Questions to Ask Before Hiring a Photo Editing Company

Hiring a photo editing company looks like a simple purchasing decision: compare a few per-image rates, send a test, pick the cheapest one that looks fine. In practice it is closer to hiring a production partner. The company you choose will touch every product image on your site, your marketplace listings, your catalogs and your ads, sometimes thousands of files a week, and the quality of that work decides whether your range looks like one brand or a pile of unrelated photos. A weak supplier costs you in re-edits, missed launch dates and, less visibly, in shoppers who do not trust what they see.

This guide is for e-commerce managers, studio and production leads, marketplace sellers, photographers and agency producers who are about to interview editing suppliers. It gives you the questions to ask, grouped by theme, and for each one explains why it matters, what a good answer sounds like and what a bad one sounds like. Use it in a call, in a written request for proposal, or as a checklist while you review a test batch.

Throughout, we quote our own published starting rates and the typical US market ranges we have reviewed, so you have a reference point for pricing conversations. Those are starting prices, not guaranteed totals; the actual cost of any job depends on the drivers explained in the pricing section.

Why the choice of photo editing company matters more than the per-image rate

The per-image rate is the number everyone compares, and it is the least informative one. Two suppliers can both quote a low rate for a clipping path and deliver completely different work: one with hand-drawn paths that follow the product edge, the other with a rough automated cut that leaves halos around the product and flattens every shadow. The price difference shows up later, as rework, as a slow launch or as a category page where nothing quite matches.

Three things make this choice hard to reverse. First, consistency: once a supplier has edited half your range to their standard, switching means either re-editing that half or living with two standards on one page. Second, the reference: a good company invests in understanding your products, backgrounds and color targets, and that knowledge does not transfer automatically. Third, your calendar: seasonal launches, sales events and marketplace deadlines leave no slack for a supplier who misses a batch.

If you are still deciding whether to outsource at all, our article on in-house or outsourced photo editing sets out the trade-offs. If you have decided to outsource, these are the questions that matter most, in the order we would ask them:

1. Will you edit a test batch of my own images, to my brief, before I commit?
2. How do you agree and lock a reference image before volume production starts?
3. How do you keep color and style consistent across hundreds of images?
4. Who actually does the editing, and who checks it before it reaches me?
5. What exactly does the per-image price include, and what makes it go up?
6. Who owns the edited files, paths and masks, and how is my material kept confidential?
7. What happens when a batch comes back wrong?

The rest of this guide expands those seven into a fuller interview, theme by theme.

Fit and process: will this company work the way your images need to be worked?

These questions establish whether the supplier is set up for your kind of work and has a real production process, rather than just a team of editors and an upload link.

What kinds of images do you edit most, and can I see work like mine?

Why ask it: Editing is specialized. A company that mostly cuts out hard goods on white may struggle with fashion, where ghost mannequin compositing and fabric texture matter, or with jewelry, where reflections and metal tones are the whole job. You want evidence of your category, not a generic before-and-after gallery.

A good answer sounds like: They name the categories they handle every week, show examples close to your products, and are frank about what they do less often. They ask what marketplaces or channels the images are for, because that changes the output specification.

A bad answer sounds like: Everything is possible and every category is a specialty. The portfolio is a handful of heavily polished hero images with no batch work, and nobody asks what the images are for.

Will you edit a test batch of my own images before I commit?

Why ask it: A portfolio shows the best work a company has ever done. A test on your own files shows what they will do with your lighting, your backgrounds and your problems. Choose test images that include your hardest cases: hair, glass, dark garments on dark backgrounds, reflective products.

A good answer sounds like: Yes, on a small set of your images, edited to your written brief, returned with a note of any assumptions made and any problems in the source files. They ask for the brief first rather than just editing to their own taste.

A bad answer sounds like: They only want to test on easy images, they edit the test with far more care than the production batch will get, or they refuse any test and ask for a large upfront commitment.

How do you agree and lock a reference image before volume production?

Why ask it: The reference is the single most important control in batch editing. It fixes background value, crop, margins, shadow, color and retouching depth, and every later image is matched to it. Without a locked reference, each editor makes their own choices and the batch drifts.

A good answer sounds like: A clear sequence: they edit a reference image, you approve it, and only then does volume production start. For our own work the published timeline is reference image approved in 1–2 business days, first batch back 2–3 business days after the reference is signed off, and volume production at typically 200–500 images per day once the reference is locked. A good supplier can describe a sequence like that for their own process.

A bad answer sounds like: There is no reference step. They start the whole batch immediately and promise to fix anything you do not like afterward, which usually means you become the quality control.

How do files move in and out, and how are they named and organized?

Why ask it: Handling thousands of files by hand is where errors creep in: missing images, overwritten masters, files renamed so your platform cannot match them to SKUs. File handling is dull, and it is where many projects go wrong.

A good answer sounds like: A defined transfer method, your file names and folder structure preserved exactly, a manifest or count check on intake and delivery, and a clear rule for what happens to originals. They can also handle naming conventions and embedded data if you need it; our guide to bulk metadata editing covers the decisions involved.

A bad answer sounds like: Files arrive renamed, some are missing and nobody notices until you do. Or the answer is simply "we will send a link" with no mention of counts or naming.

That is our own published project timeline, and it is a useful benchmark when you ask other suppliers how their process runs. As an illustrative worked example: a batch of 2,000 product images, once the reference is locked, would take roughly 4 to 10 production days at 200 to 500 images per day, plus the time to approve the reference and review the first batch. Exact throughput depends on complexity, so a batch of simple cut-outs sits at the faster end and a batch of hair and glass at the slower end. If a supplier promises much more with no reference step, ask how.

Craft and quality: how the editing is actually done

These questions get past the sample images to the methods behind them. The answers tell you whether quality will hold on image 1,500 as well as on image 1.

Are your clipping paths hand-drawn, automated, or a mix?

Why ask it: Automated background removal has improved, and it is fine for some uses. But its failure modes (soft halos, missing gaps between straps or handles, chewed edges on fine detail) are exactly what makes a product page look cheap. You need to know which method is used on which images.

A good answer sounds like: Specific: hand-drawn vector paths for product cut-outs, with anchor points placed on the true edge; masking rather than paths for hair, fur and soft edges; any automated step followed by manual checking. Our clipping path work, for example, uses hand-drawn vector paths priced by how many curves the subject has.

A bad answer sounds like: Vague claims about AI tools that do everything, or an insistence that every image is hand-drawn at a price that could not pay for it.

How do you keep color consistent across a batch?

Why ask it: Consistency across a range is what sells it. Matching four hundred images to one reference costs more than editing four hundred images well individually, and it is also the part that customers notice when it fails: a navy that is blue in one shot and nearly black in the next.

A good answer sounds like: They work to a reference and to your color targets, use calibrated monitors, ask for physical sample references or a color chart in your captures, and check images side by side rather than one at a time. They may use lookup tables for shared corrections; our article on LUTs and color lookup for photos explains where those help and where they do not.

A bad answer sounds like: "Our editors have a good eye." No mention of references, calibration or side-by-side review.

What does your quality control look like, and who does it?

Why ask it: Every editor makes mistakes: a missed dust spot, a stray pixel of old background, a crop a few pixels off. What matters is whether someone other than the editor catches them before you do.

A good answer sounds like: A separate checking step by someone other than the editor, a written checklist based on your brief, and a record of what failed and why. They can tell you what they check at full zoom and what they check as a grid.

A bad answer sounds like: The editor checks their own work, or quality control means "the client reviews and sends back anything wrong".

How do you handle the hard cases: hair, glass, transparency and ghost mannequin?

Why ask it: Hard cases are where the price differences are real. Hair and fur need masking, glass and transparent materials need rebuilt reflections and edges, and ghost mannequin work means compositing necks, cuffs and hems from separate exposures. For fashion, our guide to fashion lookbook editing describes the extra steps.

A good answer sounds like: They describe the method for each case, show examples on similar products, and are honest about what cannot be fixed in editing, such as a missing inside-neck shot. They price these cases separately rather than hiding them in a flat rate.

A bad answer sounds like: One flat rate for everything, with no distinction between a box and curly hair on a busy background. That usually means the hard images are rushed.

Team: who will actually edit your images

The people behind the rate matter as much as the rate. These questions tell you whether the company can keep quality steady when volume rises or when someone is away.

Who will actually edit my images, and are they your own staff?

Why ask it: Some editing companies are a sales front for a changing pool of subcontractors. That is not automatically bad, but it makes consistency and confidentiality harder to guarantee, and you should know which model you are buying.

A good answer sounds like: A clear description: in-house editors or a named, managed partner network, how editors are trained on a new client's brief, and whether the same people usually stay on your account from batch to batch.

A bad answer sounds like: Evasion, or a promise that "the best editors" will work on your images with no explanation of who they are or how they are chosen.

How does capacity scale when I send a batch several times my usual size?

Why ask it: Seasonal launches and sales events produce spikes. A supplier who handles your normal week well may fall apart when you send five times as many images, and the drop in quality often appears only after delivery.

A good answer sounds like: They explain how they add capacity (more trained editors, shifts, a planned schedule), ask for a forecast of your peaks, and are honest about how much notice they need. They mention that the reference keeps new editors on standard.

A bad answer sounds like: "Unlimited capacity" with no explanation. Unlimited capacity at a fixed quality level does not exist.

What happens when the editor who knows my account is unavailable?

Why ask it: If all the knowledge of your standard lives in one person's head, your quality depends on their holiday calendar. Documentation is the difference between a supplier and a freelancer.

A good answer sounds like: Your brief, reference images and any exceptions are documented and stored with your account, and at least one other editor or lead is familiar with your work.

A bad answer sounds like: "It will not be a problem", with nothing written down.

Communication: how problems surface and get fixed

Most editing projects that go wrong do so because a problem was noticed late. These questions establish how quickly issues will reach you and how easy it is to get them fixed.

Who is my point of contact, and when are they available?

Why ask it: Many editing teams work in a different time zone from their clients, which can be an advantage: you send files at the end of your day and wake up to finished work. It only works if there is overlap for questions and a named person who answers them.

A good answer sounds like: A named account contact, stated working hours in your time zone, a stated response time for questions, and a route for urgent issues during a launch.

A bad answer sounds like: A generic support inbox, or a contact who changes with every batch.

How do I request revisions, and how are they tracked?

Why ask it: Revision requests are where rework multiplies. If feedback is scattered across emails and chat messages, the same mistake comes back in the next batch.

A good answer sounds like: A single place for feedback, with each revision tied to a file name, a clear turnaround for a normal round (our published figure is 1 business day for a normal round), and a habit of updating the brief when feedback reveals a new rule.

A bad answer sounds like: Revisions handled informally, with no record, and the same correction needed batch after batch.

Will you tell me when my source files have a problem?

Why ask it: Some problems cannot be solved in editing: a missing ghost mannequin inner shot, a product photographed out of focus, a color cast so strong the original color cannot be recovered. A good supplier flags these before spending hours on them.

A good answer sounds like: Yes, proactively, with examples, and with a suggestion for the shoot next time. They distinguish between what they can fix, what they can improve and what needs a reshoot.

A bad answer sounds like: They edit everything silently and deliver weak results, or they only mention problems when you complain.

Contracts and rights: ownership, confidentiality and what happens when things go wrong

These questions are easy to skip in the excitement of a good test batch. They matter most when something goes wrong, which is exactly when it is too late to negotiate them. This is practical guidance, not legal advice; have your own counsel review any contract.

Who owns the edited files, the paths and the masks?

Why ask it: You paid for the work, so you should be able to reuse it: to re-crop for a new channel, to re-edit next season or to move to another supplier. Paths, masks and layered masters are what make that possible.

A good answer sounds like: You own the delivered files, and layered masters or path files are available if you need them, with any additional charge stated upfront. The contract says so plainly.

A bad answer sounds like: The supplier keeps the layered files and treats them as a reason you cannot leave, or ownership is simply not addressed.

How do you keep unreleased products and my files confidential?

Why ask it: Editing suppliers see products before they launch. A leak of an unreleased collection or product design can be costly, and marketplaces and brand partners may require you to control who sees images.

A good answer sounds like: They sign a confidentiality agreement, limit access to the people working on your account, describe how files are stored and transferred, and state when files are deleted after delivery.

A bad answer sounds like: No agreement, files passed through personal accounts, or an unwillingness to say where files are stored.

What is your redo policy if a batch does not meet the brief?

Why ask it: Even good suppliers occasionally miss. What matters is whether fixing an error to the agreed brief is included, and how fast.

A good answer sounds like: Errors against the agreed brief and reference are corrected at no extra charge within a stated time. Changes to the brief after approval are handled as new work, priced openly. The distinction is written down.

A bad answer sounds like: Every correction is billed, or the reverse: unlimited free revisions, which usually means there is no firm brief and you will spend your time art-directing.

Pricing: what the rate includes and what moves it

Price is best discussed last, once you know what you are buying. Image editing is priced per image, with the rate dropping as volume rises. For reference, these are our published starting rates alongside the typical US market ranges we have reviewed. Our rates are starting prices; the final price depends on the drivers below. Our article on how much photo editing costs goes into clipping path pricing in more depth.

What exactly is included in the per-image price?

Why ask it: Two identical rates can cover very different work. One may include dust removal, shadow creation, color matching to a reference and three crops; another may cover only the cut-out, with everything else billed as extras.

A good answer sounds like: An itemized description of what the rate covers, what counts as an extra, and how many output versions are included. They point out that output versions multiply the work: one master and five crops is six versions per image, and thirty-six files once each is exported in three formats and two color spaces, not one.

A bad answer sounds like: A single headline rate with the extras discovered on the invoice.

How do complexity, volume and turnaround change the rate?

Why ask it: These are the main cost drivers. Edge difficulty is the single biggest one: a rectangular box is not a bicycle, and a bicycle is not curly hair on a busy background. How much retouching matters too, since dust removal is minutes while rebuilding a reflection, reshaping a garment or repairing a bad capture is a different order of work. Per-image rates drop with quantity, because the reference and the setup are amortized across the batch. And overnight work is real overtime for real people and is priced that way.

A good answer sounds like: They explain their tiers in these terms, show how a volume commitment changes the rate, and state the rush premium before you need it.

A bad answer sounds like: One flat price regardless of complexity or deadline, which means either the easy images are overpriced or the hard ones are underworked.

How do you decide which tier an image falls into before work starts?

Why ask it: Disputes about price usually come from images being reclassified after the work is done. Agreeing the tier in advance avoids that.

A good answer sounds like: They review a sample of the batch, classify it by complexity, and confirm the price before production. Unusual images are flagged individually rather than silently moved to a higher tier.

A bad answer sounds like: The tier is decided on the invoice.

Aftercare: the next season, the next channel and the long-term relationship

A good photo editing company is still useful after the first project. These questions test whether the relationship will get easier over time.

Do you keep my references and brief for the next batch or season?

Why ask it: Most product ranges are photographed repeatedly: new colorways, restocks, next season's styles. If the supplier starts from scratch each time, you pay for the learning curve again and the new images will not match the old ones.

A good answer sounds like: Your brief, reference images and exceptions are kept on file for a stated period, and new batches start from them. They offer a quick check of new images against the existing range.

A bad answer sounds like: Every project is treated as new, and you are asked to resend your brief each time.

Can you deliver the outputs each channel needs?

Why ask it: Marketplaces, your own site, print catalogs and social ads each have their own size, background, color space and file format needs. Our guides to image editing for marketplaces and image editing for print catalogs cover the specifications in detail.

A good answer sounds like: They ask which channels you use, know the common requirements, and propose an output set produced from one master rather than separate edits for each channel.

A bad answer sounds like: They deliver one file and leave you to resize and convert it, or they are unaware that print needs a different color space from the web.

What happens if my standard changes partway through?

Why ask it: Brands change their minds: a new background tone, tighter crops, a different shadow. The question is whether the supplier can switch cleanly and how the images already delivered are handled.

A good answer sounds like: A change is agreed in writing with a new reference, applied from a stated batch onward, with a clear, priced option to update earlier images for consistency.

A bad answer sounds like: The change is applied informally, so the range ends up with a mix of old and new standards.

The full checklist of questions to ask a photo editing company

Copy this list into your request for proposal or keep it open during interview calls. You do not need to ask every question of every supplier; drop the ones that do not apply to your images.

Scoring photo editing suppliers side by side

After the calls and the test batch, score each supplier on the same criteria so the decision does not come down to whoever gave the most confident pitch. Score each criterion from 1 (poor) to 5 (excellent), based on evidence from the test and the answers, then multiply by a weight that reflects what matters for your images. The weights and scores below are illustrative; set your own before you see the tests, so the results do not bend the criteria.

In this illustrative example, Supplier B has the most attractive price but the weakest process, and the weighted total reflects that. Supplier C wins on process and quality while costing more per image. Whether that premium is worth it depends on how many images you send and how visible they are. For a hero range on your own site, it usually is. For a long tail of simple marketplace listings, Supplier A might be the practical choice. The point of the scorecard is to make that trade-off explicit rather than letting the lowest rate decide by default.

How to run the test fairly

Send every supplier the same images, the same written brief and the same reference, and ask for the same outputs. Include at least one easy image, one typical image and one hard image. Review the results side by side at full size and as a grid, because some problems only show at full zoom and others only show when images sit next to each other. If possible, have two people score independently and compare.

When you are ready to compare, you can see how we approach this work on our image editing and retouching services page, and browse more guides in our image editing and retouching articles.