Landing page builders are one of the most common places I find broken attribution. A brand builds a dedicated page in Replo or Shogun for a Meta campaign, conversion rate looks great in the builder, and a month later the attribution tool shows that campaign underperforming. Usually nothing is wrong with the campaign. Something between the ad click and the order lost the tracking. The good news is that the fixes are mostly a checklist you can run in an afternoon.
Why landing pages break tracking
Both Replo and Shogun publish pages into your Shopify store, so in a standard setup the page loads with your theme’s tracking and the Shopify pixels you already have. Problems show up at the edges. Pages served on a separate domain or subdomain. Redirects that drop UTM parameters on the way. A/B tests that send visitors through a redirect to a variant URL. Custom add-to-cart buttons that skip the events your pixels listen for. Heavy pages with large images or video that delay pixel scripts long enough for fast bouncers to never register.
Any one of these can make a page look worse in attribution reports than it really is, because the order still happens but the click that started it no longer connects to it.
The checklist
Start by opening the live page from an actual ad preview link, not from the builder. Use Meta Pixel Helper, Google Tag Assistant and your attribution tool’s debug or live view to confirm a page view fires. Check the address bar after the page loads. If the UTM parameters and click IDs like fbclid or gclid are missing, something in the path is stripping them, and that is your first fix.
Next, click the page’s add-to-cart or buy button and confirm an add-to-cart event fires. Custom buttons built in the page editor sometimes add the product to cart without triggering the standard event. Then go through checkout and place a test order, using a discount code that brings the total close to zero. Confirm the purchase shows up in your attribution tool with the right source and campaign. If it shows up as direct, the click was lost somewhere between landing and checkout.
Finally, check speed. Run the page through PageSpeed Insights on mobile. If the largest element takes several seconds to load, some visitors will leave before any script fires, and those sessions vanish from every report. I covered the tracking side of speed in attribution tools and site speed.
If you run pages outside the theme
Some brands host landing pages on a separate domain or a headless frontend. That works, but every tracking script has to be installed there too, and cookies set on one domain won’t carry to another by default. Keep the landing page and checkout on the same root domain wherever possible. If you are already headless, the setup issues are covered in attribution on headless Shopify, and checkout-specific tracking is in checkout extensibility and attribution tracking.
Comparing landing pages fairly
Once tracking works, compare pages on orders and revenue from the attribution tool, not on the builder’s own conversion metric. Builders tend to count a conversion when a visitor clicks to cart or reaches checkout, which is useful for design testing but says nothing about whether the visitor bought. Run each page against the same ad set and audience where you can, so the page is the only real difference. Shopify’s sessions by landing page report gives a rough conversion rate per page as a sanity check.
Where ThoughtMetric fits
ThoughtMetric installs on your Shopify store and tracks visits through to orders with its own pixel and server-side tracking, so purchases tie back to the campaign that sent the visitor even when browser-side tracking is spotty. (Disclosure: ThoughtMetric sponsors this site.) It reports by channel and campaign with multi-touch attribution, which makes it a reasonable place to compare a campaign before and after you switch its landing page. Pricing is based on monthly pageviews, starting at $99 a month for 50,000 pageviews, with every feature on every plan. It will not fix a page that strips UTMs or sits on a different domain. Run the checklist first.
A common failure I see
The pattern that catches brands most often is an A/B test that redirects. The ad sends visitors to page A, the testing tool sends half of them to page B through a redirect, and the redirect drops the click ID. Page B looks like it converts worse because half its buyers show up as direct. If your test tool redirects, confirm it passes query strings through. If it doesn’t, test by duplicating the ad and pointing each copy at its own page instead.
Picking an approach
- New landing page going live? Open it from an ad preview and confirm UTMs and click IDs survive.
- Using custom buttons? Confirm add-to-cart events fire before you spend on traffic.
- Running A/B tests with redirects? Make sure query strings pass through, or split tests at the ad level.
- Judging pages on the builder’s conversion rate? Compare orders and revenue in your attribution tool instead.
- Want campaign-level attribution with server-side tracking behind your pages? ThoughtMetric (our sponsor) covers that.
Leave a Reply