Short answer: why not use both? Google Analytics 4 and Amplitude were designed to be used by different types of people in the web-analytics space. In turn they emphasize different things. Both are free, in 99% of cases, which is nice – as most site owners do not get past their free quota.
Since the release of Google Analytics 4, I prefer Amplitude for both big and small projects. It's more intuitive, and they've recently released a special marketing dashboard at the time of writing (April 2024). But I also install Google Analytics on sites in most projects because it's free, and performs better in one or two important cases (provided it does done in a privacy friendly way).
In practice I open Amplitude more often than GA4, because it get answers faster. If I am confused by an answer, I can dive into the data to see what’s going on to debug it. In GA4 there is no easy option to dive into the data without a lot of extra configuration; for small website owners this is unlikely to happen.
For some practical examples of the difference in the interface and user experience while asking common question types, jump to here. It should give you a sense of which tool you prefer the 'feel' of.
Background / Context to Understand The ‘GA vs Amplitude’ Question
There are different ways to analyse how people are using your website or app. Google offers a product to do this. It’s called Analytics.
A company called Amplitude also offers a product that does analytics.
Historically Google Analytics and Amplitude were used by different teams to do different things.
Google Analytics was (and is) mostly used by marketing teams and ‘webmasters’ to answer questions such as
- how are people discovering my website – search, search ads, referrals, display ads, social media, email (and so on);
- is my (eg newsletter email / Facebook / TikTok / affiliate) campaign working well;
- how long are people staying on this webpage for;
GA is used by marketing teams to answer hundreds of questions like these.
I use the term ‘webmaster’ above to emphasise that GA has this ‘legacy-web’ ring to it. It was launched in 2005. Back then, one or two people would run everything to do with a website.
Modern data stacks
In modern tech companies, you sometimes have teams of tens of thousands, across marketing, data, development, product and growth all working on ‘the website’ which is the same thing as ‘the business’.
Google’s Analytics product is also used by marketing teams to build re-marketing lists and campaigns – to target and advertise people as they go around the wider web. This is because it natively ties Google’s wider advertising / tracking ecosystem. Here it is useful for what they term 'bottom of the funnel' campaigns – eg abandoned cart adverts – as well as brand awareness campaigns at 'the top of the funnel'.
Many website builders will easily ‘plug into’ Google Analytics, such as WordPress, Wix, SquareSpace, WebFlow and the like. Because it’s been around for so long, and because many website builders have made it very easy to install, it is by far the most used analytics tool on the web. About 60% of all websites use it. In turn Google collect data on almost all the web, and in turn this dataset will influence its product development in other areas (especially machine learning/AI).
Amplitude is much newer. You can tell it's been built for the modern paradigm of the web, apps and digital products more generally. In its brief time around, it’s often used by ‘product’ and ‘growth’ teams in technology-first companies / startups, often to understand how products are being used. It’s far less 'mainstream' than Google Analytics, so congratulations on being here. Knowing about Amplitude in my judgment makes you more savvy than 95% of internet marketers out there. Startups using Amplitude in my experience often have a better sense of what they're doing than sites only using Google Analytics. You can tell in the type of questions they ask.
These product/growth teams tend to ask questions such as ‘what feature on my mobile app is most popular’, ‘what percentage of users are doing ‘X’ (eg adding a song to a playlist) within their first 7 days of downloading the app’, ‘what is the ‘funnel’ between free trials and paying customers’, ‘what sort of traits does this type of customer have’.
Google Analytics sort of provides the answer to these questions, but in a clunky roundabout way, and without the ability to properly interrogate the data (without building complicated stacks on top of it). Some basic examples at the end.
Events now come first
Go forward to 2017 or so – the web and digital products became increasingly ‘event driven’ – which Amplitude was designed for in mind. Google Analytics, written in 2005, was written with the understanding that websites consisted of ‘pages’ and these things called ‘sessions’, which are randomly defined. In the context of mobile apps, mobile browser tabs that are often left open, and an unpredictable user navigation journey, the idea of 'sessions' became kind of stale.
Google Analytics in this newer ‘event driven’ web still had its use cases – such as bucketing traffic into different ‘blocks’ and seeing which of these blocks are really signing up or paying for a product – which is useful for CMOs and CEOs who want the big-level picture.
As of 2021(ish) the two products began to look more similar. Google introduced Google Analytics 4, which is an ‘event-driven’ system, like Amplitude. Amplitude also introduced its own ‘marketing’ tracking, importing metrics that only tools like Google Analytics had (such as utm tags, initial page source, session source). In my view, Amplitude widened its offering to appeal more to internet marketers.
The upside of Google Analytics 3 – the predecessor version that is now phased out in favour of GA4 – was that it was easy to set up, and (sort of) easy to understand for people launching their first website, who didn’t have a background in web analytics and marketing, and wanted to prioritise running their business – not go deep into analytics.
Google Analytics 4 – the latest version that you must use – is hard to use. A story for another time, but in my view this was a deliberate decision to filter out ‘beginners’ who were running low-traffic websites and so costing Google money. Imagine the costs for hosting the analytics for 60% of the internet.
If you’re on this page looking for the difference between Google Analytics and Amplitude, presumably you sit closer to the ‘beginner’ end of the spectrum.
What are the similarities and differences between GA4 & Amplitude?
With this context out of the way, here are the differences & similarities between Google Analytics and Amplitude:
- Google Analytics is mostly installed on the ‘front end’ of the user’s web browsers – although there is a trend, for more savvy businesses, to install it on the server to bypass analytics blockers.
- Amplitude is mostly installed ‘server side’, being used by product and analytics teams (although some teams install it in on the front-end too, such as me, which is helpful for getting marketing-type information such as UTM tags, to understand where particular traffic originated from).
- Google Analytics will happily indigest millions of page views per month and not charge you for it. Amplitude is free up to a point, but at the scale of millions of events monthly, will charge you a license fee which would sit around tens of thousands of dollars annually.
- For small websites, you can happily install Amplitude without paying, like I do for hobby projects such as Flamenco With Rafael – because the site traffic is tiny. In this sense, Google Analytics and Amplitude are absorbing similar data for that site. If the flamenco site was sending tens of thousands of events per month to Amplitude, I’d be asked to pay up. But I am not, and I may never be asked, given it is a small project.
This point about pricing and event volume gets to a philosophical difference between the two tools: Amplitude was originally designed to measure and track identified website users. Say you’ve signed up for an account on an app and then logged in, the app would usually give you a user-ID corresponding to a value in the database. As you go around the app or website, you would track each event – ‘Viewed Page’, ‘Add to Cart’ – and tie that to the user ID. Amplitude was designed with this in mind. For real users. Not random website visitors. And if you were tracking tens of thousands of signed-in-users using your app monthly, you probably have enough money to pay for the license fee.
Google Analytics also supports user-IDs but we’re sure 99% of its 'webmasters' do not get that far; 99% of websites using GA just track “website visits” or “traffic” and have no idea who that “traffic” actually is. Most websites are just pages; not interactive apps.
Realistically most people visiting websites do not identify themselves by logging into an account.
For instance – you’re reading this page right now – but I have no idea who you are. You’re visiting from an IP address, sure, but that is not meaningful information. Just a series of numbers. There’s no user-ID in any database for that on our end. Your visit isn't stored anywhere.
Google will happily eat up all that information and provide big-picture analytics.
Amplitude will likewise happily eat up all that information and provide big picture and very detailed, accessible analytics; but up to a point, afterwards it will cost you money.
I am a fan of Amplitude. And if you’re reading this blog post, aimed at smaller website owners, there’s no doubt that you’ll be able to use it for free. If you have hundreds of thousands of website visitors a month, and want to track all of that, that is a different question. But then you’d probably not be reading this post.
Other things of note
Google Analytics is often installed via Google Tag Manager; Amplitude can also be installed via Google Tag Manager, if you’re installing it on the front-end, as I do with my hobby project, the flamenco guitar course.
Both tools allow custom tracking of events and user IDs. In terms of interrogating the behaviour of a specific user, Amplitude wins hands down. See a practical example video at the end.
Only one tool offers an intuitive user interface (Amplitude); Google Analytics 4 is, in my view, extremely bulky compared to version 3. In fact, it is so bad that my working theory is that Google wants to discourage ‘beginner’ website users from using it. Either that, or they messed up the product’s UI/UX, which for Google is unlikely, but certainly possible.
Google Analytics 4 connects to BigQuery for free, which is cool if you happen to be a data scientist or part of a data team with established data pipelines. Again this is not normal terminology for beginner website users, unless you happen to have worked in a larger technology company as a software engineer or in the data team. For Amplitude, to export to BigQuery, you’d need to buy a license.
None of this is relevant for a beginner, because you’re probably still wondering how to grow based on standard web marketing methods such as SEO.
As this is all rather theoretical, let’s now look at some common beginner questions and see the workflow for answering the question in both tools…
Some practical examples of the difference (shown by images / video)
How many people visited my website in the last week?
Google Analytics:
Here GA4 provides a good, straightforward, report, which is accessible under ‘Reports > Acquisition Report’. It provides a nice breakdown of how users are finding the website on the right (in Flamenco With Rafael’s case, it’s mostly SEO - organic search - driven).
Amplitude
Amplitude Analytics likewise provides an easy graph, and a nice breakdown of traffic by OS / device on the right.
While these graphs look like they’re doing the same thing (and they are) - as soon as you want to customise the graph or ask a different question, you’ll quickly see differences appear..
How many people came to the website last week from the UK?
The same question as above, but with the caveat that we want to focus on UK traffic only. In Google Analytics, I cannot just click that graph and add a filter to it. I have to build a custom report from scratch, which takes over a minute of fiddling, as this video below shows:
In Amplitude, I can add a country filter in 10 seconds:
Where does my site traffic come from?
Google Analytics does a nice job here out of the box, and it still excels here – which is traditionally what it was good for. This is the main reason I still use GA4 today. Automatically Google will put each traffic source into a ‘bucket’ rather intelligently:
Amplitude, on the other hand, does not do this intelligently; you can split out traffic by ‘utm_source’ and neatly distinguish between initial source and session source, but this data is mostly unpopulated because no utm_source has been specified in most cases. For example, organic traffic from search engines is not bucketed as 'organic' – it is'undefined'. Indeed, almost everything is 'undefined' because most of the flamenco website's traffic has no utm_source specified. Google Analytics wins here.
This could certainly be remedied with further configuration, but for a smaller website that needs things to ‘just work’ for a team of of 1, 5, 10 – even 20 – solutions that work instantly are preferable. People have better things to do. Here GA4 has the edge because it knows how to automatically categorise traffic as organic / direct / referral etc.
What is the most common page people land on?
Google Analytics provides this report out of the box:
And Amplitude likewise can be configured to do this, in 10 seconds:
Amplitude provides a performance graph for each page which is pretty neat. GA didn't do that.
But let’s say again we want to just look at traffic in America. In Amplitude this is easily done. Takes 10 seconds (unfortunately the site doesn’t get much US traffic though – something for me to work on):
In Google Analytics 4, doing this took me a full minute, if we deduct the time I took while lost in a different menu:
What percentage of people go onto another page from their first page?
This is an important question to gauge if the website’s ‘funnel’ is working, as well as for people involved in SEO to determine if a webpage is of reasonable quality because of bounce rates...:
Amplitude – easy! Took 5 seconds
In GA4, I had to research how to do this, as the process is less recognisable than in GA3 – but once I got it going, I got it done in 45 seconds. The output looks less nice in Amplitude. I found the logic behind this a bit confusing, less intuitive.
Can I see what a specific device / user is up to?
This type of query is good to get a ‘sense’ of what people are doing and what a typical website journey looks like. Both GA4 & Amplitude allow for this, but note GA4 here (for no reason) originally brought up no results; and then when it did, out of the box it does not provide much information about what page was being visited. Both datasets refer to anonymous users who happen to be me, so I am happy to a share a video in each case:
Amplitude: easy to understand
In GA4 below, you will see the output looks similar, but important data is missing out of the box – it says ‘page view’ but it does not tell me what page was viewed. It does not give me any idea of the user journey, as it is too abstract.
Again, this can be remedied, but for small companies where people have many 'better' things to do, to go digging into GA4’s documentation to understand why custom parameters need to be sent in some instances but not others… is not helpful for the business. It is only helpful for GA4 consultants charging by the hour!
Summary: GA4 vs Amplitude
You can't go wrong in installing both. Amplitude provides a nicer interface which will be preferred by most people. Proper data teams can use both and even join the data together with some fancy custom data stack. Both are free for most people. GA4 is still good for bucketing traffic from different sources (SEO, paid, email); otherwise Amplitude prefers more intelligible answers to most questions.
