Ad blockers and tracking loss get discussed as one topic, and treating them as identical overstates the problem considerably. Roughly three in ten internet users run a blocker. Nothing like three in ten of your conversions are disappearing because of it.
The useful question is not the global usage figure. It is how much of your own measurement is actually missing, and that is something you can measure rather than look up.
This is for you if somebody has quoted you a scary percentage and you want to know what it means for your account.
What the usage figures actually say
The most commonly cited numbers put ad blocker usage at 29.5% of internet users globally, 32.5% in the US and 28.5% in the UK, among people aged 16 to 64.
That comes from GWI panel data via DataReportal. Treat it as directional, because the panel composition is not disclosed and self-reported technology use is a difficult thing to survey accurately.
Those figures describe people, not conversions. That distinction is where most of the overstatement enters this topic.
An advert blocked is not a tag blocked. A tag blocked for somebody who was never going to buy costs you nothing measurable. And blockers vary enormously in what they actually stop, from adverts only through to analytics and conversion tracking.
So the headline number is the ceiling of your exposure, not an estimate of it.
Why ad blockers and tracking loss are separate problems
The larger source of missing data is not extensions at all.
Safari and Firefox block third party cookies by default, for every user, without anybody installing anything. That has been true for years and affects a substantial share of most audiences.
Apple’s App Tracking Transparency removed more visibility than blockers ever have. Research by Kraft, Skiera and Koschella, submitted to the FTC, found US trackability fell 55 percentage points, from 73% to 18%, across three proprietary datasets covering billions of impressions in 19 countries.
Consent refusals remove another slice, independent of both.
Blockers are the visible part of a larger picture, and they get disproportionate attention because they involve a deliberate act by an identifiable person. The bigger losses arrived silently through defaults.
Measuring your own exposure to ad blockers and tracking loss
You can get a real number for your own audience in an afternoon, which beats any benchmark.
Record the same action two ways. Fire a browser-based event, and separately record the same event server side, where nothing in the visitor’s browser can interfere. Compare the two counts over a reasonable period.
The gap is your blocked share, measured on your actual audience rather than a global panel.
Most people are surprised in both directions. Technical and developer audiences can run far above the global figure. Older, less technical and mobile-heavy audiences frequently run well below it.
That variation is exactly why borrowing a percentage is the wrong approach. The number that matters is yours and it is obtainable.
The problem nobody can measure
Here is the part that should worry you more than the volume, and it cannot be fixed by any tool.
Ad blocker users are not a random sample of your audience. They skew by age, by technical literacy, by device and by income. So the people missing from your data differ systematically from the people in it.
That means your measured conversion rate, your measured audience preferences and your measured campaign performance all describe a biased sample, and nothing in your reporting indicates which way the bias runs.
Recovering some blocked events with server side tracking narrows the hole. It does not tell you what was inside it, because the recovered events come from people who were blocking at different layers, not a representative slice of the missing group.
So the honest position is that you have a partial view of an unrepresentative subset, and you should hold your conclusions a little more loosely than the decimal places suggest.
What actually helps against ad blockers and tracking loss
Three things, in order of how much they return.
First party data collection at the point of enquiry. If somebody gives you their email on a form, that is yours, and no browser policy takes it away. Everything downstream from there holds up better than anything depending on a tag.
Server side connections where you already have the access. They recover a real share of the browser losses and they carry maintenance cost, so they suit accounts with somebody to own them.
And realistic expectations in reporting. Describing your figures as platform-reported rather than as sales costs nothing and prevents the conversation where somebody discovers the gap for themselves.
What does not help is detection scripts that nag or block visitors running a blocker. That trades real customers for a cleaner metric, which is the wrong direction.
Why this is not getting worse in the way people claim
There is a narrative that measurement is collapsing and it is worth pushing back on gently.
Blocker adoption has been broadly stable rather than climbing steeply. Third party cookies were expected to disappear from Chrome and did not, after Google retired most of Privacy Sandbox in October 2025 and kept them.
The genuine step changes happened years ago, with Apple’s tracking prompt and with Safari and Firefox defaults. Those are absorbed, and accounts have been operating under them for some time.
So the state of ad blockers and tracking today is roughly a continuation rather than a cliff. Anybody selling urgency about an imminent collapse is selling something, and the timeline they are using has already slipped repeatedly.
Steady partial visibility is the normal operating condition. It has been for years and the sky did not fall.
What ad blockers and tracking loss should change about your targets
One practical consequence, and it is about expectations rather than technology.
If a fifth of your conversions are invisible, then your measured conversion rate is lower than your true one, permanently and by roughly that margin. Any target set from that measured figure is being set against a number you already know is understated.
The same applies to comparisons against published benchmarks. Those benchmarks were produced by accounts with their own blind spots, of unknown size, so a comparison is between two partial views rather than between two measurements.
So adjust what you conclude rather than what you count. A campaign underperforming a benchmark by ten percent may be performing identically and measured differently.
And record your blocked share alongside your results, so that when somebody eventually asks why your numbers do not match the sales system, the answer already exists in writing rather than being assembled under pressure.
What to change this week
Three steps.
Run the two-count comparison and get a real figure for your own blocked share, so you stop working from a global panel number.
Then check whether anything in your reporting is described as sales when it is platform-reported conversions, and change the wording.
Then put the effort into first party capture rather than into recovering the last few percent of browser events, because the first is durable and the second is a treadmill.
Third party cookies covers what did and did not change in the browser, server side tracking is the recovery option and its real costs, what conversion numbers actually measure is the counting underneath, and invalid traffic is the opposite problem of counting things that were never people. Organic measurement loses data the same way, which my SEO work accounts for. To size your own exposure properly, book a teardown.