For many e-commerce brands, a significant share of revenue does not convert on their own website. It lands on a marketplace: Amazon, OTTO, eBay, Zalando, Kaufland, or one of the many regional platforms that dominate their category, Allegro across Poland and Central Europe, Bol in the Benelux, Cdiscount in France. And yet the measurement infrastructure most brands rely on – multi-touch attribution, Marketing Mix Modelling, incrementality testing – is built around owned channels. Marketplaces sit outside it.
That creates a measurement gap that rarely shows up on the roadmap, but quietly distorts every budget decision that depends on it. All the while, external marketing budgets run across paid social, search, video, and beyond, driving demand that lands somewhere the brand cannot see.
A problem that grows as brands grow.
Many of today’s strongest e-commerce brands didn’t start with their own shop. They started on Amazon, and for a while, that made measurement simple. One platform, one conversion point, one set of numbers.
Anker built a billion-dollar business that way: in 2016, Amazon accounted for roughly 80% of its revenue. By 2021, that share had dropped to around 54% (Marketplace Pulse, 2021), as the brand expanded into its own DTC sites, additional marketplaces, and offline retail. Snocks followed a similar path, winning Amazon’s Sales Partner of the Year in 2019, and growing to €83 million in revenue by 2024 across Amazon, OTTO, Zalando, About You and their own shop.
But as brands grow, so does measurement complexity. Which campaigns drive customers to Amazon? Which ones bring them to the website? That is a measurement problem that scales with success. And it’s worth asking whether your setup is keeping up.
Is this relevant for you? A quick self-check.
This affects any brand that sells on third-party marketplaces, in addition to, or instead of, their own online shop or offline stores. If customers can discover your brand through your marketing and then buy on a platform you do not control, the conversion happens outside your measurement.
Before going further, it is worth asking:
☐ Sales distribution: Do you generate sales on marketplaces like Amazon, OTTO, eBay, Zalando, or Kaufland, in addition to, or instead of, your own online shop?
☐ Revenue volume: Do you generate a significant absolute volume of sales or revenue on marketplaces? Even a relatively small share can represent a meaningful amount worth optimising, if the absolute numbers are substantial.
☐ Performance visibility: When you look at your marketing dashboards today, do they include marketplace sales as an outcome? Or do they only reflect what converts on your own domain?
☐ Budget decisions: When Finance asks which channels are driving revenue, can you give an answer that includes marketplace sales, or does your response quietly exclude the channel that generates the most volume?
If any of these questions gave you pause, you likely have a marketplace attribution gap. You are spending on marketing that influences marketplace sales, and you have no reliable way to measure, defend, or optimize that investment.
Why getting numbers from the marketplace is not enough
Most major marketplaces offer some form of measurement product, e.g. Amazon Attribution generates tracking tags for external campaigns and shows which clicks led to purchases on Amazon. It is free, and for brands just beginning to think about this, it is a meaningful first step.
But it has a fundamental limitation, and understanding it matters.
Marketplace attribution tools only measure what the platform can see: the direct click from a tagged campaign to a conversion. A customer who sees your Meta ad on Monday, searches your brand name organically on Amazon on Thursday, and converts on Friday does not appear in your report. Neither does the YouTube campaign that created initial awareness, nor the influencer post that triggered the search.
“The marketplace tells you what happened inside its walls. It cannot tell you what drove people there.”
This creates the same structural problem as any self-attribution setup: the platform delivers the ad and measures the outcome. The result is a clean report that reflects what the marketplace can track, not what your marketing actually accomplished.
In practice, this means:
- Channels with a direct, trackable click path (branded search, retargeting) appear highly efficient
- Upper-funnel channels (video, display, influencer, TV) appear to contribute nothing to marketplace sales
- Budget decisions follow the available signal, gradually shifting toward demand capture and away from the channels that create demand in the first place
The consequence is predictable: growth slows, but the dashboards keep looking fine. By the time the pattern is visible, the misallocation has been running for quarters.
How to close the marketplace gap
The approach that works combines two layers. The first is the granular touchpoint data that already exists in Exactag’s MTA: every digital interaction across all online channels, measured independently from the platforms. The second is a Marketing Mix Model that takes marketplace sales or revenue as the actual output variable (depending on what matters most to your business), and uses those MTA-based channel signals as inputs.
The result is not a generic MMM that treats marketing spend as a black box, but a model that can attribute marketplace sales down to individual channels and tactics, informed by real journey data, not just aggregated spend figures. This is what Exactag’s Marketplace Attribution does.
In practice, this makes three things possible:
- See the full picture: quantify the contribution of every external channel to marketplace sales, including video, influencer, and display, which leave no direct click trail
- Distinguish demand creation from demand capture: understand which channels are building incremental demand versus which ones are simply intercepting purchases that would have happened anyway
- Defend upper-funnel investment with numbers: when Finance asks what is driving revenue, give an answer that includes the channel generating the most volume
The model runs on a monthly cadence, which means it functions as a steering tool, not just a retrospective. Budget decisions are informed by current performance, not last year’s planning assumptions.
And because Exactag operates independently from any media platform, the results carry no structural bias. There is no incentive to overstate the value of any particular channel, just a neutral read of what is actually driving your marketplace sales.
Ready to explore your hidden potential?
If a meaningful share of your revenue lands on Amazon, OTTO, or any other marketplace and your current measurement does not reflect that, we are happy to take a look at your setup together and show you what becomes visible when marketplace sales are part of the picture.
FAQ Marketplace Attribution
What exactly is marketplace attribution, and why is it different from standard attribution?
Standard attribution tracks customer journeys that end on your own website. Marketplace attribution solves a different problem: understanding which of your external marketing activities, paid social, search, video, influencer, drove customers to buy on a platform you don’t own or control, like Amazon or OTTO. Because the conversion happens outside your domain, standard tracking tools cannot follow the journey. Marketplace attribution uses a combination of multi-touch data and Marketing Mix Modelling to close that gap.
We already use Amazon Attribution. Isn’t that enough?
Standard attribution tracks customer journeys that end on your own website. Marketplace attribution solves a different problem: understanding which of your external marketing activities, paid social, search, video, influencer, drove customers to buy on a platform you don’t own or control, like Amazon or OTTO. Because the conversion happens outside your domain, standard tracking tools cannot follow the journey. Marketplace attribution uses a combination of multi-touch data and Marketing Mix Modelling to close that gap.
Does this only apply to Amazon, or to other marketplaces too?
It applies to any marketplace where your customers convert: Amazon, OTTO, eBay, Zalando, Kaufland, and regional platforms like Allegro, Bol, or Cdiscount. The measurement problem is the same wherever the conversion happens outside your own domain. Exactag’s Marketplace Attribution can model against sales or revenue data from any marketplace, not just Amazon.
How does the model actually work if it can’t track individual users on Amazon?
It doesn’t rely on individual user tracking. Instead, it uses Marketing Mix Modelling with marketplace sales or revenue as the output variable and your external marketing activity as the inputs. The model analyses statistical patterns over time to determine which channels explain movements in marketplace revenue, separating the effects of paid media from organic demand, seasonality, competitor activity, and pricing. The result is a channel-level view of what is actually driving sales, without needing to follow individual users across platform boundaries.
Which channels become visible that weren’t visible before?
Any channel that drives demand without leaving a direct click trail. That includes TV and video campaigns, display advertising, influencer activity, and organic search that was triggered by awareness you created. These channels are structurally invisible to click-based tools like Amazon Attribution. With MMM, their contribution to marketplace sales can be quantified and defended.
How is this different from what Exactag already does for online shops?
For brands selling through their own shop, Exactag’s MTA tracks the full customer journey at touchpoint level. For marketplace-first brands, the conversion happens outside that tracking environment. Marketplace Attribution extends the framework by integrating marketplace sales as the output variable in the MMM layer, so the same neutrality and independence that applies to owned-channel measurement also applies to marketplace revenue. The two approaches can be run in parallel for brands that sell through both.
