Conversion is not a single metric: what everyone responsible for traffic should take away from this CRO conversation
Expert commentary on the “O e-commerce” podcast episode with Ewa Wysocka and Michał Czajkowski (DigiSpot)
I listened to Ewa’s podcast episode with Michał Czajkowski, and one line stayed with me long after: you can raise your conversion rate very easily – just cut cold traffic from your campaigns. It sounds like a joke, but it’s one of the most important lessons for anyone who judges marketing by its results.
In my day-to-day work, I focus on performance marketing and brand growth, working with a range of e-commerce brands – including on tribe47 projects. That’s why I look at an online store from the side its traffic comes from. Michał talks about the user journey from the product listing to checkout. I bring in the other half of the equation: who we let onto that journey in the first place, and with what intent.
Because conversion doesn’t happen in a vacuum. It starts long before a user ever sees a product page.
Let’s start with hard data. Ewa opens the episode with the e-commerce equation: revenue = traffic × conversion rate × average order value. These three elements are interconnected, so judging any one of them in isolation leads to poor decisions.
Market data makes this very clear. According to Contentsquare’s 2025 benchmarks, paid social traffic converted at around 0.4%, while paid search traffic converted at 2.0–2.1%. That’s more than a fivefold difference within paid channels alone. Meanwhile, Triple Whale data (covering 33,000+ DTC brands) shows a conversion rate of 3.9% on desktop versus 1.8% on mobile.
Same stores, same products – yet completely different results depending on where customers come from and which device they use.
What changes from a performance perspective
Michał puts it plainly in the podcast: a global conversion rate is misleading and, on its own, tells you very little. I fully agree – and I’d add that from a campaign perspective it can be downright dangerous, because it encourages you to optimize for the metric rather than for the business.
Here’s what I see when working with campaign data:
- Cold and warm traffic serve two different goals. Prospecting campaigns build awareness and acquire new customers, while bottom-of-the-funnel campaigns close the sale. Comparing their conversion rates one-to-one is like comparing a first conversation with a client to signing the contract. If you cut cold traffic today because it “hurts conversion”, you may well be hurting your long-term growth.
- The traffic source determines what you see in your store. Michał gave the example of a company with “great SEO” whose traffic landed mostly on the blog – one of the weakest-converting sections of the site. There was traffic, but no sales. Before you start rebuilding your product page, check whether your campaigns and content are sending people to places where they can actually buy.
- Attribution is the foundation, not an add-on. The episode makes an important point: some testing tools rely on GA4 attribution, which distorts the picture in itself. Michał’s team pulls data directly into BigQuery. From my perspective, working with a solid analytics tool (or a well-built spreadsheet 🙂) is the standard worth aiming for: if you don’t know which channel actually drives sales, every budget decision is guesswork.
Micro-conversions: where performance meets UX
The most interesting part of the conversation is about micro-conversions. Michał describes a furniture store where the biggest game-changer wasn’t in the cart or at checkout, but in the step from the product listing to the product page. In another case, as many as 70% of users dropped off at the fifth step of a product configurator.
For performance marketers, that’s a very concrete takeaway. If a campaign delivers high-quality traffic but users get lost on the listing page, the problem isn’t the targeting. And vice versa: even the best-designed journey won’t hold up if the people landing on it have no purchase intent. That’s why UX insights and campaign data belong at the same table.
Michał also highlights the question of scale: a continuous A/B testing program makes sense at around 4,000–5,000 orders per month in the store itself (excluding marketplaces). At a smaller scale, an audit with a list of quick wins works better. This matters for ad budgets too. Not every conversion uplift has to be “bought” through testing – sometimes all it takes is getting your traffic in order.
My three tips to get started:
- Segment before you judge. Before you conclude that “conversion has dropped”, break it down by traffic source, new vs. returning users, and device. Very often it turns out the store is performing well – it’s only the traffic mix that has changed.
- Measure each campaign against the goal it’s meant to achieve. Different campaign types can have different objectives: a new-customer acquisition campaign may run at a higher CPA – and so may Reels on Meta – while a static promotional ad should drive sales at a high ROAS. Applying one KPI to every campaign is a fast track to burning through your budget or, worse still, to stalling your growth.
- Keep seasonality and context in mind. Michał illustrates this vividly: a collaboration that kicks off right before Black Week will bring success “on its own”. Always compare campaign results with the equivalent period and with what was happening in the market – not with the previous week.
Conversion isn’t a single number. It’s the result of an entire ecosystem working together: traffic, the purchase journey and the offer. Data over intuition – but with human context.
#CRO #ConversionRateOptimization #PerformanceMarketing #GrowthMarketing #UXDesign #ABTesting #Ecommerce #Ecommerce2026 #DataDriven #DigitalMarketing
Natalia Kobierska
Performance & Growth Expert, tribe47
Sources
- “O e-commerce” podcast, S02, episode 41: UI/UX and CRO with Michał Czajkowski (DigiSpot): Spotify | YouTube
- Contentsquare, 2025 benchmarks (Q4 2025), via MarketingCharts: paid social conversion rate of approx. 0.4% – Best For Ecommerce: Social Share of Traffic
- Contentsquare, 2025 benchmarks (Q4 2025), via MarketingCharts: paid search conversion rate of 2.0–2.1% – Best For Ecommerce: Paid Search Share of Traffic
- Triple Whale, 2025 (33,000+ DTC brands): desktop 3.9% vs. mobile 1.8% – PodVector: Ecommerce Benchmarks 2026
