Product Photo Editing Cost: In-House vs Outsourced Retouching at Scale
Product Photo Editing Cost: In-House vs Outsourced Retouching at Scale
Built for e-commerce teams pushing hundreds or thousands of SKUs to Amazon, eBay, Etsy, Shopify, Walmart, Zalando and Otto. No fluff, just the math.
Why the "cost per image" question is harder than it looks
Ask three e-commerce managers what they pay to edit a product photo and you will get three wildly different answers: "nothing, my assistant does it," "about two euro," and "I have no idea, it is bundled into a retainer." All three are wrong in the way that matters, because none of them is looking at the fully loaded cost.
The price of a single edited image is not the sticker price on an invoice or the per-hour rate of the person clicking in Photoshop. It is the total money and time it takes to move one raw capture to a store-ready asset that passes marketplace validation, matches your other listings, and does not come back for a second round. When you frame it that way, the cheap in-house option often turns out to be the expensive one, and the "premium" outsourced option often turns out to be the cheapest per finished, accepted image.
This article breaks the number down honestly. We will cost out each type of edit, show how volume compresses the per-image price, build a full in-house budget with real euro figures, compare freelancers and a specialised service, and finish with a break-even formula you can run on your own numbers in about five minutes.
What actually goes into the price of one edited image
Before comparing suppliers, you have to compare like for like. "Photo editing" covers a range of tasks that differ in difficulty by an order of magnitude. A flat lay on white that needs a clean cutout is not the same job as a diamond ring shot that needs stone-by-stone retouching. Pricing that ignores the type of work is meaningless.
Here is the ladder of complexity, from simplest to most demanding, with typical market ranges per image at moderate volume.
| Edit type | What it involves | Typical range per image | Time per image (skilled) |
|---|---|---|---|
| Background removal / clipping | Cut the product from its background, place on pure white or transparent | 0.30-1.20 euro | 1-3 min |
| Full product card | Cutout, colour correction, shadow, crop to marketplace spec, sizing | 1.00-3.50 euro | 3-8 min |
| Ghost mannequin | Combine inside and outside shots so apparel keeps its 3D shape with no visible model | 2.50-6.00 euro | 8-15 min |
| Jewellery / high-complexity retouch | Reflection cleanup, stone clarity, metal smoothing, dust and scratch removal | 4.00-12.00 euro | 15-40 min |
Three cost drivers explain that spread.
Labour time. A background swap is a two-minute job for someone competent. Jewellery retouching on a faceted stone can eat forty minutes because every facet reflects the studio and each reflection has to be managed by hand. Time is the single biggest input, so complexity scales price almost linearly with minutes.
Skill scarcity. Anyone can learn to clip a shoe. Far fewer people can retouch a watch dial or build a convincing ghost mannequin from two shots. Scarcer skill commands a higher rate, and you cannot fake it with a cheap operator without the rejects piling up.
Consistency overhead. On a catalogue you are not editing one image, you are editing a set that must look identical: same white point, same shadow, same crop ratio, same margin. Enforcing that sameness across hundreds of files is real work that a per-image quote for a single hero shot never captures.
This is why a blanket "how much does photo editing cost" question has no answer until you specify the mix. A store selling plain apparel lives mostly in the first two rows. A jeweller or a watch reseller lives in the bottom two and should budget accordingly.
How volume drives the price down
The per-image numbers above are for moderate batches. Volume changes them because the fixed costs of any editing operation, whether internal or external, get spread across more units.
Setup and briefing are one-time costs. Defining the look, building the crop templates, agreeing the shadow style and the white point, running the first approval loop, all of that happens once. Amortised over 50 images it is expensive per unit. Over 5,000 images it disappears into the noise.
Throughput lets a supplier lower rates. A predictable pipeline of thousands of images a month means a specialised service can keep a trained team fully utilised, which is exactly when they can discount. This is why bulk pricing exists.
Volume pricing normally works in tiers or on a retainer.
| Monthly volume | Typical outsourced rate (full product card) | Structure |
|---|---|---|
| Under 100 images | 2.50-3.50 euro | Pay per image, no commitment |
| 100-1,000 images | 1.50-2.50 euro | Tiered per-image, minor discount |
| 1,000-5,000 images | 1.00-1.60 euro | Volume tier or light retainer |
| 5,000+ images | 0.70-1.20 euro | Retainer with reserved capacity |
A retainer reserves a fixed slice of a team's monthly capacity for you at a locked rate. You commit to a floor of volume, they commit to turnaround and price. For a store with steady catalogue churn this is usually the cheapest structure and the most predictable for planning. See how a dedicated pipeline is organised for bulk product photo editing.
The takeaway: if you are quoting your own economics, use the rate that matches your real monthly volume, not the single-image sticker price. A team doing 3,000 images a month is not paying the 3.50 euro rate, they are paying closer to 1.20.
The true cost of doing it in-house
This is where most cost comparisons fall apart, because people compare an outsourced invoice against an in-house salary and stop there. The invoice is fully loaded. The salary is not. To compare fairly you have to build the complete in-house cost and divide by realistic output.
Let us cost a single mid-level retoucher in a Western European market. Adjust the absolute numbers for your own region, the structure holds everywhere.
Direct costs
| Line item | Annual cost (euro) | Notes |
|---|---|---|
| Gross salary, mid retoucher | 38,000 | Junior 28k, senior 50k+ |
| Employer taxes and social charges | 11,400 | Assume roughly 30 percent on top |
| Software (Creative Cloud, plugins) | 720 | Per seat, per year |
| Workstation, calibrated monitor, amortised | 900 | Hardware over 3 years |
| Colour calibration, backup, storage | 400 | Ongoing |
| Desk, office overhead allocation | 3,600 | Space, power, IT |
| Direct subtotal | 54,020 |
The costs people forget
The subtotal above is already 42 percent higher than the bare salary, and it is still incomplete. Three hidden costs decide whether in-house actually pays.
Management time. Someone has to brief, review, approve and correct. If a manager on a 60,000 euro salary spends 20 percent of their week overseeing image production, that is 12,000 euro a year of loaded management cost attached to the operation. Call it 10,000 euro conservatively.
Idle time and seasonality. Product photography is not a flat load. You shoot a big drop before a season, then volume falls off a cliff. A salaried retoucher is paid 52 weeks a year but only has enough work for perhaps 32 of them. That off-season idle time is pure sunk cost, and it is the single most damaging line in the in-house model for a seasonal catalogue.
The productivity ceiling. One person has a hard output limit. A strong retoucher does maybe 40 to 60 finished product cards a day, so roughly 900 to 1,300 a month at full tilt, and full tilt is not sustainable. When a 4,000-image drop lands, one in-house seat physically cannot clear it on time, and you are back to overtime, rush freelancers, or delayed launches. You cannot flex a salary up and down with demand.
In-house cost per image, the real number
Take the loaded annual cost including management: roughly 54,000 plus 10,000, so 64,000 euro a year.
Now the honest output. In theory a retoucher clears around 12,000 images a year at 1,000 a month. In practice, with holidays, sick days, meetings, the seasonal idle weeks, revisions, and the fact that nobody sustains peak pace, realistic accepted output is closer to 8,000 to 9,000 finished images a year.
- On paper: 64,000 / 12,000 = 5.30 euro per image
- Realistic: 64,000 / 8,500 = 7.50 euro per image
That is for standard product cards. The person who quoted themselves "about two euro" was looking only at a slice of salary divided by peak theoretical output, and ignoring taxes, software, management, idle season and revisions. The real fully loaded in-house cost is three to four times what they think.
And note the ceiling: this seat maxes out around 8,500 accepted images a year no matter how much you spend. Growth past that means hiring a second seat, doubling the fixed cost, whether or not the second person is busy year-round.
The freelancer model: cheap rate, variable everything
Freelancers sit between in-house and a managed service. Their headline rate is attractive because you carry none of the employer overhead. But the loaded cost hides in reliability.
The good. No taxes, no idle season, no hardware, no seat you pay for in the quiet months. You pay per image or per hour, only for work delivered. A skilled freelancer can be excellent, and for a small, occasional volume they are often the right call. Rates commonly land at 1.50 to 4.00 euro per product card depending on skill and region.
The problems that cost money.
Consistency risk. One freelancer means one style, which is fine until they are unavailable and you bring in a second, whose white point and shadow differ subtly. Across a catalogue those subtle differences read as a messy storefront. Enforcing a single look across a rotating cast of freelancers is a management job you now own.
Capacity risk. A freelancer is one person with their own ceiling and their own other clients. When your big drop coincides with their busy week, your deadline is not their priority. You have no reserved capacity and no service guarantee.
Turnaround variance. Solo operators get sick, travel, and take other jobs. Turnaround that was three days last month can be nine days this month with no warning, right when a launch depends on it.
Management load. Sourcing, vetting, onboarding, briefing, quality-checking and paying several freelancers is real recurring work. Factor 3 to 5 euro of your own loaded time per image in coordination on a badly run freelancer operation, and the "cheap" rate is not cheap.
Freelancers are genuinely good for low, irregular volume where a missed deadline is not catastrophic. They struggle exactly where volume e-commerce lives: high count, tight and repeatable deadlines, and strict visual consistency.
The specialised service model at volume
A specialised bulk editing service is a managed team with a defined pipeline, a quality layer, and reserved capacity. You are not buying one person's hours, you are buying a guaranteed output at a locked rate. This is the model built for catalogue-scale work. A closer look at how the relationship is set up is here: outsource product photo editing.
What you are actually paying for beyond the pixels:
- Consistency by design. A team following one style guide with a QA checkpoint produces a uniform catalogue regardless of which editor touched which file. The sameness is enforced systematically, not left to chance.
- Elastic capacity. A 4,000-image drop and a 200-image top-up are handled by the same pipeline without you hiring, firing or begging a freelancer. You scale up and down by volume, not by headcount.
- Predictable turnaround. Committed delivery windows because the capacity is staffed and reserved, not dependent on one person's calendar.
- No overhead, no idle. You pay only for accepted images. Software, hardware, training, taxes, holidays, the quiet season, all of it is the supplier's problem, not a line on your budget.
- Marketplace fluency. A service that ships to Amazon, eBay, Etsy, Shopify, Walmart, Zalando and Otto every day already knows each platform's image rules, so fewer listings bounce on validation.
At real e-commerce volume the per-image rate lands between roughly 0.70 and 1.60 euro for product cards depending on tier, with complex categories priced up. Critically, that number is fully loaded and final: it already includes management, QA, revisions and capacity. There is no hidden multiplier on top the way there is with in-house salary or freelancer coordination.
Three models side by side
| Factor | In-house | Freelancers | Specialised service |
|---|---|---|---|
| Loaded cost per image (product card) | 5.30-7.50 euro | 2.00-4.50 euro incl. coordination | 0.70-1.60 euro |
| Scales up for a big drop | No, hard ceiling | Poorly, one person's limit | Yes, elastic capacity |
| Scales down in off-season | No, you pay idle salary | Yes, pay per job | Yes, pay per image |
| Consistency across catalogue | High, one person | Low to medium, varies | High, enforced by QA |
| Turnaround reliability | Medium, single point of failure | Variable | High, committed windows |
| Management burden on you | High | High | Low |
| Fixed cost commitment | Very high (salary) | None | None or light retainer |
| Best fit | Constant huge volume, brand-critical hero work | Low, irregular volume | Steady mid to high volume |
The pattern is clear. In-house wins only when volume is both enormous and constant enough to keep a seat saturated all year, and even then it cannot flex for peaks. Freelancers win at the low, irregular end. For the broad middle where most growing stores actually operate, steady mid to high volume with tight deadlines and consistency demands, the specialised service is both the cheapest per accepted image and the lowest risk.
The hidden cost nobody budgets: rejects and rework
Every model above has a shadow cost that rarely appears in the comparison, and it can dwarf the per-image rate: listings that get rejected or images that come back for rework.
When an image fails marketplace validation, the wrong background, the wrong dimensions, a watermark, a shadow that violates the platform spec, the cost is not one re-edit. It is the editor's redo time, plus the manager's time to catch and route it, plus the delay before the SKU goes live, plus the revenue lost while a sellable product sits dark. A listing that goes live three days late on a fast-moving item can lose more margin in those three days than the entire editing budget for that image.
Do the math on a modest 5 percent rework rate over a 2,000-image month. That is 100 images redone. At even a conservative 10 euro of total loaded cost per rework cycle (redo plus coordination plus delay), that is 1,000 euro a month of pure waste, on top of your editing spend, and it is invisible unless you track it.
This is where the cheap option quietly becomes expensive. An underpriced operator or an overloaded freelancer produces more rejects, and the rework bill lands on you. A service with a QA layer that catches errors before delivery may charge a higher headline rate and still cost less all-in, because your reject rate drops toward zero and your listings go live on the first pass. When you cost editing, cost the rejects too.
Find your own break-even point
Enough ranges. Here is how to find the exact volume where outsourcing beats in-house for your business. Run it on your own numbers.
The formula
Step 1. Your fully loaded in-house cost per image.
In-house per image = ( annual salary
+ employer taxes
+ software + hardware
+ management time
+ office overhead )
/ realistic annual accepted output
Use realistic output, not theoretical. For one mid retoucher, roughly 8,000 to 9,000 finished product cards a year after holidays, revisions and off-season idle.
Step 2. Your outsourced cost per image.
Take the tier rate that matches your monthly volume from the pricing table above. This number is already fully loaded, so no additions.
Step 3. Compare at your real monthly volume.
Monthly in-house cost = monthly volume x in-house per image Monthly outsourced cost = monthly volume x outsourced per image
But there is a catch with in-house that the multiplication hides: the fixed cost does not scale down. If you hire a retoucher, you pay their full loaded salary whether you send them 200 images or 1,000. So the real in-house cost per image rises sharply as your volume falls below their capacity.
Worked example
A store ships 700 product cards a month, so 8,400 a year, with a strong seasonal peak.
In-house. One seat, loaded cost 64,000 euro a year. At 8,400 images that is 7.62 euro per image. But 700 a month is below the seat's capacity, so you are also paying for idle time you cannot recover. And when the seasonal peak hits 2,000 in a month, one seat cannot clear it and launches slip.
Outsourced. 700 a month sits in the 100-1,000 tier at roughly 1.80 euro per image (a bit above the floor because it is mid-tier). Monthly cost: 700 x 1.80 = 1,260 euro. Annual: 15,120 euro.
In-house annual: 64,000 euro. Outsourced annual: 15,120 euro. Outsourcing is cheaper by roughly 49,000 euro a year, absorbs the seasonal peak without slipping a single launch, and carries zero fixed commitment. At this volume it is not close.
Where does in-house catch up?
In-house per image only drops to competitive levels when the seat is saturated all year. At a genuine, sustained 1,000+ images a month with no seasonal collapse, loaded cost approaches 5.30 euro per image. That is still above a mid-tier outsourced rate. For in-house to truly win on cost you generally need multiple saturated seats doing very high constant volume, plus a reason the work must stay internal, such as brand-critical hero imagery or tightly coupled studio-to-edit workflows.
For the overwhelming majority of stores under a few thousand images a month, or with any seasonality, the break-even never arrives. Outsourcing is cheaper at every realistic volume they will hit.
When in-house is still the right call
To be fair, in-house genuinely wins in specific cases, and it is worth naming them honestly.
- Constant, saturating volume. If you reliably produce well over 1,000 images a month, every month, with no off-season, a saturated seat becomes cost-competitive and gives you tight studio-to-edit integration.
- Brand-critical hero and campaign work. The flagship images that define your brand may justify an in-house specialist who lives inside your creative team, even while the bulk catalogue goes out.
- Extreme confidentiality or same-hour turnaround. Unreleased products under strict embargo, or a workflow that needs edits back within the hour in the same room as the shoot, can favour internal staff.
- A studio that shoots continuously. If you run your own high-volume studio, an on-site editor closing the loop between capture and delivery has real workflow value beyond pure cost.
The smart pattern for many scaling brands is hybrid: keep a small in-house capability for hero and campaign work, and push the bulk catalogue, the 90 percent of SKUs that just need clean, consistent, spec-correct cards, to a specialised service. You get brand control where it matters and volume economics where it counts. Building those consistent cards at scale is exactly the product card creation service job.
Frequently asked questions
What is a fair price to edit one product photo?
For a standard product card at moderate volume, expect 1.00 to 3.50 euro outsourced, dropping toward 0.70 to 1.20 euro at high volume on a retainer. Simple background removal is cheaper, from around 0.30 euro. Complex jewellery or ghost mannequin work runs 2.50 to 12 euro because it takes far more skilled time. Beware any quote far below the range, it usually signals rushed work and a high reject rate.
Is in-house really more expensive than outsourcing?
Almost always, until your volume is both very high and constant. The trap is comparing a salary against an invoice. Once you load the salary with taxes, software, hardware, management time, office overhead and, crucially, off-season idle time, and then divide by realistic rather than theoretical output, in-house lands around 5 to 8 euro per image. That is above outsourced rates at every volume most stores actually run.
How does volume lower the price?
Fixed setup and briefing costs spread across more images, and a supplier can discount when a steady, predictable pipeline keeps their trained team fully used. That is why per-image rates fall from around 3 euro at low volume to near 1 euro on a high-volume retainer. Committing to a monthly floor via a retainer usually unlocks the best rate and the most reliable turnaround.
What about the cost of rejected listings?
Underbudgeted and frequently the largest hidden cost. A rejected or reworked image costs redo time, management time, and lost revenue while the SKU sits offline. A 5 percent rework rate on 2,000 images a month can quietly burn 1,000 euro. A service with a real QA layer often costs less all-in despite a higher headline rate, because near-zero rejects mean listings go live on the first pass.
Can freelancers handle bulk catalogue work?
For low or irregular volume, yes, and often well. At catalogue scale they struggle with the three things that matter most: consistency across a rotating cast, reserved capacity for big drops, and reliable turnaround under tight deadlines. The coordination load also lands on you. They are a good tactical tool, not a scalable catalogue engine.
How do I calculate my own break-even?
Divide your fully loaded annual in-house cost by realistic annual output to get your true in-house per-image cost, then compare it against the outsourced tier rate for your monthly volume. Remember that in-house fixed cost does not scale down, so below a saturated seat your real per-image cost climbs. For most stores under a few thousand images a month, or with any seasonality, outsourcing wins at every volume.
The bottom line
The cost of editing a product image is not the rate on the invoice, it is the fully loaded cost of a finished, accepted, on-spec asset that never comes back. Measured that way, in-house is the expensive option for almost everyone except the highest, most constant volume operations, freelancers are a tactical tool for small irregular batches, and a specialised service is the cheapest and lowest-risk choice across the broad middle where growing e-commerce lives.
Run the break-even on your own numbers. If you ship more than a few hundred images a month, or your volume swings with the seasons, the math will point the same way it does for nearly every store that actually does the calculation.
Want the exact number for your catalogue? Send us your monthly volume and image mix for a fully loaded quote, and take a free test batch so you can compare the finished quality and turnaround against your current cost before you commit anything.
Free online photo tools
Edit your photo right in the browser — no install, no signup: