Across the world, boutique and medium-sized image libraries are increasingly creating their own bespoke keywording standards. They recognise that competing solely on volume with the global giants is impossible, so they’re differentiating themselves through precision and personality — building vocabularies that better serve their specialist subjects and client bases.
A fine-art library might develop language that reflects artistic movements, materials, and periods — “Impressionism, oil painting, watercolour study.” A sports archive could choose terminology favoured by its clients, using “grassroots football” rather than the broader “soccer.” And a heritage collection in Britain might insist on historically correct phrasing such as “Georgian townhouse” instead of “old building.” In each case, the keywords reflect not just what’s pictured, but how their audience searches.
By contrast, the major image aggregators — Getty Images, Shutterstock, Adobe Stock and others — work to strict, industrial-scale keywording standards. Their vocabularies are designed for consistency across millions of files, enforcing singular nouns, controlled terms, and tight ordering by subject, location, and concept. It’s an efficient system at scale, but inevitably generic.
This is where automation alone falls short. Most AI keywording tools are trained on vast, global datasets that default to broad, universal tags — “landscape,” “travel,” “person,” “nature.” For libraries with niche vocabularies or unique standards, those outputs are far too bland. AI can generate a lot of tags quickly, but without human intervention it rarely speaks the language of a specialist archive.
Picsell Media’s Gold AI Professional service solves this by combining AI speed with human intelligence. Before any work begins, our team reviews a client’s keywording guide — or helps create one — and then builds prompt cards and custom vocabularies that instruct the AI exactly how to tag within that framework. Those prompt cards embed the library’s preferred order, term limits, and phrasing rules, ensuring the AI produces output that fits the standard from the start.
Once the automated pass is complete, our human editors audit and refine every batch — correcting hierarchy, removing duplicates, and fine-tuning relevance to guarantee consistency with the client’s bespoke system.
This hybrid approach means smaller and specialist libraries can now enjoy the speed and cost benefits of AI keywording without losing the individuality of their brand. It’s the best of both worlds: automation for efficiency, and human expertise for accuracy, nuance and style.