Stock Photo Keywords: The Complete Beginner's Guide

You took a great photo. You uploaded it to a stock platform. Nobody found it.

The image wasn’t the problem. The keywords were. Or more likely, the keywords didn’t exist at all.

Stock photo keywords are the single most important factor determining whether your images get discovered by buyers. They’re also the part most beginners skip, rush, or get wrong. This guide covers what keywords actually are, why they matter so much, and how to write them correctly from day one.

What Stock Photo Keywords Are (And What They’re Not)

Stock photo keywords are descriptive words and phrases attached to your image as metadata. When a buyer searches for “woman working from home” on Adobe Stock, the platform checks every image’s keywords to find matches. Your image either has those terms or it doesn’t.

Keywords are not hashtags or SEO tricks. They’re not optional fields you fill in later. They’re the bridge between what a buyer needs and what you uploaded. Without that bridge, your image sits in a database of 350+ million assets with zero chance of appearing in search results.

Every major stock platform requires keywords as part of the upload process. Adobe Stock requires a minimum of 5 keywords per image (though 15+ is recommended for real visibility). Shutterstock requires at least 7 and recommends 25-45. Both cap at 50. The numbers vary by platform, but the principle is universal: more specific, accurate keywords means more visibility.

Why Keywords Matter More Than Image Quality

This sounds backwards, but it’s true. A technically average image with excellent keywords will outsell a stunning image with poor keywords every time.

Stock platforms are search engines first, marketplaces second. A graphic designer looking for a “modern kitchen interior” types that query, the platform filters its entire library, and the results page shows maybe 50 images out of millions. If your beautiful kitchen photo doesn’t have “modern,” “kitchen,” or “interior” in its keywords, it never appears. The buyer never sees it.

Image quality determines whether someone clicks “download” after finding your image. Keywords determine whether they find it at all. You can have the best photo in the world, but if the metadata doesn’t describe it accurately, it’s invisible.

Ask any contributor who went back and re-keyworded their early uploads. Better metadata on the same images leads to more downloads. The images don’t change. The keywords do.

The Three Types of Metadata

Keywords don’t exist in isolation. They’re part of a metadata package that every stock platform requires. Understanding all three parts helps you write better keywords.

Title. A short, natural-language description of the image. “Young professional working at laptop in bright home office” works. “Beautiful amazing stunning workspace photo” doesn’t. The title is the strongest text signal for search on most platforms.

Keywords. The 15-50 individual words and phrases that describe everything relevant about the image. This is where you cover subject, setting, mood, style, concepts, and technical details.

Category. A dropdown selection from the platform’s taxonomy. Business, Lifestyle, Nature, Technology, and so on. Getting this wrong means your image appears in the wrong section when buyers filter by category.

All three work together. A strong title with weak keywords underperforms. Strong keywords with the wrong category cost you filtered searches. For a platform-specific breakdown of exactly how each field works, see our Adobe Stock keywords guide and Shutterstock keywording guide.

Five Types of Keywords Every Image Needs

Beginners usually write one type of keyword: literal descriptions of objects in the frame. That’s a start, but it covers maybe 30% of the searches your image could match. A complete keyword set covers five layers.

1. Subject Keywords

What is literally in the image. Be as specific as possible.

“Espresso cup on marble countertop” is better than “coffee.” “Golden retriever puppy” is better than “dog.” Specific terms match specific buyer searches. Vague terms compete against millions of images and lose.

If the image shows a person, include observable details: approximate age range, activity, setting. “Woman in her 30s typing on laptop” works. Avoid guessing demographics you can’t see.

2. Setting and Context Keywords

Where and when the scene takes place.

“Home office,” “outdoor cafe,” “urban rooftop.” “Morning light,” “rainy day,” “autumn season.” These keywords catch buyers who search by scenario rather than by subject. A marketing team looking for “remote work” imagery isn’t searching for “laptop.” They’re searching for the situation.

3. Style and Technical Keywords

How the image looks, separate from what it shows.

“Natural light,” “overhead shot,” “shallow depth of field,” “warm tones,” “minimalist composition.” Buyers often filter by visual style, especially for editorial or brand work. Two images of the same coffee cup look completely different in flat lay versus close-up. Style keywords capture that difference.

4. Conceptual Keywords

What the image represents beyond its literal content.

“Productivity,” “morning routine,” “work-life balance,” “wellness,” “independence.” These are the keywords that power advertising and brand campaigns. A company building a campaign around “sustainability” searches for that concept, not for “woman holding reusable water bottle.” If your image can represent an abstract idea, keyword it.

5. Emotional Keywords

The mood or feeling the image communicates.

“Calm,” “energetic,” “contemplative,” “joyful,” “serene.” These overlap with conceptual keywords but focus specifically on emotional tone. They’re useful because many creative briefs are written in emotional terms: “We need something that feels warm and inviting.”

The Seven Mistakes That Kill Visibility

These are the patterns that keep beginner portfolios invisible. Fixing them has an immediate impact on downloads.

Too few keywords. Using 7-10 keywords when the platform accepts 50 is leaving discoverability on the table. Target 30-45 keywords per image. Not padding with junk, but covering all five layers described above.

Too generic. “Nature,” “business,” “technology” as standalone keywords match millions of images. You’ll never rank for them. “Monstera plant on white shelf” beats “nature.” Specificity is how small portfolios compete.

Keyword stuffing. Adding every remotely related term degrades your relevance score. Stock platforms track how often buyers who search for your keywords actually download your image. Irrelevant keywords bring the wrong traffic and tank your relevance metrics.

Brand names. Including “Apple,” “Nike,” “Starbucks,” or any trademarked name in commercial image keywords gets you rejected. Even if the product appears in the frame, the brand name can’t appear in commercial metadata.

Describing what’s not there. Keywords should describe visible content. If there’s one person in the image, don’t keyword “group” or “team.” Platforms cross-check keywords against visual content, and mismatches trigger rejections or penalties.

Inconsistent terms across similar images. If you have 10 images from the same shoot, calling the subject “freelancer” in some and “remote worker” in others splits your search presence. Pick consistent terminology across related images so they reinforce each other in search results.

Ignoring keyword order. On most platforms, keywords listed first carry slightly more weight. Put your most important, most specific terms at the top of the list. Generic terms go at the end.

How to Write Keywords for Your First 10 Images

Theory is useful. Practice is what builds a portfolio. This is the process for keywording images when you’re starting from zero.

Step 1: Describe what you see. Open the image and list every object, person, activity, and setting detail visible in the frame. Don’t overthink it. Just write down what’s there.

Step 2: Add context. What scenario does this image depict? What would a buyer use it for? A photo of a salad isn’t just food. It’s “healthy eating,” “meal prep,” “lunch,” “clean eating,” “plant-based diet.”

Step 3: Add style terms. How is the image shot? Overhead? Close-up? Natural light? Moody lighting? Bright and airy? Add 5-8 terms describing the visual approach.

Step 4: Add concepts. What abstract ideas does this image communicate? Health, productivity, relaxation, creativity, growth? Add 3-5 conceptual terms that a brand manager might search for.

Step 5: Check the count. You should have 25-45 keywords. If you’re under 20, you’re missing a layer. Go back through the five types and fill the gaps.

Step 6: Check for mistakes. Scan for brand names, terms describing content not in the image, and duplicate/near-duplicate terms (“coffee” and “cup of coffee” and “coffee cup” all at once). Remove anything that doesn’t belong.

This process takes 3-5 minutes per image when done properly. Tedious? Yes. But that time investment is what separates images that sell from images that collect dust.

The Scale Problem

Three to five minutes per image sounds reasonable until you’re uploading regularly. At 50 images per week, you’re spending 4 hours just on metadata. At 100 images across two platforms, it’s 8 hours. And the quality degrades. By image 30 in a session, most people are copying keywords from earlier uploads and skipping the conceptual layer entirely.

This is the hidden cost of manual keywording that stops most contributors from building a portfolio fast enough to generate meaningful income. The images are ready. The bottleneck is the metadata.

Keywords for AI-Generated Images

If you’re creating stock images with AI tools, the keywording strategy is identical. The same five layers apply. The same mistakes apply.

Two extra considerations. Most platforms now require you to flag AI-generated content during upload. And some platforms restrict AI content to specific categories or apply additional review criteria. The keywords themselves follow the same rules. For a full guide on building an AI stock photo business, including keywording strategy, see our complete guide to making money with AI stock images.

Quick Reference: Platform Keyword Limits

Platform	Min Keywords	Max Keywords	Recommended
Adobe Stock	5	50	15-35
Shutterstock	7	50	25-45
Zedge	2	10	8-10 tags
Displate	None	15	10-15 tags
TeePublic	None	15	10-15 tags

Each platform weighs keywords differently and has its own rules for what gets accepted or rejected. Our platform-specific guides cover the details for each one.

What to Do Next

Stock photo keywords are straightforward once you understand the system. Describe what’s in the image across all five layers. Be specific. Avoid the seven mistakes above. Do this for every upload and your images will start appearing in searches instead of sitting invisible in a database.

If you’re uploading to Adobe Stock specifically, read our Adobe Stock keywords guide for the exact field requirements and layered strategy that works on that platform. For video content, our stock video keywording guide covers the additional considerations for motion footage.

The difference between a portfolio that earns and one that doesn’t is rarely the images. It’s the metadata.

Try it free: Generate AI keywords for your stock photos, no account needed. Works with 5 platforms.