The Fab Buyer's Field Guide

Why most listings are a gamble (and how not to lose the bet)

There are 79,792 listings on Fab as of May 2026, and 1,754 more arrived in May alone. That is a wonderful problem and a real one: somewhere in that pile is exactly the asset your project needs, and around it sit hundreds of near-lookalikes that will quietly eat your budget and your weekend. This guide is the set of questions I run before I spend a penny — and, because we sell on Fab ourselves, the same questions I'd want you to run on our listings before you trust us.

Start with the uncomfortable shape of the market. In our own compiled data, 70% of the listings launched in 2023 have never received a single review — and that figure has climbed steadily, up from 43% for the 2015 cohort. The catalogue has a long, silent tail: thousands of packs that shipped, surfaced for almost no one, and were never heard from again. The bestseller you've seen recommended ten times is the visible tip; most of the iceberg is dark.

Here is the trap, and it's the single most important idea in this guide. A brand-new asset with no reviews is not a bad asset. Our demand and review data is frozen at September 2024; it has not moved since. A large share of today's catalogue — well over a third — was released *after* that line and therefore shows zero questions and zero reviews because our coverage stopped, not because buyers rejected it. Silence on a recent listing is missing information, not a verdict. Silence on an *old* listing — something that has had a year or two in the wild and still has nothing — is a much louder signal. So the first move is always: check the listing's age, then read the silence accordingly.

The rest of this guide is how to fill in that missing information yourself, from the things a listing actually shows you.

• ✓Check the listing's launch date before you read anything into its review count.
• ✓Old + silent (a year or more, still no reviews or questions) is a genuine warning sign.
• ✓New + silent is simply unproven — judge it on the asset itself, not on the absence of a crowd.
• ✓Assume the bestsellers you keep seeing are a tiny, visible fraction of the catalogue, not the norm.

Ratings barely help

The instinct is to sort by stars and trust the number. On Fab, that instinct misleads you. In our frozen review data, 83% of all reviews are five-star. When nearly everything is five stars, the rating stops being a measurement and becomes wallpaper — it cannot separate the excellent asset from the merely adequate one, because both wear the same badge.

There are honest reasons for this. People who bought an asset they didn't like often just move on rather than write a one-star review; the buyers motivated to post are disproportionately the happy ones; and a small handful of reviews can be moved by a single enthusiastic or disgruntled voice. None of that is sinister — it's just how a low-volume, opt-in review system behaves. But the consequence for you is concrete: a 5.0 average across three reviews tells you almost nothing, and even a 5.0 across thirty tells you less than the rating system implies.

Treat the star average as a coarse filter at best — useful for spotting the rare genuinely-low score, useless for ranking the top of the pile. The real signal lives one click deeper, in the questions buyers actually asked. That's the next section.

• ✓Don't rank assets by star average — with 83% five-star, it can't separate good from great.
• ✓Treat a 5.0 over a tiny review count as 'unproven', not 'proven excellent'.
• ✓A rare low rating is worth reading carefully — it's swimming against a very strong tide.
• ✓Use stars to exclude, not to choose.

Read the buyer questions, not the stars

If ratings are wallpaper, the buyer questions are the structural inspection. They are the single most honest thing on a Fab listing, because they are written by people who were close enough to buying that they cared about the answer — and the answers reveal what the marketing copy carefully doesn't.

There's a deeper reason questions matter, and it's one of the most striking patterns in our whole dataset: demand inverts supply. Sellers crowd into 3D models — they're 57.1% of the entire catalogue — yet in our frozen demand data buyers ask about them least, at roughly 2.5 questions per listing. The most asked-about category by a mile is game templates at around 43 questions per listing — about 17 times the 3D-model rate — even though templates are thin on the ground. Audio sits at the quiet end, around 0.3 questions per listing. High question volume is where engaged buyers actually congregate; it's a far better signal of a living, used product than any star count.

When you read the questions on a listing, you're looking for two things. First, the *content*: do buyers ask whether it works with your engine version, your render path, your target platform? Those are the real-world failure points, surfaced by people who hit them. Second, the *response*: did the seller answer, and how? A creator who replies promptly and specifically is a creator who'll still be there when you file a bug. Unanswered questions on an older listing are the clearest 'this seller has moved on' tell you'll find. One caveat, always: our question data is frozen at September 2024, so a recent listing with no questions hasn't been abandoned — it simply postdates our window.

• ✓Open the questions tab before the description — it's the most honest part of the page.
• ✓Look for questions about your engine version, render path and platform; those are the real failure points.
• ✓Judge the seller by their answers: prompt and specific is good; ignored questions on an old listing is bad.
• ✓High question volume signals a living, used product far better than a high star average.
• ✓Remember the inversion: the crowded category (3D models) is the least-asked-about; thin categories like templates draw the most engagement.

Check engine support — it's the number-one way to waste your money

This is the cheapest mistake to avoid and the most common one made. 46.7% of the version-tagged catalogue is still tagged for Unreal Engine 4 — nearly half. The marketplace carries a large modernisation backlog: plenty of perfectly good assets that were authored for UE4 and never migrated. Buy one for a UE5 project without checking, and you've bought yourself a conversion job at best and an incompatible pack at worst.

The flip side is just as true for the modern catalogue: UE5 now accounts for 93.2% of new monthly releases, so newer assets overwhelmingly target UE5 — but the engine moves fast, and a 'UE5' tag spans everything from 5.0 to 5.7. Subsystems shift between those versions; a pack built against an older 5.x can still need adjustment on the current stable release. The fix is mundane and reliable: read the supported-versions field, match it to the exact version your project ships on, and don't assume 'UE5' means 'my UE5.'

If a listing is vague about engine support, that vagueness is itself information — treat it as a question to ask before buying, not an assumption to make after.

• ✓Find the supported-engine-versions field and match it to your project's exact version.
• ✓Don't treat a bare 'UE5' tag as 'works on my 5.7' — the range spans many releases.
• ✓For a UE4-tagged asset, budget for a migration pass, or skip it for UE5 work.
• ✓If engine support is left vague, ask in the questions before you buy — vagueness is a yellow flag.
• ✓Code plugins are stricter than content here: a version mismatch usually means it won't compile at all.

Does it have a video? (Especially for anything that moves)

Screenshots are easy to flatter. A hero render with the right lighting and a generous depth of field can make almost any asset look like the one you want. Motion is much harder to fake — which is exactly why a video is the strongest single trust signal a listing can offer, and why its absence on certain kinds of asset should give you pause.

Nowhere is this clearer than effects. In our VFX study, 89% of VFX listings embed a video — against roughly 48% market-wide — and that's no accident: a particle effect, a material that animates, a Niagara system, a piece of locomotion, all live or die in motion, and the good sellers know a still frame can't sell them honestly. For anything that *moves*, a listing with no video is asking you to buy on faith. The category-wide habit is so strong that a video-less VFX pack stands out as the odd one out, not the norm.

For genuinely static assets — a single prop, a material you'll only ever see still — a video matters less, and its absence isn't damning. But even there, ask what the screenshots are *not* showing you: wireframes, polycounts, the asset in a neutral grey-box scene rather than a beauty shot. The more a listing relies on a single flattering angle, the more questions you should ask before paying.

Disclosure: we sell Niagara VFX packs, audio packs and content packs on Fab, so this section overlaps what we make. We put video on our own effect listings for exactly the reason above — and if one of ours ever doesn't, hold it to the same suspicion you'd hold anyone else's.

• ✓For anything that moves — VFX, materials, animation, locomotion — require a video; ~89% of VFX listings have one, so a missing one is the exception.
• ✓Treat motion as the honest demo and a single hero still as the sales pitch.
• ✓For static assets, look for wireframes, polycounts and a neutral grey-box shot, not just a beauty render.
• ✓Ask what the screenshots deliberately aren't showing you.
• ✓Hold our own listings to this rule too — that's the point of writing it down.

Price sanity — and the cross-marketplace check

Once an asset has passed the trust checks, the last question is whether the price is fair — and here you have a firm anchor. The median price on Fab is $19.99, which is also the modal price: the catalogue clusters hard on the .99 line, and $19.99 is the gravitational centre of it. That's not a ceiling or a floor — substantial, complex packs reasonably cost more, and 3.1% of the catalogue is free — but it is the number to measure against. An asset priced far above $19.99 should be visibly, defensibly more than the median pack: more content, more support, more proven engagement in its questions. If it isn't, the price is doing work the asset can't back up.

Then run the cross-marketplace check, because it can quietly save you real money. Across the 2,260 assets we found cross-listed on two stores, the Fab listing is typically ~150% more expensive than the very same asset sold elsewhere. That premium isn't universal and it isn't a scandal — convenience, the in-engine workflow and Epic's ecosystem all carry value — but for a 3D model, a texture set or a generic content pack that plainly exists on more than one storefront, a sixty-second search before you pay is one of the highest-return habits in this whole guide.

A note on what price is *not*: it is not a sales figure and it is not revenue. A high price tells you what a seller is asking, never what anyone paid or how many bought. Don't read an expensive listing as a successful one, or a cheap one as a desperate one — read the questions and the engine support, then let the median be your fairness yardstick.

Disclosure: our own packs are priced on this same marketplace, around the same conventions. Apply the median test and the cross-store search to us as readily as to anyone — a fair price should survive the check.

• ✓Anchor every price to the $19.99 median; anything well above it should be visibly more pack than the median.
• ✓For generic, portable assets (3D models, textures, content packs), search other stores first — the Fab listing is often ~150% dearer.
• ✓Don't pay a premium for convenience you don't actually need.
• ✓Never read price as sales or revenue — it's an asking price, nothing more.
• ✓Free isn't a red flag — 3.1% of the catalogue is free, and some of it is excellent; still run the same trust checks.

The whole guide on one card: do and don't

Run these in order and you've done, in a couple of minutes, the diligence that separates a good buy from a graveyard pack. None of it requires inside knowledge — only the willingness to read past the hero shot and the star average to the things a listing reveals when you look.

And the standard cuts both ways. We sell on Fab; everything below applies to our listings as squarely as to anyone's. If one of ours fails a check — no video on an effect, a vague engine field, a price the pack can't justify — that's a fair reason to pass, and we'd rather you caught it than trusted us blind. A guide that only protected you from other people's listings wouldn't be worth the bytes.

• ✓DO check the launch date first, and read silence on new listings as 'unproven', not 'bad' — freeze isn't decay.
• ✓DO read the buyer questions before anything else, and judge the seller by their answers.
• ✓DO match the supported engine version to your project's exact version.
• ✓DO require a video for anything that moves, and look for wireframes and grey-box shots on static assets.
• ✓DO anchor price to the $19.99 median and search other stores for portable assets.
• ✓DON'T rank by star average — 83% are five-star and it can't separate good from great.
• ✓DON'T treat a recent listing's lack of reviews or questions as a verdict.
• ✓DON'T read price as sales, revenue or success — it's only what's being asked.
• ✓DON'T assume 'UE5' means your UE5, or that a bestseller's prominence is the catalogue norm.
• ✓DON'T take our word for any of this — run the checklist on us too.

The five-minute pre-purchase checklist

Run top to bottom before you buy anything on Fab. Demand-side signals (questions, reviews) are from the MythicLemon Marketplace Index and frozen at September 2024; supply, pricing and engine-version data are as of 2026-05.

FAQ