Selling on marketplaces like Amazon or bol.com sounds simple.

Until you actually start doing it.

In a recent video, Stephan Spijkers and Chris Jobse (founders of PIMvendors) sat down with Luc Smeets (Managing Director at KatanaPIM) to talk about what’s really happening in ecommerce right now.

One thing became very clear: marketplaces bring opportunity, but also complexity. And without a solid PIM strategy, that complexity quickly gets out of control.

Marketplaces are just the new “shop-in-shop”

Marketplaces aren’t new.

They’re just the digital version of what department stores used to be.

As Stephan Spijkers put it, they act as a “digital shop-in-shop.” A place where brands get a podium to sell, usually on a commission basis.

Sounds great, right?

But here’s the catch: every marketplace has its own rules.

Different attributes.
Different languages.
Different image requirements.

And suddenly, managing product data becomes a full-time job.

As Chris Jobse explains:
"You get a proliferation very quickly if you are also managing that alongside your own channels"

The classic mistake: fixing the wrong problem first

When companies scale to multiple channels, they often start with a channel integrator.

Makes sense. You want to connect everything.

But here’s where it goes wrong.

A channel integrator can move data.
It can’t fix bad data.

Luc Smeets sees this all the time:
"We often saw that a channel integrator was implemented first, and only then did people start thinking about a PIM"

And that’s where things break.

Because if your data isn’t right, marketplaces won’t forgive you.

Products won’t go live.
Or worse, you risk getting banned.

That’s the reality of “garbage in, garbage out.”

Not all product data is equal

If you want to win on marketplaces, your product data needs to level up.

The experts describe it in three layers:

1. The baseline
Make sure all required data is there. No missing fields. No “yellow or red cards.”

2. Enrichment
Add enough detail so your products can actually be found.

3. Optimisation
Tailor your content per channel. Improve SEO. Adapt to AI-driven search.

And this only works if your data structure is solid from the start.

As Chris Jobse says:
"You want to record your data at the most granular level. From there, you can merge things again according to the requirements of your various output channels"

Think: flexible, reusable data instead of fixed, messy fields.

Your tech stack shouldn’t do everything

Another big shift: composable commerce.

Instead of one big system trying to do it all, you connect the best tools for each job.

PIM.
OMS.
WMS.

Each tool has its role.

And PIM? That’s your foundation.

Even with AI becoming more important, that doesn’t change.

If anything, it makes PIM more critical.

As Luc Smeets puts it:
"All those tools... are very handy and can do a lot, but they need data... if you work with products, make sure you have your data in order, because then it also becomes much easier to be the director or conductor of that orchestra of different tools"

So where do you start?

You don’t need to overcomplicate it.

Start small.

Many companies begin with one marketplace. Often by selling overstock. Just to learn how it works.

Then scale.

But the goal stays the same: one central place for your product data.

A single source of truth.

Because the future is moving fast. Think AI agents buying from other AI agents.

And in that world, your product data is everything.

As Stephan Spijkers concludes:

If you want a consistent and strong omnichannel experience, you need to manage that complexity at the core.

That core is your PIM.

💡 A PIM helps you centralize, enrich, and distribute product data from one place — making it easier to scale across marketplaces without losing control

Try it now or book a demo!