Listing on Bol.com without losing your mind

Bol.com is, by some distance, the most important marketplace in the Netherlands. It is also the most opinionated. Listings get rejected for reasons the dashboard does not explain. Ranking moves for reasons the algorithm does not publish. This is the playbook we use across our customer base.

1. EANs are non-negotiable

Bol's catalogue is keyed on EAN. No EAN, no listing. Two different products with the same EAN, no listing. A wrong EAN, even a typo, can get your listing merged with someone else's product. Once merged, splitting is a support ticket that takes weeks.

Rules: every variant gets its own EAN. EANs live in your PIM, not in a sheet. Every listing payload validates the EAN against Bol's catalogue before it ships.

2. Content blocks beat description length

Bol scores content quality by block count, not by character count. A 2000-character paragraph counts as one block. A 200-character block followed by a bullet list followed by a comparison table counts as three.

What we ship for every product:

1. A 1–2 sentence hook (what is it, who is it for)

2. A 3–5 bullet feature list

3. A specification table mapped from PIM attributes

4. A use-case block if applicable

5. An "in the box" block if there are multiple items

That's five blocks per listing, generated from structured data, in Dutch and Belgian Dutch. Done in a workflow, not by hand.

3. The title is a structured field, not a sentence

Bol's title format is Brand + Product type + Key attribute + Variant. Not "Beautiful soft cotton t-shirt for summer days." Buyers search by brand and category. The algorithm rewards titles that match that pattern.

Example. Don't write: "Stylish men's leather watch with brown strap."

Do write: "Casio Mannenhorloge Leer Bruin 42mm."

4. Images: square, white, exactly to spec

Bol requires:

• Minimum 500×500, recommended 1000×1000, square aspect ratio

• White background on the hero image

• JPG or PNG, under 4MB

• Product fills 80% of the frame

• No text, watermarks, or logos burned in

You can ship gallery images that break those rules, but the hero must pass. If your DAM doesn't auto-generate a Bol-compliant hero from a master asset, every product launch becomes a manual photoshop session.

5. Categorisation matters more than you think

Putting your product in the wrong Bol category isn't a "minor labelling issue." It changes which buyers see it, which attribute fields are required, and which competitors you rank against. We've watched listings 5x their conversion just by moving from a generic parent category to the right leaf.

The trick is mapping your PIM taxonomy to Bol's category tree explicitly, not "auto-detected from the title." Map once, then ship from the map.

6. Pricing pressure is real

Bol uses the buy box logic familiar from Amazon: lowest qualified offer wins by default, but content quality, fulfilment speed and seller rating all weight the win. You will not win on price alone against a Chinese drop-shipper. You can win on content + fulfilment + rating, even at a premium.

7. Validation before publish, every time

The number-one reason a product disappears from Bol is a validation failure that nobody noticed. Image too small. EAN duplicate. Required attribute missing. Prohibited claim in the description.

Don't push to Bol and find out later. Run a pre-publish gate against Bol's full rule set. If it fails, route it to whoever can fix it. If it passes, ship it. Auto-publish what passes, queue what doesn't.

8. Sell-through, fed back

Bol's seller API exposes per-listing orders, returns and conversion. Don't let that data die in the marketplace dashboard. Pull it back into your PXM so you can see which listings are working and which content patterns convert. The catalogue should learn from its own performance.

The TL;DRBol rewards structure: EANs, titles in their format, content in blocks, images to spec, categorised to the leaf, validated before publish. Get those right and you'll outrank competitors with bigger budgets.

None of this is impossible by hand. It's just slow, error-prone, and impossible to scale past a few hundred SKUs. Which is what the platform is for.