Plausible: The more accurate alternative to Google Analytics
“Every number in your Google Analytics account is wrong.” That is exactly what an independent study by Orbit Media found.
This happens for several reasons: consent banner declines, spam and bot traffic, tracking scripts blocked by privacy-first browsers and ad blockers. Google Analytics has no solid solution to any of them. Site owners who switch to Plausible consistently see higher numbers than they ever did in GA.
Here’s what we do differently.
- Why does accurate data matter?
- What Google Analytics gets wrong
- Cookie consent banner dependency
- Spam and bot traffic skewing data
- Tracking script being blocked by ad blockers and browsers
- Location inaccuracy
- Conversion attribution inaccuracy
- Misclassified Direct traffic
- Complex traffic exclusion settings
- Delayed data processing
- Time on page tracking inaccuracy
- Complicated setup contributing to setup errors
- Data modeling and sampling
- In conclusion
Why does accurate data matter?
Second-guessing the numbers on an analytics dashboard defeats the purpose of having them. When data is underreported or inaccurate, the effects compound: GA misattributes acquisition channels, marketing budgets get allocated against skewed numbers, agencies show clients figures that don’t match the backend and publishers make content decisions based on audience data that’s missing a significant share of visitors.
The bigger the scale, the bigger the cost of getting it wrong.
P.S. If you’ve ever wondered why analytics numbers differ between tools, this post explains why.
What Google Analytics gets wrong
Cookie consent banner dependency
Implementing a GA tracking code on your website requires a cookie consent banner. This is legally required in many regions to comply with privacy regulations like the EU’s GDPR and California’s CCPA.
Many website visitors don’t want their activities to be tracked and want to protect their privacy. So they decline the banners. This causes the tracking script to be disabled.
This gap leads to underreporting in traffic numbers and incomplete data on user interactions, skewing analytics insights.
In fact, according to the Orbit Media study shared above, the accuracy of GA4 as compared to Plausible was exactly at 55.6% due to the cookie consent banner’s presence. More than half the data was missing, making Google Analytics stats unreliable.
You can estimate the size of your own data gap with our cookie banner traffic loss calculator.
How do we tackle this at Plausible? We made Plausible tracking completely cookieless. This means that we track your website usage data without cookies or any other persistent identifiers.
Neither do we ever track or store your visitors’ data for any reason, making us inherently privacy-friendly. This means you don’t have to use a cookie consent banner on your website just because you’re using Plausible.
No consent banners means no declines and no missing data in the first place.
Spam and bot traffic skewing data
GA4 excludes known bots automatically. Its detection stops there, and a significant volume of spam and bot traffic still gets through.
At Plausible, we detect and automatically exclude bots by:
- Blocking traffic based on the User-Agent header
- Filtering out known referrer spam domains
- Blocking traffic originating from data centers
- Detecting and excluding unnatural traffic patterns
Plausible also excludes ~32,000 data center IP ranges by default. GA4 does not.
In a test we conducted, we simulated bot traffic to a website and observed that GA4 failed to detect it, displaying the bots as real users. Plausible correctly identified and excluded them from the stats.
GA4 users constantly deal with skewed data because of this. For example, when a site uses a cookie management platform (CMP), the CMP might send data center traffic that GA4 is unable to detect and exclude.
So GA users need to manually find out which IP addresses are being used by the CMP and exclude them within the GA4 settings. You may end up excluding 10 IP addresses based on the help guides or direct information available from the CMP, before another one attacks your stats.
And all this depends on noticing the problem in the first place, then wading through jargon and support guides to fix it. All you wanted was to see real site stats.
Plausible excludes 32k IP ranges, bots, crawlers, referrer spam and other known non-human activity from the get-go. In a test we ran, we saw 18x more pageviews with server-side tracking than with Plausible, which shows the volume of spam traffic that Plausible excludes.
Tracking script being blocked by ad blockers and browsers
Google Analytics is infamous for leaving cookies on users’ browsers and devices. Privacy-conscious users block it using ad blockers and privacy-first browsers like Safari and Firefox.
Apple’s Intelligent Tracking Prevention, Firefox’s Enhanced Tracking Protection and Brave all block it specifically because of how GA tracks users.
If your users use privacy-friendly browsers or ad blockers, Google Analytics script won’t be able to track their traffic on your website.
Plausible’s position: Being cookieless and privacy-first means we’re not blocked by privacy-friendly browsers or most ad blockers in the first place.
One exception: ad blockers that make no distinction between privacy-friendly and invasive scripts, blocking all JavaScript tracking.
For those cases, we offer a proxy option that bypasses the block. This means accurate data even for visitors with ad blockers enabled.
Setup takes a couple of minutes. No technical knowledge required.
Location inaccuracy
Users connecting through proxy servers or VPNs can conceal their true location, leading Google Analytics to log the location of the proxy or VPN server rather than the user’s actual location.
City-level reporting is also less accurate within GA4. This usually happens due to IP anonymization measures (since IP addresses are personally identifiable information) across the globe.
Google Analytics changes the last digit of the IP address to zero when these addresses are sent to the GA4 servers in the US.
What Plausible does: We group VPN traffic under a single “Anonymous VPN Service” category. This ensures that Location reports are more accurate and reduces the noise that VPN users might otherwise add.
We also use the MaxMind database to determine the visitor location based on the IP address. This lookup happens on our servers and no data is ever sent to MaxMind.
The IP address itself is discarded to make sure we don’t store any personal data. We never store IP addresses in our database, logs or anywhere on disk at all.
Conversion attribution inaccuracy
Google Analytics makes you manually exclude unwanted referral sources like payment processors to keep your data clean.
If not, you’ll see traffic sources like PayPal, Stripe or Paddle as a source of the conversion instead of the original channel.
This happens when users briefly leave for payments or password resets, causing GA4 to start a new session.
On top of it, this has to be done within a cluttered settings panel.
But Plausible handles this automatically, keeping your original traffic source and session intact. No extra setup needed.
With Plausible, referral sources are counted only when they start a new session on your site. This prevents external domains such as payment gateways showing up in the list of sources.
One makes you work to get accurate data, the other just works.
Misclassified Direct traffic
In Google Analytics, traffic from mobile apps like Gmail, Slack and Telegram is often misclassified as “direct,” leaving you with unclear data.
Plausible’s algorithms recognize traffic from Android apps and categorize it under “android-app” in the “Sources” tab. This can help recover up to 10% of previously unrecognized traffic for sites with large mobile audiences.
Complex traffic exclusion settings
In GA, it’s possible to exclude internal traffic from polluting your analytics by excluding the IP addresses. If you wanted to exclude a country’s traffic from being recorded in GA4, there’s no setting available and the workarounds are:
- Add IP ranges of an entire country to your IP block list. (not reliable)
- Filter, segment and use Comparisons to exclude countries while viewing reports. (complicated)
- Get a developer to do it. (unnecessary)
In Plausible: you can exclude traffic by IP address, country, page/section or hostname. No developer needed: it’s a dropdown menu.
Delayed data processing
Google Analytics can take up to 48 hours to fully process data and update your reports. This generally means a delay in making important decisions, or worse: not knowing that the data may not be fresh and making misinformed decisions.
Plausible’s data is always fresh and constantly updated. Real-time analytics, no waiting.
Time on page tracking inaccuracy
The average engagement time metrics in Google Analytics 4 can be misleading since their calculations don’t include the time from sessions that lasted below 10 seconds.
In our experiment, we found that such so-called “non-engaged sessions” can be half of sessions, causing the final metric to underreport by as much as 80%.
How Plausible calculates it:
- We track active engagement only when the page is in “focus”. So we don’t count the time a visitor spent away from your website, say by switching tabs or switching apps.
- We include time on page for bounced visits (i.e. visits where only a single page was viewed).
- We count all time spent, even very short visits, down to a single second. This is the major difference in calculating time on page in Plausible from GA4.
Complicated setup contributing to setup errors
It is quite complicated to set up a well-functioning GA dashboard, which is why there are so many expensive courses on the web trying to teach GA.
It usually requires learning Google Tag Manager, Google’s terminology, character limits, how to configure settings correctly, DebugView and often working with a developer just to get reports working.
So it’s easier to miss out on essential settings or do something wrong than it is to get the setup right.
The Plausible approach: The dashboard is very simple to set up and understand.
Because Plausible handles many things by default, the setup is minimal. This leaves little room for error and ensures all you see is real and accurate data.
Data modeling and sampling
Google Analytics actively shows modeled key events. Quoting them: “Google uses modeling to estimate online key events that can’t be observed directly. Modeling allows for accurate attribution without identifying users.”
GA4 has been including paid and organic channel modeled key events since the end of July 2021. Event-scoped dimensions also include modeled data.
They do this to cover the data gaps stemming from rejected cookie banners, browsers limiting the time window for first-party cookies, ad interaction and the key event happening on different devices.
You cannot differentiate modeled data from your regular data.
The conversion numbers aren’t final either. It takes up to 12 days for conversion data to process and stabilize after each conversion is initially recorded. Why? Because GA4 is also feeding this data into attribution models (for example, to determine which channels contribute most to conversions).
Much of what GA reports is estimated, not measured. Here’s a full breakdown.
Plausible does not use data sampling or modeling by default. We collect and store 100% of the data regardless of how many pageviews you have. The stats in your dashboard reflect 100% accurate data of what happens on your site.
However, on some dashboard views that have a lot of data such as those with more than 10 million pageviews, we apply limited data sampling to help the dashboard load as fast as possible.
In conclusion
If you want accurate data without filtering bots yourself, managing consent banners or second-guessing whether the numbers are real, Plausible is built specifically for that.
More accurate data means more confident decisions. That accuracy doesn’t require asking your visitors to trade their privacy for it.
Try Plausible free: no setup overhead, no consent banners, no estimated numbers. You can also follow our GA4 to Plausible migration guide.
On WordPress? Install our official plugin.
