StockForge — Stock Library & Digital Asset Command Center

Shot once. Sold ten thousand times.

StockForge is a complete, private, 100% offline operating system for a stock library — photographs, illustrations, vectors, mockups, textures and presets, sold through agencies, packs and direct licences. 15 integrated modules and 39 tools in a single HTML file, plus a live 13-sheet Excel workbook. Nothing installs. No account. No file, shoot, rate or customer ever leaves your computer.

How few of your files are actually earning?

Everybody in stock knows a small number of files pay for the whole library. Almost nobody has measured their own version, because doing it properly needs one correction that is easy to get wrong and fatal to get wrong.

In the worked example inside — 2,366 files, eleven shoots, thirty-six months of downloads across four agencies — the top 1% of files earned 21.6% of the money and the top 5% earned 48%. Half of everything came from 119 files, which is 5.5% of the library. A quarter of it, all old enough to have been found, has never been downloaded once.

And the young files are not counted as failures

This is the correction. A file uploaded six weeks ago with no downloads has not failed — it has not been indexed, surfaced or found yet. Counting it beside a three-year-old file that never earned turns a portfolio into a graveyard, and makes every share wrong in the same direction by an amount that depends entirely on how much you happen to have uploaded lately.

So every figure is computed only over files old enough to have been discovered, at a window you set and can move, and the young ones are reported separately. Move the window and watch every share on the screen move with it. Any tool that hides that choice is making it for you.

Volume up, money down — and the treadmill that creates

In the example, downloads rose 23% and revenue fell 4.5%. Revenue per download went from $0.46 to $0.36, and from $0.55 when the ledger opened. Which means this: you now have to grow downloads by 28.7% every year simply to earn what you earned last year. That is not a target. It is the treadmill, and no agency dashboard will show it to you, because the rate is the number they set.

Your season is probably not the month you think

Add up every January, every February and so on and the example says the peak is May. It is not. The library trebled across the window, so raw monthly totals measure growth and label it season. De-trended against a centred twelve-month average — the correction that costs six months at each end and that almost nobody applies — the peak is November and the trough is July. Six months out. StockForge shows you both answers side by side and the correction between them, then gives you a shoot-by date.

What generation changed, measured properly

Splitting your subjects into exposed and not-exposed and comparing what each earns per file is close to worthless: the two groups never had the same demand, and medical always outsold abstract backgrounds. StockForge runs a difference in differences instead — each group indexed to itself across a shared baseline, so a permanent demand gap and the growth of the library both divide out. Then it prints the divergence, prints the file counts behind it, and refuses to name a cause, because there is no control group and no experiment. That judgement is yours, and it is a real one.

Twenty-one screens

The library — What Actually Earns (the flagship, with a Lorenz curve and a Gini coefficient) · The Asset Register, filtered and sorted over thousands of files · Shoots & Production Runs, which refuses to judge a shoot too young to be judged · Rejections & Dead Weight, which keeps refused frames and accepted-but-never-found files strictly apart.

What it earns — Revenue Per Download · The Agencies, with share of downloads against share of money · Exclusive Or Not, which prints the break-even instead of a verdict and tells you the rate an agency would have to offer · Packs & Direct Sales, where the normalisation is printed beside the answer rather than quietly performed.

Being found — Keywording, a measured 3.7× spread reported as a correlation and priced only as a conditional · Subjects & Demand · Seasonality · What Generation Changed.

The business — How A File Ages, censored the same way · What A File Costs, the ladder from per-frame-shot to per-file-that-paid-for-itself · Payouts & Thresholds · Releases & Rights.

Nine calculators

What a file really costs · Where the rate is going · An exclusivity offer · Keywording deeper · A pack against the agencies · A payout threshold · Shifting the mix · A file over its life · The dead weight. Every one opens with your own figures already in it.

Twelve copy-ready pieces

A shoot brief · a model release · a property release · an appeal against “similar content” · an appeal against “commercially saturated” · a note to an agency about an exclusivity offer, written in break-even language · a pack listing · an extended licence summary · a keywording checklist · a cull checklist to run before you upload · chasing an unsigned release · a quarterly review addressed to yourself.

What you get

• StockForge-Stock-Library-Digital-Asset-Command-Center.html — the whole application, one file, offline
• StockForge-Companion-Workbook.xlsx — 13 live formula-driven sheets
• Quick-Start-Guide.html — eight steps to your first real number
• How-To-Use-Guide.html — the full manual, screen by screen
• README.txt and LICENSE.txt

Genuinely private. It makes no network requests of any kind. Your shoots, files, rates and agency figures stay in your own browser and are never transmitted anywhere.

Honest by design. Young files are never counted as failures. Per-download and per-pack are never averaged without the normalisation printed beside the answer. The keywording relationship is labelled a correlation and priced only as a conditional. The seasonal index shows its own working, including the wrong answer it corrects, and refuses to call a month a peak on too few observations. Exclusivity is reported as a break-even rather than a verdict, because the verdict depends on a number nobody knows. Nothing is extrapolated past the months your ledger covers, so the lifetime hourly figure is described as a floor everywhere it appears. And there is deliberately no hourly rate for a single year, because the money that arrived this year was earned by files shot across three — the two figures describe different populations and dividing one by the other describes neither.

StockForge performs arithmetic on figures you supply. It is not legal, tax or accounting advice. Agency contracts, exclusivity terms, model and property releases, trademark and editorial-use restrictions and what an extended licence actually permits are matters to take proper advice on rather than to settle from a template.