No results found
Try adjusting your search terms

BoxQuiz is a Danish online shop selling puzzles and games. It is run by Bettina Cronquist, who also runs Puslespilsbutikken, a physical puzzle store in the heart of Helsingør. The business runs on Shopify with the Cornerstone theme by Fluorescent Design, and reaches customers through Google, Meta and organic search.
Danish shoppers buying puzzles and games, both puzzle enthusiasts and gift buyers. They arrive through Google Shopping and search, Meta, and organic. These are considered purchases rather than impulse buys, so knowing which channel actually earned a sale is what makes the ad budget defensible.
Bettina came to me through Shopexperts with a specific worry: that her tracking pixels might be at risk of throttling, and that Meta, Google and her analytics might not all be receiving complete data. The goal was to establish what was actually true across the whole tracking and attribution layer, fix whatever was broken, and give her a setup she could trust when deciding where her ad budget goes. The brief was an audit plus the fixes, not a report to act on later.
The store was running two AdNabu apps, Nabu for Google Pixel and Retargeting for Google, to cover Google tracking and retargeting. In practice both duplicated work the native Google and YouTube sales channel already does, so the store was paying for tracking twice and carrying the risk that two tools firing the same event would double-count conversions. Alongside that, the Google tag was flagged for migration with a red alert, Countries and languages sync was switched off in Merchant Center, Customer Match was not running, and the Judge.me review widgets were not showing on the storefront. Nobody had confirmed whether GA4 and the Meta Pixel were actually receiving complete data, so every decision about ad spend rested on numbers no one had checked.
Baseline Metrics(before the project)
I audited the full tracking and attribution layer before changing anything. The pixels turned out to be working and not throttled, so the real problem was elsewhere: a Google tag migration left unfinished, configuration switched off in Merchant Center, and two apps being paid for to do a job the native Google and YouTube sales channel already handles.
The cost-saving step was deliberately held back. Rather than uninstall the two AdNabu apps the moment the native channel was live, I let the new setup run and watched the data for two weeks before greenlighting the removal, so the call that it was safe to remove them was backed by real data rather than an assumption.
Completed the Google tag migration and cleared the red-alert flag
Enabled Countries and languages sync in Merchant Center
Switched on Customer Match with daily auto-sync to Google Ads
Verified GA4 end to end: every event flowing, no data gaps
Verified the Meta Pixel Purchase match quality
Restored the Judge.me review widgets on the storefront
Retired both AdNabu apps once two weeks of live data confirmed the native channel was handling the job
Used Shopify apps and theme
After two weeks of stable live data, the picture was clean and confirmed at the source (verified February 27, 2026). GA4 was recording every event end to end with no gaps. Google Ads was tracking purchases and add-to-carts cleanly, with no missing conversions. The Meta Pixel returned a Purchase match quality of 8.9 out of 10. Every Google tag fired exactly once, with no duplicates. And with the native Google channel confirmed handling the job on its own, both AdNabu apps came off the bill.
For a small retailer that is two wins at once. Fixed costs came down, and the advertising data became trustworthy, so every kroner of ad spend is now measured against a true picture rather than one inflated by duplicate fires.
Baseline Metrics(after)