Every e-commerce founder eventually discovers the same uncomfortable math: a catalog of 60 products, sold across four channels, needs somewhere between 500 and 1,000 image files – each channel with its own dimensions, aspect ratios, background rules, and file-size ceilings. Amazon wants a pure white background and at least 1,000 pixels on the longest side. Etsy crops to 4:3 in search results. Instagram rewards 4:5 portraits. Your own Shopify theme wants squares.
Hiring a designer to produce and maintain that matrix costs more than most bootstrapped brands’ entire marketing budget. Yet the brands that skip it pay a quieter price: mismatched thumbnails, auto-cropped heads, listings suppressed by marketplace image checks, and a storefront that signals “hobby” instead of “business.”
The good news is that product imagery is not a creativity problem. It is a pipeline problem – and pipelines can be systematized by one person with a phone, a window, and a repeatable process. Here is the framework.
The five-stage pipeline
Stage 1 – Capture once, capture wide
The most expensive mistake in DIY product photography is shooting for a single destination. Shoot every product in one session, in the highest resolution your phone offers, with generous space around the subject on all sides. That margin is what lets one master photo become an Amazon square, an Instagram portrait, and a website banner later. Natural window light plus a £15 sweep of white card outperforms most cheap lightbox kits.
Stage 2 – Create one master file per product
Do your color correction, dust removal, and background cleanup exactly once, on one large master image. Every downstream version derives from this file. The moment you find yourself editing a channel-specific copy, your pipeline has broken – fixes now have to be repeated in five places forever.
Stage 3 – Derive channel versions mechanically
This is where solo operators lose entire evenings, and where the right tooling collapses hours into minutes. Batch-processing tools apply one rule – “resize everything to 2,000 × 2,000, pad to square with white” – across an entire folder at once. A free image resizer like this that runs in the browser handles this without software installs or uploads, which matters when the folder in question is your entire unreleased product line. The operating principle: channel versions should be *generated*, never *hand-made*. If producing your Etsy set takes more than ten minutes, the process – not the person – is wrong.
Stage 4 – Name files like a database
`SKU_channel_variant.jpg` – for example `MUG-014_amazon_main.jpg`. Six months from now, when Amazon flags one image or a product gets a redesign, findability is the difference between a two-minute fix and an afternoon of archaeology. Founders consistently rate this the highest-leverage boring habit in the whole pipeline.
Stage 5 – Audit quarterly
Marketplaces change image requirements without much announcement, and themes get updated. A quarterly pass – open each channel, view your listings as a customer, note anything cropped, soft, or inconsistent – keeps entropy from compounding. Put it on the same calendar as your bookkeeping.
The make-vs-buy decision, quantified
| Approach | Upfront cost | Per-catalog-refresh cost | Break-even logic |
| Freelance designer | £0 | £300–£1,500 | Wins only if refreshes are rare and products complex |
| Studio photography service | £25–£60/product | Same again per refresh | Wins for premium brands where imagery is the differentiation |
| DIY pipeline (this framework) | ~£30 gear + a weekend | ~2 hours | Wins for catalogs over ~20 SKUs with regular channel changes |
The strategic insight is that imagery is a recurring cost center, not a one-off project. Channels multiply, specs drift, and catalogs grow – so the compounding advantage goes to whoever drives the marginal cost of a new image version toward zero. A pipeline does that; a designer invoice does not.
Where founders over-invest and under-invest
Over-invested: equipment. Under £100 of gear captures everything a marketplace listing needs; the £800 camera changes almost nothing that a buyer scrolling on a phone can perceive.
Under-invested: the white-background requirement. Amazon’s main-image check is automated – it samples background pixels, and a photographed white wall reads as grey (roughly RGB 230–250) and fails. This single technicality suppresses more DIY listings than any lighting or composition sin. Pure white must be produced digitally, not photographed.
Under-invested: consistency across the grid. Buyers rarely judge one image; they judge the *set*. Twelve products shot at twelve angles with eleven background tones reads as improvisation. One angle, one background, one crop rule – enforced mechanically at Stage 3 – reads as a brand.
The one-person rule
If a step in your visual pipeline cannot be written down as a rule a stranger could follow, it will silently decay the week you get busy. The founders who keep professional-looking catalogs at scale aren’t more artistic – they’ve simply converted taste into a checklist, once, and let batch tooling execute it forever after.
