Ad blocker conversion tracking is a measurement problem before it is a marketing problem. When a visitor runs a blocker, the browser pixel on your store often cannot report the purchase. The sale still happened. The ad platform just does not know it came from your ad.
This post explains how ad blockers cause missing conversions, why that distorts ROAS and automated bidding, which numbers are actually sourced and which are not, how to measure your own gap, and what server-side tracking can and cannot recover. We are deliberately careful with figures here. Many articles on this topic repeat percentages that no one can trace back to a primary source.
Why conversions are lost
A conversion pixel is a script that loads from an ad platform's domain and then sends the purchase back to it. Ad blockers work from filter lists of known advertising and tracking endpoints. If the script or its requests are on a list, the blocker stops them. The shop page loads normally, the shopper buys, and nothing is reported.
Missing conversions have several browser-side causes. Blockers are one of them, and it helps to keep them apart:
| Cause | What happens | Fixable on the server? |
|---|---|---|
| Ad blockers | Pixel script or its requests are blocked | Yes, for the purchase |
| Browser tracking prevention | Cookies and storage written by scripts expire sooner | Partly. See the Safari ITP post. |
| Declined consent | Marketing pixels are not allowed to run | No. A declined visitor stays declined. |
| Closed tab or off-site payment | Shopper never reaches the thank-you page | Yes |
What the sourced data says
The most useful primary source we could verify is the 2023 eyeo Ad-Filtering Report, the successor to the PageFair Adblock Report, now published through Blockthrough. Here is what it states, and what it does not.
- By Q2 2023 the report estimates 912 million active ad-blocking users worldwide, across desktop and mobile. That is a count of users and devices, not a share of your visitors.
- Its ad block rate reference table, built from traffic on sites that use Blockthrough's detection, lists rates by country and device type. For Germany it shows 32% on desktop and 17% on other devices. For Austria it shows 22% on desktop and 15% on other devices.
Read those numbers with care. The rates come from anonymised traffic on ad-supported sites, and the detection works by checking whether requests to bait URLs are interrupted. A webshop's audience differs from a publisher's audience. A blocked ad is also not the same as a blocked conversion pixel, since not every blocker rule set targets the same endpoints. The figures show that ad blocking is common. They do not tell you how many of your purchases went unreported.
Sources: 2023 eyeo Ad-Filtering Report (PDF) and the Blockthrough report page. We do not quote a blanket "percent of conversions lost" figure, because we have not found a primary source that supports one for e-commerce purchases.
Three assumptions worth dropping
- "My pixel reports conversions, so blockers are not a problem." You only see the conversions that were reported. Blocked ones leave no trace in the ad platform, so the platform view cannot tell you how many there were. Only your order data can.
- "This only affects desktop." The same report estimates over 495 million monthly active users of mobile ad blocking and over 416 million on desktop at the end of Q2 2023. Mobile is not exempt.
- "Server-side tracking sidesteps consent." It does not. A declined visitor stays declined, and forwarding is gated on consent.
Ad blockers ROAS impact: the mechanics
ROAS is conversion value divided by ad spend. Ad blockers act on the top of that fraction. When purchases go unreported, the platform sees less value for the same spend, and the reported ROAS drops.
A purely hypothetical example, to show the arithmetic and not a benchmark: say a campaign drives 100 orders worth 50 each, for a spend of 2,500. The true ROAS is 2.0. If the platform only receives 80 of those purchases, it reports a ROAS of 1.6. Nothing about the ads changed.
There is a second effect. Automated bidding learns from the conversions it sees. If it sees only part of the picture, and that part is skewed toward visitors who do not run blockers, it optimises for that audience. That may not match the people who actually buy from you.
Measure your own gap
Do not rely on an industry figure to size your own problem. You can estimate it from your own data in under an hour.
- Pick a recent period of at least a few weeks, with stable campaigns.
- Export orders from your store. Exclude test orders and cancelled orders.
- Export purchases the ad platform reports for the same period, using the same attribution setting.
- Run a test purchase in a clean browser and another with an ad blocker enabled. If only the first reaches the platform, you have confirmed the leak on your own pixel.
The remaining difference includes more than blockers. Attribution windows, view-through credit and orders from other channels all widen it. The post on why Meta shows fewer conversions than Shopify walks through separating those causes. For a quick look at which tags your storefront loads, use the free tracking audit.
If the gap turns out to be small, you have your answer and can move on to other work. If it is large, treat that as a measurement project, not a campaign problem: fix the data first, then judge the ads.
What server-side tracking fixes, and what it does not
With server-side tracking, the purchase is sent from your store's order data to the ad platform through a relay, so a blocker in the shopper's browser has nothing to block. SyncBeacon does this for Shopify, WooCommerce and Shopware orders, then forwards to Meta, Google Ads and the other launch destinations.
What it recovers
- Purchases from shoppers whose pixel was blocked, whose tab closed early or who never reached the thank-you page.
- Deduplication, so a purchase that both the pixel and the server report counts once. Meta documents the requirement in its deduplication guide. The
event_idmust match on both sides.
What it does not recover
- Visitors who declined marketing consent. SyncBeacon stores the event for your records but does not forward it.
- Click IDs when the optional browser script is blocked. The script is still a script. Purchases come from the store, but without a click ID Google Ads may have nothing to match unless hashed customer data is present.
- Ad views and browser events such as page views, which still depend on the browser.
- Any promise of a specific uplift. We do not state one, and you should be wary of anyone who does without showing their method.
Destination detail: Meta Conversions API and Google Ads Enhanced Conversions. If you want the longer argument on how this looks for Google Ads, see our Google Ads conversions post.
Related reading
- Server-side tracking: the overview
- Safari ITP and ecommerce tracking
- Shopify conversion tracking not working: a diagnostic checklist
- How server-side conversion tracking works
Get started
Close your conversion gap today
Connect Meta CAPI, Google Enhanced Conversions, and more - in one setup, no developer required.
Figures are quoted from the cited report and describe ad-supported websites, not your store. The ROAS example is hypothetical arithmetic. Not legal advice.