Choosing a tool
Looking for a Google Analytics alternative
Eight questions that will pick one for you, and an honest account of where this product fits — including the five cases where the answer is to stay where you are.
The itch
Why people go looking
Six reasons, in the order they usually bite. None of them is that GA4 is a bad product — it is a very good product built for an advertising business, which is not the same thing as a measurement business.
- 01
The consent banner
GA4 writes
_gaand_ga_<id>to the visitor's device, with a two-year default expiry. Writing to the device is what Article 5(3) of the ePrivacy Directive requires consent for, so the banner is not a nicety — it is the price of the cookie. Every visitor meets an interruption before they meet your page, and a meaningful share leave there. - 02
It stopped being readable
Universal Analytics had reports. GA4 has explorations you assemble yourself, and the most common complaint in every review corpus for this category is not price or privacy — it is that nobody on the team can find the number they used to look at on a Monday morning.
- 03
Thresholding and the (other) row
GA4 withholds rows that might identify an individual, and buckets high-cardinality dimensions into a row literally labelled
(other)once it hits an internal limit. Standard reports are unsampled, but explorations sample above a per-query event limit. All three are reasonable engineering decisions and all three mean the number on your screen is not the number that happened. - 04
Fourteen months, and only if you knew to ask
A standard GA4 property keeps event data for 14 months at most and 2 months by default. Raising it is not retroactive, so whatever passed while the setting sat at the default is gone. Most people discover this the first time they try a year-on-year comparison.
- 05
A large share of visits never arrive
Every mainstream blocklist carries
google-analytics.com. On a technical audience that can be more than half your traffic missing, and it is missing silently — the reports look complete. Ours is 117 times smaller at 1270 bytes gzipped against roughly 145 KB, and far less widely blocked, which is why sites that run both usually see more traffic reported here, not less. - 06
The sunset taught everyone a lesson
Universal Analytics stopped processing on 1 July 2023 and the historical data was deleted afterwards. Nobody can import it now — be sceptical of anyone who says they can. The durable effect was not the migration cost; it was a lot of people deciding they would rather their measurement history did not live somewhere it could be switched off.
How to judge one
Eight questions, whichever tool you end up with
We do not rank other people's products here. This is the list we would use if we were buying rather than selling, it names nobody, and one of the eight is a question we answer badly.
-
Q1
Does it write anything to the visitor’s device?
This one question decides whether you need a consent banner, and it is not the same question as "does it collect personal data". A cookie, a
localStorageentry or a device fingerprint all engage Article 5(3) regardless of what is stored. Ask for a yes or a no, then verify it yourself in a network tab. -
Q2
What exactly is metered, and are events included?
The headline number is usually "pageviews". If custom events also draw on the allowance — and they do almost everywhere, including here — then instrumenting your site well is what pushes you up a rung. Work out your real event volume before you pick a plan, not after the first invoice.
-
Q3
What happens on the month you go over?
There are three possible answers and they are very different: the dashboard locks, you are billed for the overage, or nothing happens for a while. Ask which. A launch or a front-page link should not cost you access to the numbers describing it.
-
Q4
How long is data kept, and is a change retroactive?
Retention is the number nobody checks until they need last year. Ask for the figure, ask whether it applies to imported history too, and ask whether raising it recovers what has already passed. Usually it does not.
-
Q5
Can you get your own data out?
An API, an export, and whether both are on the plan you are actually buying rather than the one at the top of the page. This is the question we answer worst — ours is a paid-tier feature, said plainly below — and it is the one most worth asking, because it decides how expensive your next migration is.
-
Q6
Where does the data live — a jurisdiction or a hint?
Most infrastructure offers a location hint, which expresses a preference and constrains nothing about replication. A jurisdiction is a boundary. If EU residency matters to you, the difference between those two words is the whole answer, and vendors use them interchangeably.
-
Q7
Is cross-site tracking switched off, or impossible?
A promise not to do something the system remains perfectly capable of doing is a policy, and policies change with ownership. Ask what the architecture makes unavailable. It is a different question and it gets a different quality of answer.
-
Q8
What does it cost at your traffic, not at the headline?
Pricing in this category is a ladder, and the first rung is the one on the marketing page. Find your own volume on the ladder, add your events, and compare that figure. Then check the annual discount, because most of the category prices annual meaningfully below monthly.
Our answers
The same eight, about us
- Writes to the device?
- Nothing, with two exceptions that belong to the visitor rather than to us: the flag recording an opt-out, and the flag that excludes your own visits. No cookies, no IP address stored anywhere, so no banner is required for our script.
- Metered on
- Pageviews and custom events, added together. Events are not free here and we are not going to imply they are — if you instrument heavily, count your events before you pick a rung. Bot and AI-crawler traffic is filtered out before anything is metered.
- Going over
- Nothing breaks and the dashboard never locks. Three consecutive months over the rung moves you to the one that fits, pro-rated, with notice and one click to decline. Spikes and launches are on us.
- Retention
- 15 months on every plan, on every plan, with nothing to configure — and it applies to imported GA4 history exactly as it applies to what you collect yourself.
- Getting data out
- This is our weakest answer. There is a full query API, and it is on the Business tier only, alongside funnels, custom properties and ecommerce revenue. If an API on the cheapest rung is a requirement, we do not meet it. There is no raw event firehose at any price.
- Where it lives
- A jurisdiction, pinned at signup, with every store created inside it — not a location hint. That distinction is the whole of an EU residency claim and it is why we use the word we use.
- Cross-site tracking
- Arithmetically unavailable, not switched off. The daily token is one-way, scoped to a single site, and rotates — yesterday's cannot be recomputed, and the same person on another site running this product produces a different token. The mechanism is drawn out in full and you can check it from your own browser.
- What it costs
- From $9 a month for 10k pageviews and custom events, after a 30-day trial that takes a card and converts unless you cancel. 1 / 3 / 10 by tier. The whole ladder is on the pricing page — find your own volume rather than reading the first rung.
Side by side
The same questions, scored
Two of these ten rows go to Google, and they are shaded so you can find them without reading all ten.
| Absolutely Analytics | Google Analytics 4 | |
|---|---|---|
| Cookie banner required | No | Yes |
| Script size (gzipped) | 1270 B | ~145 KB |
| On tracker-blocking filter lists | Not currently listed | Listed by default in the major ones |
| Data retention | 15 months on every plan, nothing to configure | 14 months maximum, 2 months by default |
| Time to a useful report | Zero — one screen, no setup | Explorations, custom dimensions, a course |
| Data freshness | Realtime, no processing lag | Up to 48 hours for standard reports |
| Owns your visitors' data | You do | An advertising company |
| Price at 1M pageviews and events | $69/mo | Free — you pay in data |
| Google Ads and Search Console integration | Not available | Native, and genuinely better |
| Cross-product identity across Google properties | Impossible by design | Yes, if that is what you want |
If you buy Google Ads and optimise on conversion imports, keep GA4 for that. Most sites do not, and are paying for it in a consent banner and 145 KB of JavaScript on every page load, in exchange for reports they never open.
Be honest
Five reasons to stay on Google Analytics
Every one of these has been asked for by somebody we then told to stay where they were. If you are in one of them, you are better off on GA4 and we would rather say so here than after you have paid.
-
You optimise Google Ads on conversion imports
Keep GA4. The native Ads and Search Console integration is genuinely better than anything we offer, and bidding on imported conversions is a workflow we do not replace. Plenty of sites run GA4 for the ad account and us for everything else.
-
You need more than the retention window
Ours is 15 months, on every plan, with nothing to configure — a month longer than a standard GA4 property can hold, so a like-for-like switch loses nothing. But if you need a three-year trend, neither tool has it and we are not the answer. Say so before you migrate rather than in March.
-
You want raw events in a warehouse
GA4 exports to BigQuery and you can write SQL against it. We have a query API rather than an event firehose, and it is a paid-tier feature. If your analytics plan starts with "then we join it to the CRM", this is the wrong shape of tool.
-
You want to self-host
There is no self-hosted build and no plan for one. If running it on your own infrastructure is a requirement, say so now — there are tools in this category that do that well and we are not one of them.
-
You need one identity across several products
Signed-in users stitched across apps and properties needs a durable identifier. We do not have one and cannot build one without giving up the thing this product is for. That is a real capability and it is really unavailable here.
If you are leaving
What the switch actually involves
One script tag to start collecting, a read-only connection to pull your GA4 history, and a month with both installed so you can see where the two disagree and why. The import costs nothing on any plan, including the work of recreating your goals.
What comes across and — more usefully — what cannot, is listed in full before you start, along with why your new numbers will not match your old ones. Four separate reasons, all of them real.
What is the best Google Analytics alternative?
There is no single answer, and a vendor who gives you one is answering a different question — their own. The one that decides it for most sites is the first on the list above: does the tool write to the visitor’s device? If it does, you keep the consent banner and you have swapped one tool for another without removing the thing that was costing you visitors. If it does not, you are choosing on reporting, price and how much you trust the numbers.
Is there a free Google Analytics alternative?
There are several, and we are not one of them. There is no free tier here — a 30-day trial on the plan you choose, a card taken up front, and it converts unless you cancel. We would rather say that on this page than have you find it at the checkout.
Free in this category generally means one of three things: open source you host and maintain yourself, a small free allowance on a commercial product, or your visitors’ data as the payment. The third is the one worth reading the terms for.
Will I lose my Google Analytics data if I switch?
No. Removing the GA4 script stops collection; it does not delete anything Google already holds, and your property stays readable. You can also import your GA4 history into your dashboard here at no charge on any plan — what comes across and what cannot is listed in full before you start.
Do I have to remove Google Analytics to try this?
No, and we would rather you did not at first. Run both for 30 to 60 days. You will get a day-by-day comparison with each gap explained — which part is ad blocking, which is bot filtering, which is a metric that is simply defined differently in the two tools. Not knowing whether the new numbers are true is what stops most people switching.
Does switching to a cookieless tool mean I can drop my cookie banner?
Only if nothing else on the page needs one. Our script writes nothing to the device, so it does not require consent — but an advertising pixel, an embedded video player, a chat widget or a tag manager each have their own cookies and their own requirement. Removing one analytics script does not remove a banner that four other scripts are holding up. The reasoning is set out in full in what makes web analytics GDPR-compliant.
How long does the switch actually take?
About two minutes to be collecting — one script tag, 1270 bytes gzipped. The GA4 import runs for up to 48 hours in the background because Google’s API is rate-limited, and it resumes on its own if interrupted. The part that takes real time is the month you spend running both before you trust it.
In this series
Leaving Google Analytics
Whether to switch, what moves across, and how to read the numbers once you have.
- What imports, and what cannotThe full list before you start, not a footnote afterwards.
- What is a good bounce rate?GA4 redefined it, and most advice online describes the old number.
- How to check website trafficYours can be counted exactly. Everyone else’s can only be modelled.
- Your first week after switchingWhat to expect in the seven days people decide whether to trust it.
- What counts as a visitPageviews, visits, visitors and events, defined exactly.
- Why two analytics tools never agreeFive reasons the same site reports different numbers in different tools.
Ready
Try it beside Google Analytics for a month.
Leave GA4 exactly where it is. Paste one 1270-byte script tag, watch both sets of numbers, and decide with evidence rather than with a marketing page.