The store wants to sell you what's already selling. That's the whole problem in one sentence. A recommendation engine's job is to reduce risk, and the safest bet it can make is the game a million people already bought, which is exactly the game you didn't need help finding.
Real indie discovery means going looking for things with a few hundred reviews, not a few hundred thousand. The good news is that the tools to do it already exist. They're just not the big button on the home page.
Why the storefront front page is working against you
The front page is optimised for conversion, not taste. It surfaces what's trending, what's on sale, and what people with your purchase history also bought. All of that skews toward things that are already popular, because popularity is the signal the system trusts most.
That's fine when you want the safe pick. It's useless when you want the weird puzzle game two people you've never heard of made in a garage. The algorithm has no way to distinguish "nobody bought this because it's bad" from "nobody bought this because nobody's seen it yet." To the machine those look identical.
So the first move is to stop treating the home page as discovery. It's a bestseller shelf with better lighting.
Start with the reviews you can actually read
Filter by review count, not review score. A game sitting on a couple of hundred reviews is in the zone where every review was written by someone who chose that game on purpose, not because a trailer chased them around the internet for a month.
Sort a store's search results by "most recent" or by release date rather than relevance, and set a review-count ceiling if the filters allow it. You're deliberately looking below the popularity waterline. The score matters less than the shape of the reviews: a game with three hundred reviews and a lot of specific, slightly obsessive praise is usually onto something.
One caution. A small review count also means less of a safety net, so it's worth knowing how to tell a good indie game from a bad one before you buy before you spend anything on a title with barely any footprint.
Follow people, not systems
The single best discovery tool is a human with a bias you understand. A curator who loves cosy management sims will steer you wrong on horror and perfectly on cosy management sims, and knowing which is which is worth more than any neutral algorithm.
Find two or three creators whose taste overlaps yours by maybe seventy percent. Not a hundred, because someone who only likes what you already like tells you nothing new. You want the person who's slightly braver than you, who plays the thing you'd never have clicked and occasionally comes back saying it was extraordinary. Playkora's creator directory is one place to start building that shortlist.
The overlap number is the whole trick. Follow someone with wildly different taste and every recommendation misses. Follow your own clone and you learn nothing. The useful person is the one who's right often enough that you trust them, and different enough that they surprise you.
Play the demos, especially during festivals
Demos are the most honest thing in games. Nobody makes a good demo out of a bad game by accident, and a bad demo tells you exactly as much as a good one, faster.
Storefronts run regular demo festivals where hundreds of unreleased games put a slice up for free at once. This is the closest thing to a discovery goldmine that exists, because it's built entirely around games that don't have an audience yet. Set aside an evening, install six demos on a whim, and uninstall five of them in ten minutes each. That fifteen minutes with the sixth is what you came for.
The uninstalling is not failure. It's the process working. You're spending minutes to filter what would otherwise cost you money and regret.
Dig through tags like a rabbit hole
Every store lets you browse by tag, and tags get weirdly specific once you leave the big categories. "Roguelite" is a crowd. "Deckbuilder" plus "cosy" plus "no combat" is a back room where the interesting stuff lives.
Stack three or four tags that describe the exact thing you're in the mood for, then sort by something other than popularity. The more specific your combination, the less the popularity signal matters, because there simply aren't enough games in that niche for a bestseller list to form. You've filtered your way past the algorithm by being too particular for it to help you.
Then use the game you land on as a jumping-off point. Look at what its store page lists as similar, look at what the developer made before, look at who published it. Small publishers tend to have a house style, and finding one you like is like finding a shelf instead of a single book.
Watch what developers themselves recommend
Indie developers play each other's games and talk about them constantly, often more generously than any marketing channel. A developer boosting another small game is putting their own reputation next to it, which is a stronger endorsement than a paid banner.
Follow a few developers whose released work you've enjoyed and pay attention to what they get excited about. It won't be filtered through a sales team. It'll be a person who understands how hard the thing is to build telling you someone nailed it.
Know what you're buying before it's finished
A lot of the most interesting small games are still in progress, sold while they're being built. That can mean getting in early on something special, or it can mean paying for a promise that never fully arrives.
The difference between those two outcomes is knowable in advance if you read the signs, which is why it's worth understanding what Early Access actually means before you buy. Treat an unfinished game as a bet on the developer, not on the current build, and check whether they've shipped anything before.
Keep a list, and let it get out of date
The last piece is boring and it's the one that actually works: write games down when you spot them, and don't buy immediately. A wishlist is a cooling-off period disguised as a to-do list.
Half the things you add you'll never buy, and that's the point. The ones that keep pulling at you weeks later, the ones you keep checking for a sale, those are the real hits. Impulse buys favour whoever shouted loudest. A list favours whatever you actually can't stop thinking about.
Browsing a curated catalogue like Playkora's games section works the same way. You're building a personal shortlist that owes nothing to what's trending.
The one thing to do first
Go find your first demo festival, or if one isn't running, install three demos of games with under a thousand reviews right now. Not games you've heard of. Games you found by stacking tags or following one creator whose taste you trust.
You'll uninstall most of them. The one you don't is the whole reason indie games are worth the search.
Common questions
Are low review counts a red flag or a good sign?
Both, and you tell them apart by reading. A low count with detailed, specific praise usually means an undiscovered game. A low count with vague reviews, or a lot of complaints about crashes and missing features, means proceed carefully. The number alone tells you nothing; the content of the reviews tells you everything.
How do I find indie games that aren't on the big storefronts?
Itch.io and developer sites host a lot of experimental and free work that never reaches the major stores. Game jam collections are especially rich, because they're hundreds of small ideas made under a deadline, unfiltered by any commercial pressure. It's rougher terrain, but the hit rate on genuine novelty is high.
Should I trust review scores or ignore them?
Use them as a floor, not a target. A score tells you a game clears a basic bar of not being broken. Beyond that, the reasons behind the score matter more than the number, because a game can be divisive precisely because it's doing something unusual, and unusual is often what you're hunting for.
How much should I spend on a game nobody's heard of?
As much as you'd happily lose. Small games are lower-risk bets by price, so treat a purchase as a punt rather than a considered investment. If it flops for you, it cost about the same as a coffee. If it lands, you found something before everyone else.