CHAPTER FIFTEEN

How Discovery Works Now

Almost everything you’ve heard about “the algorithm” is wrong. Underneath all the noise, discovery runs on a small handful of durable principles — and the machine is only ever trying to do exactly what you do behind the bar.

Welcome to Part Three. We’ve settled where to be — the platforms, and the patch of ground you own. Now we turn to the thing that actually feeds all of it: the content itself. And before we talk about how to make it, we have to understand how it gets seen, because everything you make will live or die by a set of invisible machines we call algorithms — and almost everything you have ever heard about them is either wrong, out of date, or being sold to you by someone who profits from your confusion. That confusion is a whole little industry, and it exists precisely because the truth is too simple to sell. Once you understand the handful of things that are actually true, the mystery evaporates — and with it, most of the anxiety that keeps good bartenders from ever posting a single thing at all.

Here is the good news, and it’s a genuine relief: you do not need to “beat the algorithm.” You don’t need the secret hacks, the magic posting times, the tricks some twenty-two-year-old is packaging into a ninety-nine-dollar course. Underneath the platform-by-platform noise, discovery still rests on a small set of durable principles: clarity, relevance, retention, response, and consistency. Learn those, ignore the rest, and you will already be ahead of very nearly everyone else fumbling around out there.

So this chapter does one thing: it separates the two. The durable principles that hold true everywhere and endure, on one side; the platform-specific mechanics that change every few months and are not worth one minute of your chasing, on the other. And I’ll give away the punchline right now, because it’s the soul of the whole thing: the algorithm is trying to do exactly what you do behind the bar every night — read the room and give each person something they’ll love. Understand that one sentence, and all the rest is just detail.

15.1 The Big Shift: From Who You Know to What You’ll Love

Start with the most important change, because everything else flows from it. For years, social media ran on what people call the social graph: you followed accounts, and you saw the stuff those accounts posted. Clean and simple. That world is essentially gone. Today it runs on the interest graph: the platform shows each person whatever it predicts that person will enjoy, whether or not they follow you. Read that again, because it’s the whole ballgame. Your content is no longer shown to your followers first. It’s shown to whoever the machine thinks will want it. The technical name for the change is the move from the social graph to the interest graph, but you can just think of it plainly: the old internet asked “who do you know?” and the new one asks “what do you like?” Nobody has to be your follower to see you anymore. They only have to be the kind of person who’d love the thing you made.

Here’s what that means down on the ground, and it cuts both ways hard. It is why a bartender with two hundred followers can post one clip and wake up to two hundred thousand views, and why an account with fifty thousand followers can post and reach almost nobody. Follower count simply is not your reach anymore; it’s barely even a factor in it. What the platform does now is treat your followers as a test audience — a small starting group it shows your new post to first, to gauge whether the thing is any good — rather than as a guaranteed set of eyeballs you’ve earned the right to reach. For a honky tonk, this quiet mechanical fact is enormous. It means the video you shoot of your Saturday-night crowd two-stepping can reach a country-music lover three states away who has never once heard your bar’s name — not because they follow you, but because the machine clocked that they love exactly this. Your reach is no longer capped by how long you’ve been at it; it’s set, post by post, by how good the thing is.

And this is wonderful news for someone just starting, once it sinks in. It means you do not have to grind for years building a following before you’re allowed to reach anybody. You can reach exactly the right people today, on the strength of a single good piece of content, if that content earns it. The game stopped being “how many followers have you racked up” and became “how good is this specific thing, for the specific people it gets shown to.” And that second game — that’s a game a good bartender with something real to say can win on day one, from a standing start, with two hundred followers and a phone. This is the most hopeful fact in the whole book, so let it land properly. The gatekeepers are gone. There is no follower threshold you must cross before you’re allowed to matter, no dues you have to pay in years of grinding first. There’s just the work, the people it’s right for, and a machine whose entire purpose is connecting the two. Make the thing good, and the door is already standing open.

15.2 How the Machine Actually Works

Under the hood, every one of these platforms does fundamentally the same thing with a new post, and once you see it you’ll never un-see it. It shows your post to a small seed group first. It watches, closely, how those people react in the first hour or so — a high-pressure little audition. If they respond well, it hands the post to a larger group, and if they respond well too, a larger one still, the thing rippling outward to more and more people who resemble the ones who already liked it. If they don’t respond, it quietly stops and lets the post die. That’s the entire machine, on every platform, underneath all the different logos and branding. Seed, test, expand — or seed, test, stop. Understanding this quietly heals the thing that breaks most people’s hearts online. When a post flops, it does not mean the platform hates you, or shadowbanned you, or that you’ve been secretly blacklisted — that’s the paranoia the hack-sellers love to stoke, because frightened people buy solutions.

It just means the seed group didn’t bite, so the machine moved on. That’s all it is. It isn’t personal, it isn’t permanent, and your next post starts fresh with a clean audition. Two useful things follow straight from that first-hour audition. First, post when your own people are actually awake — not at some universal “best time” off a blog — that’s nonsense — but at the hours your own crowd is on their phones. Second, be there in that first hour to answer the early comments, because a lively opening tells the machine the post is worth pushing wider.

So everything comes down to what “responding well” means, and it’s a short list in rough order of power. First and biggest: attention. Did people actually watch, and — for video especially — did they keep watching to the end? Completion is close to everything. Second, and this is the one people underrate: the quiet, powerful signals. Did someone save your post to come back to, or share it — above all, send it privately to a friend? A share is worth a mountain of likes, because it’s hard proof the thing carried real value, and on several platforms a private send is now the most powerful signal there is. A like is cheap and costs nothing. A share is a vote cast with a person’s own reputation attached to it. So if you want a single number to aim at, aim at shares and saves, not likes. A thousand likes feels lovely and does relatively little; fifty shares can set a post on fire, because every one of them tells the machine “this was worth interrupting someone’s day to send to them.” Everything that travels far travels because people handed it to each other. Likes are applause. Shares are fuel. Which signal weighs most shifts by platform — completion on TikTok, private sends on Instagram, session time on YouTube — which is what those platform chapters mapped. But the categories themselves are universal.

Then there’s the newer layer, the one that arrived in the last couple of years and changed the game for the better. The platforms got stricter and smarter about a single question: not just whether you grabbed attention, but whether people felt their time was actually well spent. Some now literally ask them — a little pop-up that rates your video one to five stars. The upshot is that the cheap tricks are dying off. Clickbait, fake hooks that don’t pay off, engagement-bait like “comment YES if you agree” — they’ll buy you one hollow moment and then get you quietly buried. The machine is steadily learning to reward the one thing that’s hard to fake, which is real value honestly delivered. And this is the best news in the chapter for someone like you, once you sit with it. Every shift the platforms have made lately punishes the fakers and rewards the real — which is exactly the ground a veteran bartender with true stories and real craft stands on most comfortably in the world. The machine is slowly being rebuilt to prefer the very thing you can’t help but be. Let the trick-chasers panic over the new rules. They were written in your favor.

15.3 The Handful of Things That Never Change

Here it is, then — the whole durable list, the physics you can build an entire career on without ever reading another trends article. Strip away every platform’s particular quirks and the same short set of principles is left standing, and it has not meaningfully changed in years: hook people fast, hold them with something worth their time, make it good enough that they’d share or save it, be clear and consistent about who you are, show up regularly, say what you are in plain words the machine can read, and make it native to the platform you’re on. That is the list. Master those seven things and you are free — you never have to chase another hack as long as you live. And notice that not one of those seven is a trick, a loophole, or a setting hidden in some menu. Every one of them is just a description of making something good for someone specific — the same instinct that already tells you which regular wants to talk tonight and which one wants to be left alone with their beer. You have the underlying skill. This is only about pointing it at a camera.

Let’s unpack the three that carry the most weight. The hook: you win or lose a person in the first two or three seconds, on every platform, and that will never change, so open with the interesting part and cut the slow windup entirely. Value: give them something real in exchange for the time they hand you — a laugh, a lesson, a feeling, a look at something they can’t see anywhere else — and don’t waste a single second of it. And the share, which is the highest bar and the one truly worth aiming at every time: would a real person send this to a friend? Build things worth sharing, and the machine passes them around for you, for free, all day long. And see how those three stack into a chain: the hook earns the watch, the value earns the finish, and the finish plus something worth keeping earns the share. They aren’t three separate tricks; they’re one linked sequence. Get the hook wrong and none of the rest ever even gets seen. Nail all three and you’ve built something the machine is practically obligated to spread. That chain is the whole craft of a clip, and we’ll take it apart bolt by bolt a couple of chapters from now.

Then the principle almost everybody misses, and it’s the quiet difference-maker: clarity and consistency about who you are. The machine is constantly trying to work out what you’re about, so it knows who on earth to show you to — and it figures that out by reading your recent posts as a body of work, your last dozen or so, looking for a pattern. If you’re all over the map — bar content today, your lunch tomorrow, a random rant the day after — it can’t classify you, so it shows you to no one in particular and your reach dies. Pick your lane, the one you found back in Chapter Five, and stay in it. Be reliably about something. A bartender who is consistently and clearly about their room and their craft is dead easy for the machine to recommend; a scattered, about-everything account is impossible to place, and the impossible-to-place get placed nowhere.

The happy news is that a honky-tonk bartender has one of the easiest lanes on earth to stay inside, because your world is so rich and so specific — the room, the music, the craft, the regulars, the whole living culture of the place. You could not run dry inside that lane if you tried. You only have to resist the pull to wander out of it chasing some trend that has nothing to do with who you are. Consistency isn’t a cage here. It’s a gift you’ve already been handed. And one more of the seven earns a quick word, because it’s the easiest to fumble: make it native. Every platform quietly favors content built for it and buries content that isn’t — the watermark from another app on your clip, the post stuffed with a link dragging people off somewhere else. Post it clean, where it lives, and let the platform keep its people.

15.4 What’s Actually New: Search and AI

Two new shifts are worth adapting to, because unlike the hacks, these are durable and pointed in a clear direction. The first is search. Every platform is quietly turning into a search engine. Something close to half of Americans now search on TikTok or Instagram the way they used to search Google — “best margarita near me,” “how to two-step,” “country bars in Fort Worth” — and the platforms now read everything to answer those searches: your caption, the text on your screen, and, newest of all, the actual words you speak out loud in the video. So the move is simple: say what you are, in plain words. Put the keywords in your caption and speak them in the clip. Make the how-to content people are literally typing into that search bar. Content built to be searched keeps getting found for months, not hours. This is a real gift for a bartender, because your craft is endlessly searchable. Someone, somewhere, is typing “how do I order whiskey,” “what is a Ranch Water,” “two-step for beginners” into that search bar right this minute. Answer those questions plainly in a clip, and you get found by the exact person hunting for exactly what you know — over and over, for as long as the video exists. For a working bar, evergreen beats viral every time.

The second new thing is AI, now sitting squarely in the middle of all of it, and there are two sides you need to know. On one side, AI increasingly decides what gets recommended, and it’s even starting to answer “where should I go tonight?” when a person asks a chatbot instead of a search engine — and it pulls those answers from clear, plainly-labeled content, which is one more reason to state simply and everywhere what you are and where you are. On the other side, and this is the good part: as the whole internet floods with AI-generated slop, the platforms are putting a growing premium on the real and the human. Genuine, slightly imperfect, obviously-made-by-a-person content is now outperforming slick AI polish, precisely because people are starving for something that feels real. You’ve felt this yourself, as a viewer. You can smell the fake content now — the too-smooth voiceover, the stock-footage sheen, the words no actual human would ever say out loud — and you scroll right past it, a little worn out by it. So does everybody else. Which means the bar for standing out is not higher production values. It’s more plain humanity: a real face, a real voice, a moment that obviously actually happened. The realer you are, the harder you cut through the noise.

Which loops right back around to the heart of this entire book, in a way that ought to make you smile. The newest, most sophisticated, most aggressively AI-driven version of these algorithms turns out to be hungrier than ever for exactly the one thing you happen to have in abundance: a real human being, in a real room, doing something real in front of real people. All that technology got more advanced and landed right back on authenticity as the thing it most wants to reward. You do not have to out-tech the twenty-two-year-olds with their AI tools and their trend spreadsheets. You only have to be more real than they can fake — and behind that bar, you already are.

15.5 Stop Chasing the Algorithm

Let me hand you the verdict plainly, because it will save you years. The specific hacks — the one magic posting time, the trending sound you absolutely must use by Thursday, the three secret words for your bio — change constantly, flatly contradict each other from one guru to the next, and are usually obsolete by the time they ever reach you. Chasing them is a treadmill, and the treadmill leads exactly two places: burnout, and a scattered, soulless feed that the machine can’t place and people don’t care about. The durable principles, by lovely contrast, are boring, free, and permanent. Put your energy there and nowhere else. Every hour spent hunting a hack is an hour stolen from making something actually good. And read “show up consistently” carefully, because it does not mean “post constantly”: a few good posts beat a daily flood of junk, and the machine now punishes the flood. That’s the exact trap the next chapter is built to keep you out of.

Here’s the reframe that sets you free from the whole anxious game. Stop picturing the algorithm as an enemy to outsmart, and start seeing it for what it honestly is: a machine built to do your job. It is trying, all day long, to read each individual person and hand them something they’ll love, so that they linger and so that they come back. That is precisely, exactly what you already do every night behind your bar. So the winning move was never to trick it and never could be. The winning move is to get so good at giving people something worth their time that the machine simply can’t help but keep passing you around. Serve the people in front of you, and the algorithm follows along behind like it was trained to. Because, in a sense, it was. Think of the best host you’ve ever worked a floor with — the one who remembered every name, read every mood, sat the right person beside the right person, and sent the whole room home wanting to come back next week. The algorithm is a clumsy machine straining to be that host at a scale of billions of people at once. When you make something that would delight the real people in your real room, you’re handing that machine exactly what it’s desperate for, and it pays you back in reach. Host well, and you win.

So that’s discovery, demystified and de-fanged: interest over followers, a simple test-and-expand engine humming under every platform, a short list of durable principles that matter, search and AI both pushing — of all the possible directions — back toward realness, and one firm instruction to quit chasing hacks and start being good. Which leaves the only question that matters from here: how do you make good content, consistently, on top of a full-time job behind a loud and crowded bar, without ever running dry or burning yourself to the ground? That is not a small question. It is, in fact, the entire rest of this part of the book. Let’s go build the engine that answers it. Because here’s the quiet truth sitting underneath this whole chapter: understanding discovery is the easy part, and you now understand it better than most people who do this for a living. The hard part — the part that truly separates the ones who make it from the ones who quit by spring — was never knowing what good content is. It’s making it, over and over, when you’re tired and slammed and it’s the very last thing you feel like doing. That’s a solvable problem, and solving it is exactly where we go next.