Cumulative, lagging, and volume-driven

The model, and its real numbers

Stock is the purest version of decoupled income: make an asset, put it where buyers search, and get paid each time somebody licenses it. Photographs, illustrations, vectors, video clips, 3D models, music, sound effects, typefaces.

The mechanism is honest and the numbers are widely misunderstood, so start with them.

Microstock photography and illustration pays a commission on each download, and for contributors without exclusivity the per-download earnings are frequently well under a unit of currency: a few tens of cents is a common experience. A library of a few hundred images typically produces a trickle; meaningful monthly income usually requires a library in the thousands, built over years.

Video footage pays substantially more per licence than stills, and the libraries are smaller because shooting is harder. This is one of the better ratios available.

Music and sound effects vary enormously by library and by whether the licence is per-track or subscription-funded, with the subscription model spreading a pool across contributors by usage share.

3D assets can pay very well per sale, because the buyers are studios with budgets and the production cost is high enough to limit supply.

Typefaces are the highest ceiling and the highest bar: years of work per family, a market that judges quality harshly, and the possibility of a single family earning for decades.

The property that decides everything

Stock income is cumulative and lagging. Each asset earns a small amount indefinitely, so income is roughly proportional to library size times age.

Two consequences follow, and both are counter-intuitive.

First, the first year looks like failure. Two hundred assets uploaded and a negligible monthly figure is the normal result, not evidence that you are bad at it. People quit at month eight, which is before the mechanism has had a chance to operate.

Second, volume beats perfection. Within the bounds of a library’s quality threshold, an asset that took four hours and one that took twelve earn similar amounts, because buyers are searching for a subject and choosing the first adequate result. This is uncomfortable for people who came to creative work for craft reasons, and it is the single most important operational fact about stock.

Choosing what to make

The people who earn from stock do not make what they want to make. They make what is searched for and under-supplied.

Search the library for your subject before you produce. If there are 400,000 results for “coffee cup on desk”, making another one is arithmetic you will lose. If there are 300 results for an accurately depicted regional business scene, a working practice specific to your country, or a piece of equipment used in an industry near you, that is a gap.

The strongest advantage most contributors have is local specificity. Stock libraries are heavily over-supplied with generic Western office imagery and under-supplied with authentic, correctly-detailed material from most of the rest of the world. Buyers in advertising and publishing need it and complain about not finding it.

Keyword honestly and thoroughly. Discovery is entirely search-driven, and an unkeyworded asset does not exist. This is tedious, unavoidable, and it is where many contributors lose most of their potential income.

Exclusivity

Most libraries offer higher commission for exclusive contribution. The trade is straightforward: a larger share of a single channel against a smaller share of several.

For a new contributor, non-exclusive is usually the better bet, because you do not yet know which library suits your material and you are not giving up much. Exclusivity becomes arguable once you have data showing one library produces the overwhelming majority of your sales.

Rejections, and why they happen

Every library rejects submissions, and the reasons are mostly technical rather than artistic: noise at the pixel level, over-sharpening, chromatic aberration, poor keywording, visible trade marks, recognisable people without releases, recognisable property, and “similar content already submitted”.

Read the rejection reasons rather than resubmitting. The first hundred submissions teach you the library’s technical standard, and after that rejection rates fall sharply.

Releases matter here more than anywhere. An image with an identifiable face and no model release will either be rejected or accepted as editorial-only, which drastically reduces what it earns.

Doing it with free tools

Everything needed is free. Darktable or RawTherapee for raw processing, GIMP for retouching, Inkscape for vectors, Blender for 3D, DaVinci Resolve for video, Audacity and open DAWs for audio.

The only genuine equipment constraint is in photography and video, where sensor quality affects whether images pass technical review. Even there, the gap has narrowed: a modern phone shooting raw in good light passes many libraries’ standards, and subject choice matters more than equipment for what actually sells.

Keeping the records

Track what you submitted, where, when, and what it earns. Without this you cannot tell which subjects work, which library performs, or whether the whole exercise is worth continuing, and “it feels like it isn’t working” is not a basis for a decision about a mechanism that is designed to look like nothing for a year.

The limitation

Stock is a volume business with a long lag, low per-unit returns, and rising supply, including a large and growing quantity of generated imagery, which is compressing prices at the generic end of several libraries and is subject to inconsistent and changing acceptance policies across them.

The defensible position within it is specificity: material that is hard to produce, hard to fake, verifiably real, correctly detailed and scarce. Generic imagery was already a poor bet before the supply expanded, and it is a worse one now. Treat stock as a slow accumulating asset built alongside other work, not as a plan for this year’s income.