
Keywording is the part of photo organization that eats the most time and gets abandoned the fastest. This is a smaller version: a starter vocabulary of about forty terms, hierarchies kept to two levels, the fastest ways to apply them in Lightroom, and a straight answer on when you should skip the whole thing.
Most advice on this treats keywording as a professional duty, somewhere between backing up and paying tax. It is not. It is a search index you build by hand, one photo at a time, and it only pays you back on the days you search.
So the first question is not "how do I keyword properly." It is "do I search my own archive." If you shoot client work, deliver it, and have never once gone looking for a photo of a bride laughing in the rain across six years of jobs, keywording is a weekend you will not get back. If you license images, submit to stock, pitch editors, write blog posts out of your archive, or get asked for "something like that barn wedding you did" twice a year, it pays repeatedly and quietly.
This guide is for the middle case, which is most working photographers: a small system, about an hour to set up, and a couple of minutes per shoot to maintain. It is deliberately not a complete taxonomy. If you want the wider picture of what else lives in your files, our photo metadata guide covers the copyright, contact and caption fields, which have a much better effort to reward ratio than this one does.
One piece of context worth having: keywords are a single IPTC field, dc:subject, defined in the standard as "keywords to express the subject and other aspects of the content of the image." It is a flat list of words attached to the file. Every hierarchy, parent term and synonym your software offers is a convenience layer sitting on top of that flat list.
A controlled vocabulary is just a fixed list of words you allow yourself to use. That is the entire concept, and it is the difference between a library you can search and a library where you type "beach" and get a third of the photos because the rest are filed under Beach, beaches, seaside, coast and shoreline.
Three rules make the list work, and all three are decisions you make once:
Here is a starter list of about forty terms. It is not a template to admire, it is genuinely the whole vocabulary a wedding and portrait photographer needs to begin with.
| Group | Terms |
|---|---|
| Shoot type | wedding, engagement, portrait, family, newborn, event, commercial, headshot |
| Who is in it | bride, groom, couple, child, group, guests, pet, details |
| Where | church, barn, beach, studio, garden, city, home, hotel |
| Light and time | golden hour, night, backlit, flash, overcast, indoors |
| Moment | ceremony, first look, speeches, first dance, getting ready, portraits |
| What it is for | portfolio, blog, print sale, published, model released |
Add a term only when you have searched for something twice and failed to find it. That rule alone will keep your list under a hundred words for years, and a hundred words you actually use beats a thousand you half remember.
Lightroom Classic lets you nest keywords inside each other, so "wedding" can contain "ceremony", "reception" and "getting ready". Adobe's documentation calls these keyword hierarchies, and you build them either by typing separators such as the pipe character when you enter a keyword, or by choosing Put Inside when you create a keyword tag.
What actually matters is what happens on export, and that is controlled by four options in the Create Keyword Tag and Edit Keyword Tag dialogs:
The practical limit is two levels deep, three at the absolute most. A five level tree looks impressive in the Keyword List panel and costs you a small amount of thought on every single frame, forever. The hierarchy is there to save typing, not to model the universe.
If you can type the child term and get the parents for free, it is working. If you have to stop and think about where a term lives before you can apply it, the tree is too deep.
Plenty of good photographers never keyword a frame, and their libraries are fine. What they have instead is structure.
If your archive is organized by date and client, and you find things by remembering whose wedding it was, then your folders and file names are already doing the job a keyword index would do. A folder called 2026-06-14 Ruiz-Okonkwo Wedding answers most of the questions you will ever ask it. Our guides to folder structure and a file naming system that lasts cover that route properly, and for many photographers it is the better investment.
Three other cases where the honest answer is to skip it:
The speed problem is almost never the thinking. It is the typing. Two Lightroom features solve most of it.
Keyword sets group keywords for fast access, with up to nine keywords in a set, applied with the number keys. Lightroom ships with sets including Recent Keywords, Outdoor Photography, Portrait Photography and Wedding Photography, and you can save your own. Build one set per genre you shoot and you will almost never type a keyword again.
The Painter tool lets you load a keyword and spray it across photos in Grid view by clicking or dragging. For applying "ceremony" to a run of 300 frames it is the fastest thing in the application.
Then there is the sequencing, which matters more than either tool. Apply the broad terms to the entire shoot at import: the genre, the venue, the city, the year. Apply the specific terms only to the photos you actually selected. A wedding is typically 2,000 to 4,000 frames captured and 400 to 800 delivered, so keywording after picking rather than before cuts the work by roughly three quarters before you have made a single decision differently.
That ordering is the single biggest time saver in this article, and it is worth saying plainly: pick first, keyword second. Detailed keywords on frames nobody will ever see is unpaid work, and it is the reason most keywording systems get abandoned in their first busy season.
The temptation, once the system clicks, is to go back and keyword ten years of archive. Resist it. It is forty hours of work with a payoff you cannot predict, and it is the classic way to spend a weekend and finish 2019.
Do this instead. Keyword the last twelve months, because that is what clients and editors ask about. Keyword whatever is already in your portfolio, because those are the images you will be asked to repeat. Then keyword forward from today and let the rest of the archive stay findable by date and client name, which it already is.
If you want to be more thorough than that, timebox it: twenty minutes per year of archive, broad terms only, one sitting. Anything you cannot do in twenty minutes for a given year is detail you were never going to search for. Our guide to organizing thousands of photos covers the wider cleanup if the library itself is the problem rather than the keywords.
Keywords are frequently sold as a client experience feature: your couple types "dad" into the gallery and finds every photo of him. Check that before you build a workflow on it. Whether a gallery host indexes the keywords embedded in your files varies a lot by platform, and the ones that do usually read only the flat keyword list rather than your hierarchy. Upload ten test images and search for one of your own terms before you promise anybody anything. Our gallery delivery guide covers what clients actually use in practice, which is mostly favorites and downloads.
Nor do keywords follow your photos onto social platforms. IPTC's public testing of how platforms handle embedded metadata, most recently run in spring 2019, found that Instagram and Twitter stripped it entirely on both viewing and download, along with several large free stock sites. Platforms change quietly, so test with your own file rather than trusting a seven year old result, but plan on social being a lossy channel.
What keywords reliably do is serve you. They are how you pull an album selection, assemble a blog post, answer a vendor's request for photos from their venue, and find the three frames that belong in a pitch. On your own website, embedded descriptions and keywords also play a part in making images findable, which is a separate discipline covered in photo metadata for SEO.
Systems do not fail because they were badly designed. They fail in August when there are four weddings in the diary. These five habits are what survive that.
One last thing about the order of operations. Keywording works best when your picks are settled first, which means the picking pass has to be quick enough that it actually happens before the editing does. Kepla is being built to do that first pass for you, checking every frame from a shoot and handing back a shortlist, and it never deletes, moves or renames a file, so any keywords and metadata you applied at import are untouched. The picking app for iPhone, iPad and Mac is still in development. The booking page is live and free today.
For your own archive, five to ten is plenty, and most of those should come from parent terms exported automatically rather than words you typed. Stock and licensing work is the exception, where twenty to forty terms per image is normal because search there is a numbers game. Beyond that you are adding noise, since every extra term is one more thing to spell consistently forever.
It is a fixed list of keywords you allow yourself to use, decided in advance. The point is consistency rather than coverage: one word per idea, one choice on singular versus plural, one spelling standard. Without it a library ends up with beach, Beach, beaches, seaside and coast all describing the same thing, and no single search finds all of them. Forty terms is a realistic starting list.
They can, but it depends on your settings. In Lightroom Classic each keyword has an Include on Export option, and a separate Export Containing Keywords option controls whether its parent terms travel with it. Your export preset also has to be set to include metadata rather than stripping it. Export one test file and inspect it before you assume anything, because several common presets strip IPTC by default.
No. Apply broad terms such as the genre, venue and year to the whole shoot at import, then apply specific keywords only to the photos you selected. A wedding is typically 2,000 to 4,000 frames captured and 400 to 800 delivered, so keywording after picking rather than before takes roughly three quarters of the work out without changing a single decision you make.
In practice they are the same thing under different names. Keywords is the term used by the IPTC Photo Metadata Standard and by Lightroom, and they are stored in a single field, dc:subject. Tags is the more common word in consumer apps and on the web. If a piece of software offers both, they are almost always writing to the same underlying field.
They contribute, but they are not the main lever. What matters most for image search is the page the image sits on: the file name, the alt text, the surrounding copy and the page title. Embedded keywords and descriptions support that rather than replacing it, and they are stripped entirely by many social platforms, so treat them as one part of a wider approach.
Kepla for Mac clears the obvious misses from a card, names the reason on every frame it sets aside, and leaves the choosing to you. Nothing is ever deleted, moved or renamed. Free through the private preview · the first hundred photographers keep it at $99 a year.