I keep going down rabbit holes trying to understand how the big platforms actually decide who gets seen: what an ad really costs you and why, which social channels are worth growing for free, how ChatGPT and Google's AI pick the sources they quote, how to reach influencers, and when to post. These are my write-ups, in plain English, most with interactive simulators you can play with.
Research · Pricing New The Psychology of Pricing and Why a High Price Feels Like Quality The habit of reading quality off a price tag is called the price-quality heuristic, and 54 pooled experiments put it at about 12% of what people think of quality. Six newer studies covering 2,842 people found a high price raises what buyers expect every time and what they think afterwards almost never. Across 65 years of lab tests, quality explains about 4% of the price gap between rival brands. How To · Live Streaming How to Get More Viewers on Twitch Twitch orders its browse page by who is already being watched, so a channel with nobody watching is never shown. Kai Cenat finished his first Twitch night with 5,000 followers carried from YouTube. Jynxzi averaged about four viewers for a year. IShowSpeed averaged ten for four years, then went from 100,000 to 1 million subscribers in two months once clips of him spread on TikTok. How To · Short Video What Makes a Good Hook for TikTok and YouTube Shorts A study that recorded 15,344 real swipes found 29% of them land in the first fifth of a short video, and only 6% land between 60% and 80% of the way through. TikTok says 63% of its highest click-through videos show the point inside 3 seconds. Sixteen hook rules, ranked by how strong the evidence behind each one is. How To · Local Search Local SEO for Small Businesses A local search returns two lists, the three map results and the blue links below them, and they are ranked by different things. Your Google profile is 32% of what decides the map results and only 7% of what decides the blue links, and the top three factors of each list have nothing in common. Explainer · Reddit What a Reddit Downvote Actually Costs You A downvote is engagement on most platforms and subtraction on Reddit. Reddit ranks a comment by the lowest score it could plausibly deserve, so 500 upvotes with 400 downvotes ranks below 3 upvotes with none. Each downvote also costs about one karma point, and the count on screen is fuzzed so the damage cannot be measured. How To · Local Search How to Rank Higher on Google Maps Google puts an AI Overview on only 15% of local searches, and the three map results still appear on 93% of them. A survey of 187 ranking factors puts your Google Business Profile at 32% of the local ranking, and the single strongest thing in it is the primary category dropdown. Explainer · Comments Will You Get Shadowbanned for Promoting a Product in Comments? A comment feels safer than a post, so people promote products in comments. 69.1% of 81,381 YouTube comments got zero likes, YouTube hides every reply behind a button while Reddit shows replies with no click, and it is the link that gets you removed, not the product name. Explainer · Reddit Should You Start Your Own Subreddit to Promote Your Product Answering with a link to a subreddit you own does not get past moderators, because they judge the account and not the link. One subreddit member is worth 0.4 to 2.1 clicks a month on whatever you link, and the swap takes 9 to 45 months to repay. How To · Advertising How to Hire UGC Creators and What to Pay Them A UGC video costs $25 on a marketplace and $150 to $500 hiring direct, and one company publishes both $57 and $175 as the going rate. What to pay, what to put in the contract, and why you should budget for 17 videos rather than one. How To · Platforms How to Grow on Threads and Get More Followers Replying to the comments on your own posts raised engagement 42% on Threads, the best of any platform. Every measured tactic ranked, plus the six popular tips that have no study behind them. Explainer · Platforms What Is Meta Threads and How Does It Compare to X Threads has 500 million monthly users, and more people open it each day than open X. A new account gets 5.9 times more views there than on X, and it sends almost no clicks to your website. How To · Audience How to Build an Email List, and Whether a WhatsApp Channel Beats One A WhatsApp Channel hides every follower from you and cannot be exported. Email is opened 20.7% of the time and the Gmail Promotions tab costs far less than people think. Seven ways to collect followers compared. How To · Platforms How to Write a Viral Tweet and Get More Followers on X X published the code that ranks the For You feed. A reply is worth ten likes, a copied link forty, and one report cancels 468. Paired with 11,000 controlled pairs of tweets and four peer reviewed studies. How To · Platforms How to Build a Social Media Presence Without Video Short video shows your post to strangers, and so do Reddit, Pinterest, LinkedIn documents and Facebook Groups. Ten channels compared on stranger reach, hours per post and whether you need a camera. How To · Apps How to Design a High Converting App Paywall Every element of a mobile app paywall, ranked by the evidence behind it. Built on 115,000 apps of RevenueCat data, 40 million paywall opens and named redesigns that published their numbers. How To · Platforms How to Grow a Facebook Group and Stop the Flooding Facebook lets a group ask three membership questions, hold ten rules and cap one member at one post a day. The tactics that grow a group, and every tool that stops it flooding. Analysis · Platforms What Is Bluesky, Who Uses It and How to Grow a Following Bluesky has about 46 million registered accounts but only 10.4 million monthly active users, and its active base has more than halved since late 2024. Who is on it, and the one growth tactic with real evidence. Analysis · Platforms What Is Notion Used For and Who Actually Uses It Notion passed 100 million users but reports only about 4 million paying customers, so the large majority pay nothing. The largest group a stranger can reach is template buyers. Interactive Tool · Video How to Write a Video Script That Gets Views Veritasium describes his own method as V = (M + (Q to E)) x (A + B), and all three of its parts were tested separately in real experiments. A free planner turns your topic into a timed beat sheet. Interactive Tool · Short Form Video The Best Time to Post TikTok, Reels and YouTube Shorts Across 24.4 million Instagram posts the best hour for a Reel is midnight, at 25,159 average views against 15,928 at 6 in the evening. A free heat map scores all 168 hours of your week. Interactive Tool · Forums How to Write a Reddit Title That Gets Upvotes Stanford posted the same 85 images twice and the good titles collected about three times the upvotes. A free tool scores your title against every element measured on 86,035 matched pairs. Interactive Tool · Social The Best Time to Post on Social Media Night is when everyone scrolls and the wrong time to post, because the 9pm feed carries about twice as many competing posts as the midday one. A free heat map scores all 168 hours of your week. Email · Delivery How to Stop Your Emails Going to Spam About one in six legitimate emails never reaches an inbox. Gmail and Yahoo made SPF, DKIM and DMARC compulsory in 2024, and spam trigger word lists turn out to be mostly a myth. Outreach · Email How to Write a Cold Email That Gets Replies Only 8.5% of cold outreach emails ever get a reply, and open rates stopped being a real measurement in 2021. Across 126 randomised trials, a second email beat writing a better first one. Ad Creative · Billboards How Much Text Should an Ad Have Eye trackers show a billboard gets 0.379 of a second of attention, which is long enough for three to five words. But across 10,372 A/B tests the shorter wording won only 43.9% of the time. Interactive Tool · Copywriting How to Write a Headline That Gets Clicked I measured 18 headline rules across 10,592 A/B tests where the picture was held identical. Ending with a question mark costs 13.5% of clicks, and addressing the reader as "you" does nothing at all. Organic Growth · YouTube What Makes a Good YouTube Thumbnail? Every Rule, Checked Eye tracking found the thumbnail drove 26 decisions, 5 to pick a video and 21 to reject one. It is a rejection filter, and most of the famous thumbnail rules were never measured by anybody. Ad Creative · Design How to Make a Scroll Stopping Banner: Every Element, Measured An eye tracker found real banners were looked at 13.0% of the time and a blank grey box 8.8%, a gap small enough to be chance. What actually moves the number is where you put the banner. Organic Growth · YouTube Quality vs Quantity on YouTube: Which Grows a Channel Faster? Across 10.2 million channels, the ones uploading 12 or more times a month grew views 4.1 times faster. Upload frequency is not a ranking factor, so volume buys more chances, not algorithm favour. Interactive Tool · Marketing Tactics What Is Stealth Marketing? The 6 Types and How to Use Them There are only six types: product placement, logo seeding, content marketing, astroturfing, fake reviews and undisclosed sponsorship. The three fast ones collapse the moment anyone checks. Interactive Tool · Referral Programs Can You Run Ads to Recruit Referrers? What an Ad-Funded Referral Program Really Costs Advertise the payout instead of the product and one recruited referrer costs $31. About 90% of people who join a referral program never refer anybody, so one who does costs $315. Interactive Tool · Ad Targeting What Is Retargeting and Does It Actually Work? How It Works, What It Costs and Its Hard Limits It does not work, because it costs more than it brings in. Meta reports $22 a customer, but holdout tests put the real price at $80, against $42 to $48 for a total stranger on the same platform. Interactive Tool · Ad Economics How Low Can Customer Acquisition Cost Go? Why $35 Is the Median, Not the Floor The famous $35 per customer needs an 8% payer rate, and at the 2% to 5% a freemium app really converts the identical ads cost $120. Includes a calculator that turns your funnel into a minimum price. Interactive Tool · Product Launches How to Launch on Product Hunt: Requirements, Real Upvote Numbers and Old Apps Being featured is the real gate, and an unfeatured launch is not ranked at all. I measured five real 2026 leaderboard days: the top spot cost 356 to 643 upvotes, not the 800 to 1,500 you will read. Interactive Tool · Forums Reddit Alternatives: 14 Forums Like Reddit, Ranked by Reach Reddit is only the third biggest forum, behind WhatsApp and Facebook Groups, and it should still be first. Fourteen alternatives compared on reach, anonymity and effort to first traction. TikTok · Paid Ads Are TikTok Spark Ads Worth It? What They Cost, Whether They Convert, and How They Compare to Meta and Google Spark Ads boost a real organic post, so the money leaves something behind. But TikTok's headline 134% and 157% claims count only the campaigns that worked, and the honest numbers are smaller. TikTok · Creative Will a Viral TikTok Spark Ad Sell Your Product, or Will People Only Remember the Trend? The vampire effect is a measured drop in brand recall when something in the ad is more interesting than the brand. A Spark Ad has four separate vampires and three of them are structural. Google · Display How to Create a Successful Google Display Ad (and Why the Best Ones No Longer Look Like Ads) Display clicks cost a quarter of a Search click and convert five times worse, so the cheap click is not a cheap customer. What 432 randomised experiments measured about banner design. Organic Growth · YouTube Can Faceless YouTube Videos Go Viral? Yes, and the highest a faceless channel has reached is 80.6 million subscribers. The niches where the format works, the three where it fails outright, and why most of them still show hands. Organic Growth · YouTube How Much Am I Losing If I Don't Put a Face in a YouTube Explainer Video? The itemised bill for staying off camera, drawn from roughly 100 controlled studies. A face takes 41% of viewing time without improving recall, and the thumbnail rule is far weaker than you are told. Organic Growth · YouTube Is YouTube Worth Growing? The slowest and most expensive channel to build, and the only one whose posts keep earning for years. What a video really costs in hours, how long the ramp takes, and why the compounding library is worth it. Organic Growth · TikTok Is TikTok Worth Growing? The only place a 100-follower account can hit a million views, because the feed judges the video rather than the account. The price is a 48-hour shelf life and near-daily posting. Organic Growth · Instagram Are Instagram Reels Worth Growing? The best pure discovery engine of the Meta surfaces, with about 55% of views from non-followers, quietly crippled by a link problem that keeps off-platform conversion under one percent. Organic Growth · Facebook Is a Facebook Page Worth Growing? Organic reach fell from roughly 16% of followers in 2012 to about 1.4% today, and it was throttled deliberately. Whether a Page is worth keeping, and the one Facebook format that still reaches strangers. Organic Growth · Facebook Are Facebook Groups Worth Growing? The quiet winner of the whole comparison: 40 to 50% reach to members against a Page's 1.4%, at the lowest production cost of any channel, paid for in daily community management. Organic Growth · LinkedIn Is LinkedIn Worth Growing? The one major platform where organic reach has not collapsed, with personal profiles still reaching 20 to 40% of connections. Superb for B2B, poor for a broad consumer brand, and useless from a company page. Organic Growth · X (Twitter) Is X (Twitter) Worth Growing? Data on 18.8 million posts found free accounts get under 100 median impressions per post against about 600 for Premium. Which paid tier actually buys reach, the 43-minute half-life, and the link penalty. Organic Growth · Short-Form Video Do Reels and Shorts Actually Grow You? The reach to strangers is real, and the conversion is where it falls apart. Roughly 0.1 to 2% of views become followers, so a million views is a few hundred to a few thousand. Includes a funnel simulator. Organic Growth · Tactics Can You Grow by Replying to Other People? On X and LinkedIn you can grow by replying rather than only by posting, which needs no audience to begin with. Why the video platforms effectively lack this lever, and how much it changes the calculus. Interactive Tool · Reddit The Best Time to Post on Reddit (US and UK Overlap Analysis) About half of Reddit's traffic is American and a US reader is worth roughly four times an international one, so which hour maximises the US without losing Britain? Includes a free overlap clock. Interactive Tool · Facebook The Best Time to Post on Facebook (US and UK Overlap Analysis) The largest published timing study covered two billion engagements. Why a North American user is worth roughly three times a European one, and why you should never copy your Reddit schedule. Influencers How to Contact Influencers Directly (and Skip the Expensive Platforms) How to reach a creator and get a reply without paying a platform: where to find a YouTuber's business email, why TikTok gives you one message, and a free copy-paste outreach message builder. Interactive Tool The Best Time to Contact Influencers (Free Time-Zone Tool) A free tool that converts the best creator-outreach send windows into your own local time, with a live "is now a good time?" readout. Also the best days to send and how to follow up. Organic Growth Which Social Media Platform Is Actually Worth Growing? (Organic Reach, Effort and Payoff Compared) Organic reach from zero, hours per post, time to an audience and content shelf life, compared across nine channels. Includes a simulator that ranks the platforms for your own situation. AI SEO · YouTube YouTube for AI SEO: The Single Strongest Lever for Getting Cited Why YouTube dominates AI citations, why long-form videos beat Shorts by a mile, why a relevant niche creator beats a bigger off-topic one, and how to earn the mentions that count. AI SEO · Creators Should You Pay a YouTuber for AI Visibility? Hiring a creator can work, if you match the language to your market, know that views matter less than breadth, and understand what the AI can actually see about the channel. AI SEO · Quora Does Quora Help You Rank in ChatGPT and Google AI? Quora is one of the most-cited sources in Google's AI answers, and unlike Wikipedia you are allowed to take part, if the answer is genuinely useful. AI SEO · LinkedIn Do LinkedIn Articles Help AI SEO? What Actually Gets Cited LinkedIn surged to around the second most-cited domain in AI answers, but it is a B2B lever, and an Article gets LinkedIn cited, not your own site. AI SEO · Medium Does a Medium Article Help AI SEO? (Not the Hack You Think) Why a self-written Medium article that links to your site is not the hack it looks like: the links are nofollow, the citation goes to medium.com, and Medium is a rounding error in AI citations. AI SEO · Reddit How Reddit Powers AI Answers (and the “Reddit Is Locked” Myth) Reddit is one of the most-cited sources in ChatGPT and Google's AI, so how do the assistants read it when its API is paywalled? Through licensing deals, not scraping. AI SEO · Wikipedia Should You Create a Wikipedia Page for Your Product? A real Wikipedia page is the best AI asset for ChatGPT, but writing your own is discouraged and usually deleted. What the notability bar is, and why Trustpilot and app-store sources will not clear it. AI SEO · Freshness Does Content Freshness Matter for AI SEO? (And Can You Fake It?) How AI and search engines actually judge freshness, whether it matters for your query, and why changing the date, or spinning up a new URL each year, does not work. AI SEO · Google AI AI Overview vs AI Mode vs Gemini: What's the Difference? Google's AI Overview, AI Mode and the Gemini app all run on Gemini but behave differently. What separates the three, and which ones actually matter for AI SEO. AI SEO · Gaming How People Game AI Answers: 13-Word Poisoning and the Black-Hat Playbook The 13-word Reddit poisoning, hidden instruction injection, content farms and paid seeding: how AI answers get gamed, and why it is a trap the platforms are hunting. AI SEO · Apps How to Get Your App Recommended by ChatGPT and Google AI People ask an assistant for an app before they open a store, so it reads reviews, roundups, YouTube and Reddit, not your listing. How to get recommended. Influencers How to Find Influencers to Promote Your Brand (Without Overpaying) What each tier of creator really costs, why nano and micro creators beat the big names on cost per customer, and how to spot fake followers before you pay. Two interactive simulators. AI Search AI SEO: How to Get Your Website and App Into ChatGPT and Google AI Answers How ChatGPT and Google's AI answers pick their sources, why YouTube and brand mentions beat backlinks, and what to change on your own pages. Two interactive simulators. Memes Do Meme Ads Actually Work? (What the Data Says About Memes on Facebook and Reddit) Why memes win attention and reach but fade toward the sale, why engagement does not predict sales, and how the joke can quietly eat your brand. Four interactive simulators. Google How Google Ads Really Work (and Why the Priciest Clicks Can Be the Cheapest Customers) Why quality lowers your cost per click, and how the most expensive click on the internet can still be the cheapest customer. Four interactive simulators. Reddit How Reddit Ads Really Work (and Why They Are a Different Game to Meta) Why you buy people and not subreddits, how geography decides who sees your ad, why upvotes and downvotes move your cost, and how it all differs from Meta. Four interactive simulators. Meta How Meta Ad Auctions Really Work (and Why They Work That Way) Why Meta sells attention not space, why you bid on outcomes, what the estimated action rate is, and why a smaller bid often wins. Four interactive simulators.
Vlix