Case study · Google Ads
Feeding Google Ads cleaner conversions
with server-side tracking
We measured BooleanMaths Google conversions import against GA4, Google's native Shopify App and Triple Whale.
Across all of them BooleanMaths' server-side conversion import fed Google Ads more complete, higher-quality purchase data than the native pixel or GA4 or Triple Whale. Recovering up to 82% more revenue signal, and producing the only conversion action Google rated "data quality: excellent."
+82%
more conversion value vs the data feed from GA4
+16.7%
more conversion value vs Google's native Shopify App
+5.6%
more conversions tracked vs Triple Whale
Excellent
BM is the only Purchase data where Google Ads has no quality issues
In short
BooleanMaths sends purchases to Google Ads as a server-side "Import from clicks" conversion action.
In our case studies for three live D2C accounts against the native Google pixel, GA4 & Triple Whale, Booleanmaths' server-side feed recorded up to 82% more conversion value, beat Triple Whale head-to-head, and was the only Purchase action Google rated "Active - data quality excellent."
Cleaner conversion data that leads to Smart Bidding optimising on a truer picture of what actually sold.
The gap
The pixel undercounts. The bidder underperforms.
In the Shopify ecosystem, the purchase signal gets broken because of three reasons.
Purchases route through third-party checkouts, which are not matched to Shopify sessions perfectly. Any tracker relying on Shopify data alone loses ~10% of conversions.
Ad-blockers, privacy browsers & cookie restrictions. Any pixel that fires on the Thank-You page misses out on ~30% of conversions.
Final purchase happens on a different channel & browser - leading to a new visitor Id getting generated. Any conversion tracking without identity stitching loses ~20% of conversions.
As a result, the data that reaches Google Ads through the browser pixel or GA4 is a partial, lower-quality record of what actually sold.
Google's own diagnostics say so. On the native paths, enhanced conversions and basket data get flagged "Needs attention" - and the conversion counts Ad algorithms optimise on falls short of reality.
What we feed the account
One clean, server-side conversion action.
BooleanMaths writes a server-side "Import from clicks" conversion action straight into Google Ads for every purchase, with order value, enhanced-conversion match data, and consent state intact. These conversions come from a enriched user-journeys powered by 3rd Party Checkout data recovery and an identity graph.
The Setup
Case Study setup
We measured this for 3 different D2C Brands. All on Shopify. All using BooleanMaths in parallel with another data feed to Google Ads.
25 Orders / day
High AOV - Luxury Brand. Replacing an existing Triple Whale data feed with BooleanMaths
100 Orders / day
Medium sized Brand. Replacing a broken data-feed with BooleanMaths server side tracking.
1000 Orders / day
Results - Data Quality
Google trusts one feed. It flags the rest.
These are the actual Google Ads diagnostics for two Purchase conversion actions in the same account, tracking the same orders. The difference isn't only the volume - it's whether Google can trust the data.
Google Ads conversion diagnostics — server-side feed vs the native GA4 web action, same account.
Results - more visibility
What the cleaner feed actually captures.
Figures are the actual conversions and conversion value each source recorded in Google Ads over a four-day window, inside a single account. The comparison is between conversion actions tracking the same orders.
Brand A
Head-to-head · Google Ads · 4-day window
A rare direct benchmark: BooleanMaths and Triple Whale running side by side in one account. BM recorded the most purchases of any source ₹4.86L / 75.4 conv - ahead of Triple Whale (₹4.65L / 71.4).
Brand B
Recovered value · Google Ads · 4-day window
Over the same four days, the BooleanMaths server-side feed recorded far more purchase signal than the GA4 web pixel running in the same account - signal that the ads algorithm never received through the browser.
₹44.0L vs ₹24.1L through the GA4 pixel +82% conversion value and +69% conversions, same store, same four-day window.
Brand C
Data health · Google Ads · 4-day window
The Brand had tried to increase the captured data volume using GA4 and the Google Pixel with some success, but that resulted in a loss of data quality.
Hence the volumes sit close together, but the quality is different. Every native path tracking the same orders is flagged "Needs attention". The BooleanMaths feed is the only Purchase action Google rates "Active" while also feeding higher conversions (+4.6%).
Why it matters for bidding
Smart Bidding can only spend toward what it sees.
01
Undercounted conversions, underbidding
Smart Bidding spends toward the conversions it can see. Feed it a fraction of real purchases and it pulls budget off the audiences that actually convert.
02
Missing order value, wrong tROAS
Value-based bidding needs the basket. When order value is absent or flagged, target-ROAS optimises against a distorted picture of what each click is worth.
03
Weak match data, lost attribution
With enhanced conversions off, Google matches fewer clicks to sales, so revenue goes unattributed and unlearnable. Server-side import fixes all three at the source.
Methodology. Figures are the actual conversions and conversion value recorded by each conversion action inside a single Google Ads account over a four-day window, tracking the same orders. Values in ₹ (account currency); conversion counts include Google's fractional attribution. The brand is withheld. TW = Triple Whale. BM = BooleanMaths
