical-showdown quote, then cut into cinematic movie-edit pacing — @rekuro.ae hit 17-18x this way. @fathermanuelasmr shows a second lane: slower, novelty-driven behind-the-scenes reveals about Catholic/Franciscan life, where curiosity about hidden details (secret pockets, cathedral access) carries the view even at 71s. Every outlier here is speech-driven with burned captions and no on-screen product; CTA is mostly absent (3 of 4), so the hook and topic are doing all the work, not the ask. Post this week: a movie-edit showcase using a bold-claim or question hook tied to a recognizable actor/scene, or a talking-head/behind-the-scenes novelty clip about a lesser-known Catholic practice.

The clear winner this window is the sub-10-second aesthetic study montage: text-overlay aphorism, trending music, zero speech, no CTA. @studyingwithshrisha alone landed 7 of the top 16 analyzed posts using this exact formula (42x, 19x, 11x, 9x, 8x, 7x multiples), and relatable_pain mechanism plus bold_claim hooks together dominate the outlier set (65% and 69% share respectively, both outperforming the median). POV-style relatable identity flexes (@apluslisa, @freya.loren) are the second strongest lane, especially when they subvert a stereotype like 'naturally smart' or 'too pretty to study.' Keep videos short (median 10.5s), speech-free (only 25% used speech), music-driven (100%), and skip captions and CTAs almost entirely — 94% had no CTA and still won. Product-wise, Macbook/iPad/Apple Pencil b-roll keeps recurring as the visual backdrop, so lean into that aesthetic even if the product isn't the topic.

Only 2 outliers this window, so treat this as directional, not definitive. The winner, @nicksadler.io's ChatGPT image-generator listicle, hit 5.4x by piggybacking on a trending tool and promising insider utility, while @kayden.builds' step-by-step website tutorial only managed 0.6x despite covering a real pain point (freelancing/lead gen). Both used speech, music, and burned captions at 100%, and both ended on a comment CTA, so those are safe defaults. The gap is in hook and framing: bold-claim + trending-tool hijack beat generic list_promise tutorial framing by a wide margin. This week, lean into a trending AI/tool angle with a punchy bold claim rather than a straightforward how-to.

With only 3 outliers this window, the strongest signal is that utility-driven listicles naming specific songs for a specific use case (fit pics, Halloween transitions) crush single-topic showcase content, with a median multiple of 3.3x and the top post hitting 12x. @kendallxcel's trap-song fit-pic list ran just 10s with no speech and no CTA, proving fast, silent, text-led listicles can win. @its_sophia.alice's Halloween transitions video shows the same list format can stretch to 54.7s with speech and a comment CTA when tied to a seasonal moment. Post a tight, niche-specific song list this week (transitions, fit pics, or the next event on the calendar) and skip generic novelty stunts like @anxiety_doll's Chick-fil-A challenge, which only hit 1x.

With only two outliers this window, the signal is thin but directional: @organizedchaos4's 22-second bold-claim tutorial on turning a wicker hamper into a toy chest hit 9x and 60.6k views, while @kayleenkellyorganize's 3-minute raw 'I don't know what to do anymore' reaction only hit 0.4x and 19.3k views despite a strong relatable hook. Post short, specific, product-tied hacks with a bold claim opener this week rather than long vulnerable talking-head content. Both posts used speech-only audio with burned captions and no CTA, so keep that baseline and don't add music or a hard sell. Lean into visual clutter and toy storage as topics since they carry the higher multiple.

With only 3 outliers this window, the signal is thin but consistent: both winners are talking-head videos where a person speaks directly about career/workplace pain, while the one showcase-style gifting video flopped. @lynda.leads's contrarian 'treat your job like a transaction' framing pulled a 31x multiple, and @zoraidasworld's relatable coworker-drama question hook pulled 8x — both used real speech, no music-only reels, and skipped CTAs or used a soft comment-bait question. This week, lean into direct-to-camera career takes on burnout, boundaries, or corporate disillusionment rather than product/gift showcases. Keep videos in the 60-130s range, speak plainly to camera, and don't force a CTA — none outperformed comment-asks here.

With only two outliers this window, the split is stark: @melissadorman's contrarian bold-claim take on home equity beat baseline 9.2x, while @everydaytechtip's relatable-pain pet-safety tip only hit 0.8x. Both used identical production (talking-head, 100% speech, 100% burned captions, comment CTA), so the winning variable is message framing and topic, not format. Median video length across the two is 70.2s. For this week, lean into direct, myth-busting financial claims tied to core estate/wealth topics rather than tangential tech-safety content.

The two biggest wins this window are both @kellyn_mcmullan tutorials that repurpose boring enterprise tools (Planner, Outlook, To Do) into corporate-anxiety solutions, averaging 18.2x with direct_address or pov_relatable hooks and a follow CTA. @drmichaelsays' AI-vs-radiologist reaction bits pull huge raw views (458.8k on one) but convert at a much lower multiple (9x and 2x), suggesting reach without proportional lift. @bonsanemera's income-transparency talking-head hit 13x by pairing a specific-number hook with a comment CTA that sparked debate. If you're posting this week: lead with a bold, work-life-relevant claim, show a concrete tool/system fix in under 75s, and ask for a follow rather than nothing. Bold_claim hooks show up in 60% of posts, but the direct_address hook on the #1 video is the single strongest technique here (29x).

Only 4 outliers this period but a clear thread: unscripted, first-person access into normally hidden deathcare moments outperforms produced pieces. @shiloh_harding's late-night cremation vlog (10x, 551K views) and @kapp_tv's mobile funeral procession coverage (9x, 87.8K views) both win by pulling back the curtain on taboo or culturally significant events in real time. None of the top posts use a CTA beyond one comment-bait talking-head from @allpeoplesfuneralhome, and captions are unused across the board, speech carries all four videos. Lean short-to-mid length (12s-145s range, median 79.8s) and lead with a bold claim or a striking visual moment in the first seconds. Vlog and 'other/raw coverage' formats are outperforming reaction and talking-head so far.

Tutorials make up 75% of outliers and average 8.7x, far ahead of other formats — this is the format to lean into this week. Bold-claim hooks (50% of outliers, 8.4x mean) and question-based curiosity gaps (@_topspeedgolf's 'Can You See The Difference?' at 16x) are the strongest openers. Utility content solving a specific pain point (chipping, lob shots, shallowing the club) is outperforming vague swing talk. Link-in-bio CTAs paired with training-aid or app mentions are pulling mean multiples of 16.9x, well above no-CTA posts at 4.8x. Short-game and specific-flaw topics (lob shot, chipping, shallowing the club) are punching well above generic 'golf swing' content.

With only 3 outliers tracked in this window, the clearest signal is format 'other' (100% of outliers, mean 13.6x) built on either a novelty visual spectacle or a curiosity-gap bold claim. @anumjoshua's two synchronized-entrance showcase videos and @nbcbayarea's breaking-news clip both hit double-digit multiples with zero CTA and zero burned captions. Median video length is 28.7s, though the news-style long-form piece at 110s still pulled 16x, showing length isn't the limiter if the hook is strong. Music carries two-thirds of posts while speech appears in just one, suggesting visuals plus soundtrack outperform voiceover-led content in this small sample. For commencement content this week, lean into synchronized visual moments (processions, cap tosses) and test a bold-claim opening line, skip CTAs, and let footage/music do the work.

Novelty (mean 11.8x) and bold_claim hooks (mean 10.5x) are dominating, with short showcase highlight clips from @ukfootballtrials_ and long-form talking-head/tutorial pieces from @fit_4_football both breaking out despite very different lengths. Youth football and football training topics are spiking hardest (mean 15.3x each), especially when tied to a real trending name like Rio Ngumoha for @ukfootballtrials_. CTAs are mostly absent (70% of posts, mean 7.2x) yet the single comment-CTA post from @7philtennyson still hit 13x, so a soft CTA isn't hurting. Captions are near-universal (70% burned-in) and speech is present in 90% of winners, so voice-led delivery with text overlay is the baseline expectation, not a differentiator. Utility and social-proof mechanisms are underperforming (means 3x and 1.1x), so pure tips-listicles or interview testimonials need a stronger hook to compete.

The clear winner this window is the 'innocuous relatable moment suddenly cuts to intense gym/phonk montage' format, driving @doctortren1's top two posts to 67x and 15x. Short (15-18s), speech-driven, music-backed clips with pov_relatable or trend_or_sound hooks are massively outperforming longer talking-head advice content, which still works (@b.palmer_'s bone broth post at 6x) but tops out far lower. No post in this window used a CTA, and none used burned-in captions, so don't force either. If you're posting this week, prioritize a punchy relatable cold-open that hard-cuts into a high-energy training montage set to phonk or hype music.

This is a thin 2-post sample, so treat conclusions as directional, not proven. The one real STEM/higher-ed outlier from @asisipho_mbi used a POV hook about teacher travel and school holidays, no speech, a comment CTA, and hit 32x on 72.2k views. The other post, from @laur_en_order___, was off-topic (booktok/funko) and underperformed at 0.5x, showing that non-teaching content diluted with novelty/direct-address hooks does not travel here. For this week, lean into POV-style relatable-pain hooks tied to teacher/professor life moments (breaks, burnout, semester milestones) and end with a comment prompt rather than a hard CTA.

Only 3 outliers this window, but the split is stark: @travel_with_domi's wild-camping-gone-wrong reaction (67x) and @claire.livingston's specific-age morning walk vlog (21x) both lean on relatable_pain and land far above @gwynandami's fall-foliage listicle (1.3x). All three winners are 100% speech+music videos around 60s, with burned captions on 2 of 3. If you're posting this week, prioritize a real or reacted-to moment with visible stakes (a rule broken, a routine judged) over a curated recommendation list. Anchor content to a specific, comparable life detail (age, time, mileage) rather than generic advice framing.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

Both outliers this period are tutorials built on relatable_pain: @dr.maehughes targets pregnancy hip/pelvic discomfort and @heather.woelk targets hip-thrust back pain, each pairing a bold or direct-address hook with a fast, actionable fix. Music is on 100% of winners, captions and speech are split 50/50, so audio-driven demonstration matters more than voiceover. CTA is comment on both posts, showing engagement bait tied to a specific fix works. This week, pick one common pain point in your sub-niche (postpartum or gym-form) and build a short tutorial demonstrating the fix, not just talking about it.

With only 4 outliers, the sample is thin, but the top two posts (@jbryderauthor's skit, @writeoapp's demo) both used trending audio + POV framing at under 11 seconds and hit 5-8x. @author.gemmavale proves a long-form bold-claim talking-head can still land at 7x if it taps real Goodreads/BookTok drama. The one flop, @authorpaigenicole's reaction post, shows that relatable_pain alone isn't enough without a strong hook or trending sound doing the work. None of the winners used a CTA, so the pull is curiosity/relatability, not asks. Bet the week's slate on short pov_relatable or bold_claim hooks synced to trending audio, topic-anchored to booktok/writertok/authortok.

With only 3 outliers analyzed, talking-head is the dominant format (67%, mean 3.2x) and question hooks lead (67%, mean 3.2x), but the standout mechanism is utility at 5.1x mean versus transformation (2x) and relatable_pain (1.2x). All 3 videos use speech, music, and burned captions, so that combo looks like table stakes, not a differentiator. @personalbrandlaunch0's short 19.8s 'crazy math' video massively outperforms the longer tutorial and pain-point videos, suggesting brevity plus a concrete deliverable (350 video ideas) beats length. Given the small sample, treat these as directional signals for this week, not settled laws: lean into a tight, formula-based utility video with a question hook before trying longer pain-point or tutorial formats.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

Listicles are the clear format winner (60% of outliers, mean 14x), especially when they lean utility (mean 17.6x) over humor or pure relatability. The single biggest post, @kennedy_snead's Christmas gift list for one-year-olds, hit 34x by pairing a direct-address hook with a link_in_bio CTA — and CTA data backs this up (link_in_bio mean 23.7x vs none at 4.5x). Speech-led talking content (80% of posts) still beats music-driven edits here, and captions are only present on 20% of posts, so voice clarity matters more than on-screen text. This week: build a curated, age-specific listicle (gift guide, toy roundup, or reading routine) with a bold or direct-address hook and a link_in_bio CTA to product tags.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

With only one outlier in this window, the clearest signal is that fast, visual, single-concept tone tutorials with a direct-address hook and full captions convert far above baseline. @chineseteacherkiki8's video used hand gestures synced to tone contours plus emoji cards to make pronunciation instantly graspable, all in under 10 seconds. Speech-driven, no music, burned captions, and a simple follow CTA rounded out the format. Because the sample is a single post, treat this as a strong directional signal to test rather than a statistically broad trend. This week, prioritize short tone or pronunciation micro-lessons that show, not just tell, the mouth or hand mechanics.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

Both outliers this window are short POV skits from @itsrainingbpd using the relatable_pain mechanism and a pov_relatable hook, no CTA, no burned captions. The top post (574x) pairs an insomnia/overthinking hook with an unexpected slapstick visual gag on a viral sound in just 5.37s; the second (12x) leans on a slower, dialogue-free 'ironic validation' hook about childhood emotional neglect. With only 2 outliers and 0 carousels analyzed, treat this as a directional signal, not a proven playbook: the format and mechanism repeat 100% of the time, but performance spread (574x vs 12x) shows execution and sound choice matter more than topic choice alone. If you post this week, stick to short POV skits naming a specific, ultra-specific symptom (not general 'mental health') and pair it with a surprising visual payoff.

With only 4 outliers this window, the signal is thin but consistent: novelty mechanism dominates (75% of posts, 8.7x mean) over relatable_pain, and showcase content type (75%, 8.7x) beats skit. @pubgmesportsvn's handcam/gyroscope clip and @ewc_en's tournament announcement both used bold claims or visual spectacle to hit double-digit multiples, while @janickagaming's relationship skit lagged behind at 7x. All 4 videos used music, 75% had speech, none used burned captions, and median length sits at 28.6s. CTA is mostly absent (75% none, mean only 6x) but the single 'other' CTA post from @pubgmesportsvn hit 15.3x, suggesting some form of call-to-action or tag is worth testing. Lean into showcasing rare physical/technical skill or major esports spectacle this week over relationship-skit style content.

Small sample (2 outliers) but the gap is stark: @d_wayne111's '4 Insane Offline Truck Games' listicle hit 5.1x on 16.8k views using rapid multi-game cuts and a list-promise hook, while @mohanad_pc's single-game screen-record demo of Castle Busters only reached 1.2x despite 10x the raw views. Both used music-only audio with zero speech and zero burned captions, and neither used a CTA, so those aren't differentiators yet. If you're deciding what to post this week, lead with a numbered list hook over multiple truck/offline games rather than a single long screen-record session. Keep it short — median outlier length is 14.1s — and let gameplay footage carry the hook instead of narration.

The single biggest outlier this window is @jsmnzxr's dashcam near-accident clip at 17x/1.1M views, built on relatable_pain and pov_relatable hook technique — nothing else comes close. POV format overall averages 9.4x vs just 3x for other formats, and vlog content type averages 11x vs 3x for showcase. Speech-driven posts with a clear stakes-based hook (bold claim, question, pov_relatable) are beating pure visual-spectacle showcases like @mipega.online's ASMR key-fob clip and @jsmnzxr's low-performing sportbike showcase (0x). No outlier used a CTA, so the mechanism is the hook and tension, not an ask. This week: prioritize short POV footage with an unresolved-tension hook over polished bike-porn showcases.

The two winners this period both used bold_claim hooks with real speech and no CTA — one a raw political confrontation clip (@hankyung.com, 20x), the other a screen-recorded growth hack reveal (@build_in_public, 9x). Both wordless CapCut news-template showcases from @giovanniartes underperformed at 0.2x mean, despite high production value, suggesting visual_spectacle without narration or stakes isn't landing right now. Median video length is 60s, but the second-best post ran just 24s, so don't assume longer is better. This week: lead with a spoken bold claim tied to a real news moment or a marketing/SEO hack, skip silent template showcases, and keep CTAs light or absent.

The single biggest winner is @administraciondejusticia's listicle on under-the-radar, high-salary oposiciones with low competition, hitting 112x — pulling topic means for empleo publico, justicia, carrera and sueldo above 55x. Utility mechanism as a whole averages 31.3x vs 6x for relatable_pain, and bold_claim hooks average 32.2x vs under 13x for every other hook technique. Talking-head format still dominates volume (56% of posts) but listicles post the higher mean multiple (31.3x vs 6.7x), so the format itself matters less than pairing utility + bold_claim + a specific-oposicion angle. CTAs are mostly absent on winners (none-CTA posts average 27.1x), suggesting audiences respond better to pure value delivery than to asks. This week: lean into niche/insider oposicion picks, salary/competition comparisons, and short direct-address hooks naming a specific corps (Policía Nacional, justicia, etc.).

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

All 4 outliers are 100% speech, music, and burned-caption tutorials running 65-78s, so the format itself is settled — the differentiator is the hook and the 'why your plant is dying' angle. Bold-claim hooks (@plantalytics's cactus and basil videos) outperform question hooks nearly 5x (30.8x mean vs 6.4x), and utility mechanisms (26.1x) edge out contrarian (14.9x) but contrarian myth-busting still lands well when paired with a strong before/after visual like @shoptansy's spider plant post. Repotting, cactus care, and basil/herb rescue topics are the strongest performers; generic 'gardening' tips without a specific plant or failure point underperform. This week: pick one commonly-killed plant, open with a bold claim about why people fail it, show the visual proof (soil, roots, before/after), and close with a follow CTA.

Talking-head clips built on contrarian claims (bold_claim or question hook_technique) are the safe, repeatable format right now, averaging 6.2x across 7 posts like @financepod1's renting-vs-buying clip and @melrobbins' parenting/dopamine clip. But the single story-format post from @chillcastt (17x, 483.5k views) shows curiosity-gap storytelling can massively outperform advice clips when it teases an unresolved personal story. Every analyzed video used speech-driven delivery with 100% burned captions and no music required, so captions plus raw talking are table stakes, not differentiators. Topics skew toward relationships, mental health and parenting (mean 7.3-11.4x) beating economy/investing topics (mean 3.9-5.7x), suggesting personal-stakes topics outperform macro finance takes. CTAs are mostly absent (88%) and that group still out-multiplies the one with a CTA, so skip the ask this week and let curiosity or contrarianism carry the clip.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

With only 4 outliers, the signal is thin but consistent: talking-head format beats tutorial (4.8x vs 1x mean), and relatable_pain framing (mean 6.4x) crushes straight utility content (mean 1.2x). @tommyharr05's two posts about flippers losing money or going broke pulled 10x and 3x, both anchored in bold claims or provocative questions about financial risk. Straight how-to utility content, like @nextlevelpm's financial formula breakdown and @jackiethehappyinvestor's 90-day tutorial, still worked but converted at a much lower multiple. Speech-driven talking-head with burned captions (75% of videos) is the safe format bet this week; skip long-form tutorials unless you can front-load a pain hook.

Los outliers de esta ventana están dominados por relatable_pain (40%, media 9.3x) y specific_number (20%, media 13x), casi siempre en formato talking-head o 'other' con voz real, como @motorpasion con su aviso de gasolina mezclada con agua y @mecnico4502 con el modo Eco. El formato POV es el más usado (33%) pero el que peor rinde (media 4.2x), mientras que talking-head y 'other' arrasan (19.7x y 40x). Los hooks tipo bold_claim (40%, 12.4x) y pov_relatable (33%, 16.9x) son los que más aparecen y mejor funcionan; forbidden_or_warning solo tiene un caso pero es el mejor hook técnico (24.6x) en @motorpasion. Sin CTA gana con diferencia (60% de los posts, media 14.6x): la gente comparte por el contenido, no por la llamada a acción. Esta semana: menos POV genérico, más talking-head directo con un dato o advertencia concreta sobre coches, gasolina, ITV o eléctricos.

All 3 analyzed outliers open with a bold_claim hook (100%) and are 100% speech-driven with burned captions, so scripted, caption-heavy delivery is table stakes right now. The top performer, @imanidoesmarketing's talking-head at 22.2x, runs 237s and directly contradicts standard content-marketing dogma, showing long-form contrarian advice can still outperform quick hits. @build_in_public's screen-record novelty hack (8.6x) proves a fast, visually-demonstrated 'trick' format also works well within 24s. CTAs are mostly skipped (67% none, mean 15.4x) versus link_in_bio (33%, mean 7.4x), suggesting value-first delivery beats hard asks this window. Post this week: pick one sacred-cow marketing belief to challenge, or one concrete tool trick to screen-record, and script it with a bold, punchy opening line.

The winning move right now is stitching or reacting to an already-viral clip (a scary rash, a tanning fail, a texture complaint) and resolving it fast with clinical authority — see @dermdocmalik, @denverskindoc, and @dermatologysurgeon, all pulling 200x+ multiples. Relatable_pain is the single strongest mechanism (mean 274x) and pairs naturally with reaction format and zero CTA — 71% of outliers used no CTA and still averaged 87x. Short, high-stakes binary visuals (cancer vs safe, benign vs melanoma) built as static graphics with no voiceover are also punching way above their length, per @sadovskaya_doctor's 5-7s posts. Straight talking-head advice is still the most common format (38%) but is now a baseline, not a breakout — mean only 15.2x. Skip novelty/animal tangents and vague 'other' CTAs; they're dragging averages down hard.

Data this period is thin: a single analyzed outlier from @travelingwithtals, and it underperformed its own account median (0.5x), so treat conclusions as directional, not proven. The post that did get flagged leaned on a relatable-pain, POV hook ('hyper-independent eldest daughter proves she's alive') stitched to horseback-riding-in-Mongolia visuals, full speech, music, and burned captions across a 60-second vlog. Given the near-empty sample, the safest move this week is to test more variations of this persona-plus-adventure format rather than assume it's a proven winner. Do not overweight any single mechanism or topic here until more outliers accumulate.

The top performer this period is @topps's Star Wars showcase, pulling 3.3x by opening with a specific sold-price ($76,250) on a tactile rare card, no speech needed. @blackgoldsportscards's talking-head video about a 71-year set chase only hit 0.5x despite a strong bold-claim hook, suggesting curiosity_gap framing without a visual payoff underperforms specific_number claims with visible proof. Both posts use music and a bold-claim hook style, so that opening technique is table stakes, but pairing it with a concrete number and physical card in frame is what separates 3.3x from 0.5x. With just 2 posts analyzed, treat this as directional: lean into dollar-figure showcases this week, and if doing talking-head, add a visual reveal to back the claim.

The single biggest winner is @tyesenstudyfetch's AI study-session demo (166x) and @litseyyyy's 'studying is a privilege' affirmation clip (71x) — both short, bold-claim, mechanism=utility or relatable_pain posts under 20 seconds. Meanwhile the bulk of output is @nids.secret style talking-head advice on dating/attraction/college life, which posts consistently (62% of outliers) but only averages 11-20x, well below the utility and bold-claim tier. Planner/productivity 'plan with me' content from @mayawofsywellness still works but plateaus around 5-7x unless it includes a genuinely novel tool or system reveal. For this week: pair a bold, product-driven claim with a fast demo or an identity-affirming reframe of studying/planning, keep it short, and reserve the longer talking-head advice format for pure volume/follow-growth rather than viral spikes.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

Sample is razor-thin (2 outliers, median 7.5x), so treat this as directional, not proof. Both posts open with a bold_claim hook and run 100% burned captions, suggesting that format element is safe to keep no matter what. The only genuinely on-topic outlier is @elite.recruiter's talking-head LinkedIn-cringe video (96.5s, relatable_pain, cta=comment, 4.8x), which is the template to replicate this week. The higher-multiple post from @texpatpat (10.2x) is a short novelty retail announcement about Gentle Monster opening in Dallas and has nothing to do with tech job search, so don't let its multiple pull your strategy off-topic. Prioritize relatable career pain points delivered talking-head style with a strong opening claim and a comment CTA.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

POV format is beating talking-head badly (mean 60x vs 26.1x), and story_open hooks (mean 64.2x) crush bold_claim hooks (24.4x) even though bold_claim is used twice as often. The single novelty post, @u44nhd9efitfg's ride-along demo of an officer explaining registration stickers, hit 164x and is skewing the format's numbers but proves that authentic, unscripted police-perspective content has huge upside. Utility content (mean 29.9x) and specifically 'driver safety' and 'traffic stop' topics (60.3x and 84.9x mean) are the safest bets this week over generic 'know your rights' or 'legal advice' explainer topics which lag. Skip CTAs (80% of outliers used none) and lean into raw, real-footage cold opens instead of a scripted studio intro.

Talking-head advice is the dominant format (46% of outliers, mean 19.8x) and content type (64%, mean 21.7x), typically ~60s with speech, music, and burned captions on nearly every post. Relatable_pain and contrarian mechanisms outperform (mean 43x and 22.1x respectively), especially when paired with bold_claim or forbidden_or_warning hooks (40.3x and 41.5x mean) — see @petethepilot1's tray-table warning and @petethepilot1's '10 days in Europe' contrarian itinerary post. The single story post from @raimeetravel and the aesthetic showcase from @followmeaway both massively overperformed (211x and 69x) by leaning into emotional/nostalgic curiosity rather than pure utility, suggesting there's room beyond straight advice. Listicle-format utility posts with 'day-by-day plan' framing still work but pull lower multiples (mean 15.3x, and 5x/3x for @dalia_shalaby and @raimeetravel) unless stacked with strong authority framing. Comment CTAs are most common (55%) but posts with no CTA actually show the highest mean multiple (94.3x), so don't force engagement bait.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

The single strongest lever right now is the 'coolest kids AMAZON find' formula: a bold-claim text hook over a silent 6-15s demo/showcase, no CTA, no talking. @natalia.be.be alone posted 6 of the 16 outliers this way and hit up to 189x. Novelty and utility mechanisms together account for 69% of outliers and the highest mean multiples (46.1x and 23.8x), while relatable_pain and pov_relatable hooks on longer talking dinner videos from @mummysflippinhouse barely moved (1x). Tutorials are common (38% of posts) but average multiple is much lower (7.3x) than the 'other' demo/showcase format (58.6x mean), so craft/activity tutorials need a stronger novelty angle, not just instructions. Post this week: silent product-solves-a-problem demos with bold on-screen text, seasonal (Halloween/half-term) craft or days-out listicles with specific numbers/prices, and skip long spoken dinner-idea videos unless there's a strong hook twist.

The single biggest lever right now is the pov_relatable hook technique (36% of outliers, mean 23.4x) paired with a comment-bait CTA (mean 30.6x) — @kanececi's kid-packs-my-lunch vlog hit 90x and @livcartlidge_'s in-store showcase and challenge posts also cleared 1x+ on the strength of a relatable, low-production moment. Straight decluttering utility content still works as a dependable base: @thedeclutter.co's 'Day 1 September Declutter Challenge' (18x) and @erin.makehomesimple's one-year rage-declutter follow-up (6x) show that challenge framing and long-term proof beat generic tips. Talking-head is the safest bet for volume (46% of posts, mean 4.4x) but the 'other'/vlog format has the highest ceiling (mean 48.5x) when it captures an authentic, funny or novel household moment. Skip flat listicles and silent showcases — @dana.litz's five-things list scored 0.4x and @livcartlidge_'s silent Halloween clip only 1x, both well below the median 1.4x. This week: lean into a kid-led or novelty vlog moment, add a comment-driving question, and keep any pure advice content tied to a challenge or before/after proof rather than a generic list.

Computed from 0 outliers (0 analyzed from the video or slides). Median multiple ?x.

Talking-head advice dominates (67% of outliers, mean 47.3x) and utility/relatable_pain mechanisms are the biggest multipliers (71.3x and 33.7x mean respectively). Bold_claim hooks that promise a hyper-specific, tiny fix (three words, seven words, one hand gesture) massively outperform vague framing, as seen with @neilcanhelp's speech-opener videos and @theandyharrington's gesture tip. Career/workplace pain points (calling out sick, LinkedIn judgment, broken hiring systems) also spike when framed as a relatable confession plus a script fix, per @howtoconvince and @elite.recruiter. Short-to-mid length (15-90s) with heavy captions and no music dependency is the safe zone; the two long-form outliers (200s+) only worked when paired with strong authority or challenge framing. Post this week: one micro-script tip video and one recruiter-pain confession video, both bold_claim hooks, follow or save CTA.

With only 4 outliers this window, the signal is thin but consistent: relatable financial pain beats generic advice, and putting exact euro numbers in the hook drives the top multiples. @nettowise.nl shows up twice with concrete calculator-style breakdowns (€75/uur, €157.000 winst) hitting 6-8x, both using bold_claim hooks. The single POV clip from @nettowise.nl is the top performer at 8.2x despite being only 6.33 seconds and having no speech, proving a punchy on-screen number can replace narration entirely. Talking-head still dominates volume (75% of posts) but averages lower (5.4x) than the one POV entry, so short number-first opens deserve testing. Comment-bait CTAs correlate with the best average multiple (7.2x), ahead of link-in-bio and follow.