Ecommerce Attribution: How to Know Which Channels Actually Drive Sales

A customer sees a TikTok ad, doesn't click, forgets about it, gets an email three days later, and finally buys from a Google search for your brand name. Which channel gets the credit? Most ecommerce reporting gives it entirely to the last one — the branded search — and quietly ignores the ad that actually started the whole thing.

Why ecommerce attribution is its own problem

Ecommerce brands usually run several acquisition channels at once — paid social, search, email, sometimes influencer or affiliate — and Shopify (or whatever platform you're on) can only tell you the order happened, plus whichever single source touched it last. If you're relying on platform-native reporting alone, every order gets attributed to one channel, and every other channel involved disappears from the picture.

Shopify's own attribution is explicit about this scope: sales attributed to marketing are the ones traceable to campaigns run through Shopify's own Marketing section, or external efforts tagged with UTM parameters. Anything untagged, or anything that happened on a different platform earlier in the journey, simply isn't counted.

A realistic ecommerce journey

What actually needs to be true for this to work

1. Every campaign link needs a UTM, including email and organic social

Paid platforms often self-report clicks automatically, but email links, bio links, and organic social posts usually don't unless you tag them yourself. See UTM tracking for the specific parameters and a naming convention that won't fall apart once three people are tagging campaigns.

2. Returning visits need to be tied back to the same customer

A customer who sees an ad on their phone and buys on their laptop two days later is still one customer. Without session stitching, that's two anonymous, disconnected visits — and the platform that gets "credit" is whichever one happened to be the device they finally checked out on.

3. Repeat customers need journey history, not just their most recent order

For subscription or repeat-purchase ecommerce specifically, the channel that earned a second or third order might be completely different from the one that earned the first. Attribution that only looks at "this order's last click" misses that a customer's lifetime value is often the result of several different channels doing different jobs over time.

Reading channel roles, not just channel totals

Once journeys are visible, the useful question for an ecommerce brand isn't just "which channel drove the most revenue" — it's which channels start journeys versus which ones close them. A prospecting campaign that never shows up as the last click before purchase can still be the reason a large share of your customers exist in the first place. Cutting it because its last-click numbers look weak is one of the most common — and most avoidable — budget mistakes in paid ecommerce.

As you add more channels, this compounds

With one paid channel, attribution barely matters — there's nothing to compare it against. The moment you add a second, third, or fourth channel, every additional one increases the odds that customers touch more than one before buying, and last-click reporting gets proportionally less accurate. This is exactly the pattern customer journey analytics and multi-touch attribution are built to handle — seeing the whole sequence instead of collapsing it down to one line per order.