Most wedding photographers ship images to the web with the camera’s default name. Canon outputs IMG_4523.jpg, Sony gives you DSC09812.ARW, Fuji uses DSCF1043.JPG. Those names travel through Lightroom, through the export, through the upload, and end up in production. I did this for years. I had read the Google guidance on descriptive filenames and assumed it was a marginal factor, the kind of optimization you do once you have done everything else.
In late 2025 I put photography file naming to the test with 100 of my own wedding photos, renamed across a controlled six-week window. The data is interesting. The conclusion is that filenames are not a margin. They are a measurable, isolable input to image search performance, and the gap between what they cost and what they return is the largest I have seen on this site.
What Google actually says
The Google Search Central documentation on image SEO is older than most photographers realize, and the guidance is unchanged: “Use descriptive filenames and alt text for images. The filename can give Google clues about the subject matter of the image.” The wording is gentle, but the page has been tested for a decade by SEO teams with budgets, and the consensus from the SEO community is that filenames behave like a weak ranking signal that compounds with alt text and surrounding context.
What changed in 2025 is that Google’s image search has been overhauled twice, and image impressions in Search Console now include results from Google Discover, Lens, and the AI-driven shopping and travel surfaces. The signal landscape under image results is broader than it was in 2020. Filenames touch all of it.
I want to be careful here. Filenames are not magic. Renaming a portfolio of bad images to good filenames will not move bad images. The lever is small. But for photographers who already have indexable images on a fast site, it is a lever that costs almost nothing to pull.
The photography file naming test
I selected 100 wedding photos from 12 different gallery posts on this site, all of which had been live for at least six months and had stabilized in their Search Console performance. The images were all hosted under the WordPress media library at predictable paths and were already well alt-tagged.
The control: I left 50 of these images with their original Lightroom export filenames, which were a mix of IMG_*.jpg, couple-name-23.jpg, and _FKP1234.jpg. The treatment: I renamed the other 50 with a structured pattern: [venue-slug]-[couple-or-context]-[moment]-[year].jpg. Examples:
chateau-de-villette-laura-thomas-first-dance-2024.jpg chateau-malmaison-couple-portrait-golden-hour-2024.jpg val-de-loire-elopement-vows-2024.jpg
I redid the WordPress media replacement properly, regenerating thumbnails, updating the references in post content, refreshing the sitemap, and submitting the affected URLs to IndexNow. I did not change the alt text, the surrounding paragraph, or the file dimensions. The only variable was the filename and the URL slug under /wp-content/uploads/.
I let the test run for six weeks and pulled the Search Console data with a before-and-after split.
The numbers
I will be precise about scope. These are 100 images on a site that already had all of its other on-page SEO in good shape. I am not extrapolating to a magic uplift number for the whole industry.
The 50 renamed images saw a 31% increase in image impressions in the six-week window after the change, compared to the six-week window before. The control set increased 6%, which I treat as the seasonal baseline for that period.
Image clicks on the renamed set increased 22%. The control set was flat. Two images in the renamed set picked up first-page positions for compound queries that they had been ranking on page two for, and one query in particular (“first dance Château de Villette”) moved from impression-only to a steady 18 to 22 clicks per week.
The cleanest single signal: average position improved by 4.1 places on the renamed set and was unchanged on the control. That is not a difference I can hand-wave away as noise.
Why I think it works
The mechanism is not exotic. Google’s image index uses the filename, the URL, the alt text, the caption, and the surrounding paragraph as overlapping signals to figure out what the image depicts. Each of those signals is weak alone. Together they form a confidence score. If the filename is IMG_4523.jpg, that signal contributes nothing. If the filename is chateau-de-villette-laura-thomas-first-dance-2024.jpg, the same word appears in the URL, the alt text, the caption, and the heading nearby. The signals reinforce.
There is also a second-order effect that I did not expect. The renamed files saw more inbound links from Pinterest scrapers and curation blogs that grab images and reuse the filename. When a Pinterest pin uses the original filename as its description seed, a structured filename produces a more readable pin description without manual editing. I do not have hard numbers on this, but the qualitative pattern was clear.
What an intelligent rename pattern looks like
I will not pretend there is one universal pattern. The pattern I use is structured for venue-driven search behavior, because most of my queries follow the shape “venue + moment”. A photographer working in a different niche should think about how their potential clients actually phrase a search.
Three rules I have settled on. First, the filename should read like a fragment of a sentence, not like a database key. paris-elopement-eiffel-tower-blue-hour-2024.jpg reads. IMG-PE-ET-BH-24.jpg does not. Second, the filename should match the alt text shape, not duplicate it. The alt text describes the image for a screen reader. The filename describes what the image is filed under. Third, never repeat the filename across the gallery. Each image should have a unique path. Sequential variants like paris-elopement-vows-2024-01.jpg, -02.jpg are fine; identical paths get deduplicated by the index and you lose half your impressions.
For the actual rename work, I built a small tool that batches the rename, regenerates thumbnails, fixes the references in post content, and resubmits to IndexNow. I am opening it as part of PhotoSEO Vision in the next quarter. If you have a hundred posts and you do not want to do this by hand, that is the workflow I am rolling out.
What this does not solve
I want to head off the obvious overreach. Renaming files does not fix a slow site. It does not fix a Showit-style empty HTML problem, which I covered in the AI-invisible CMS article. It does not fix bad alt text. It does not fix images that are not actually relevant to the queries you want to rank for. It is a multiplier on existing strength, not a substitute for the work underneath.
The other thing it does not solve is AI search. AI crawlers in 2026 lean heavily on the page text and the structured markup, less on the image filename. I tested this in the third article in this series: How GPTBot, ClaudeBot, PerplexityBot See Your Portfolio. Filenames matter for image search and for Google’s classical pipeline. They matter less for the AI retrieval surface, where alt text and on-page context carry more weight.
The conclusion I am willing to defend
For a photographer who already has a fast, server-rendered site and good alt text habits, photography file naming across the existing image library is one of the highest return-on-effort interventions available. The work is mechanical, the variables are isolable, and the result shows up in Search Console within weeks rather than quarters.
For a photographer whose site sends empty HTML to crawlers, this is the wrong fix to do first. Move the platform, then the filenames matter.
If you want to see what I am working on in this space and the broader research program, the about page has the context.
