Product analytics is the process of collecting data on the use of digital products to make decisions on product optimization, the core of a SaaS. In the previous post, we talked about how to implement a product measurement strategy, and today we present an analysis of the best product analytics tools for C-Level to carry out your measurement strategy successfully thanks to the best product analytics tools that we show you.
In this post I bring you a comparison of the most used SaaS product analytics tools, with their pros and cons.
Top product analytics tools for product manager or product owners
What are the most popular digital product analytics tools today?
Here are the product analytics tools that I have used and seen the most, by SaaS companies:
- Google Tag Manager
- Google Analytics 4
- Data collection, data processing and storage.
- Data analysis.
Product analytics tools comparison: Collection and processing of user events
To track the behavior of our users, it is necessary to define the product analytics strategy, the behaviors we want to measure, as we saw in this previous post on the basics of product analytics.
We can capture these events using our own code, and store them in our own database or data lake, or use third-party tools.
For example, in this image of Posthog’s value proposition, we can see an example of implementation with our own code versus their solution:
Product analytics code vs service
So let´s start with product analytics tools comparison:
I advise to focus on existing product analytics solutions, instead of doing a whole implementation of your own, so you can spend your development team’s time 100% on delivering value to the customer.
We can also track these events with Google Tag Manager and send the information to Analytics or if we use Firebase, we can also use their tracking. I have seen a lot of implementations like this, and in the Next Scenario tool it was the first implementation.
At the beginning, don’t get scratched, put focus where it contributes the most or outsource this part, don’t have a developer several months with this.
If we want something more advanced, we will have to use an event streaming tool.
Any tool will provide you with a tracking code to insert in your own code, but there are tools that are focused exclusively on capturing and distributing this data from your own products, such as Segment or Rudderstack.
Streaming tools for user events
Event streaming tools are in charge of capturing and distributing the data generated in your service by your users.
The most commonly analytics tools to capture user events are Segment and RudderStack.
Product streaming events
Segment is currently the most widely used event streaming tools. It offers a wide range of features and integrations to track user behavior, aggregate other data sources to send to other services or repositories.
It has a free price of up to 1000 visitors per month, which is a bit short.
But it also has perks for startups of $25,000. You can apply here: https://segment.com/industry/startups/
One of the shortcomings of segment is the lack of data curation, and if you are not very careful you can fill your database with erroneous records, which you will then have to clean by hand, if it has happened to you, you know what I mean, hell.
RudderStack also offers similar capabilities, but is not as widely used as Segment, as its foundation is also much more recent.
One of the most significant features is that it is also open source and you can host it on your machines, physical or cloud, with the advantages that you control the data processing, and the disadvantages, which we know has to host something, even if it is in a cloud.
Rudestack scales very well and supports a lot of event flow. It has a standard docker implementation, in Digital Ocean, which is one-click installation, with a load balancer to scale on demand.
If you don’t know what this means: If there are many users, it scales alone.
The configuration is done in the cloud, even if you use the self-hosted version. So you don’t have to worry about configuration persistence, so it makes administration easier.
Segment and Rutterstack allow you to use custom tracking code or pre-built integrations like Google Analytics and Mixpanel.
Data processing and analytics tools for Saas business
Once we have collected the data, it is necessary to process it to extract insights. For this, there are several options available:
- Google Analytics
Mixpanel and Amplitude are two of the most popular data analysis tools. They offer a wide range of features for measuring user behavior, such as cohorts and funnels, allowing you to track user journeys across platforms.
It is the leader and the reference. At UX/UI level it is a step ahead of the others, it is a very powerful and complex tool.
The set up of Amplitude is not trivial, and it takes quite some time to get it right, the best thing to do here is to hire someone with experience.
It has a free part that includes:
- 10M events per month
- Basic analytics
The problem is that when you get volume, as it can happen with this kind of tools, it is very expensive. They also have a startup program that lasts about 2 years.
When you get volume, Amplitude + Segment is usually around 100,000 dollars per year.
The most important limitations of the free layer of Amplitude are:
- You cannot save cohorts for easy reuse, but you can extract and save them outside the tool.
- You cannot apply readjustments to the data schemas. For example you cannot rename events and properties. This is useful in early phases, because you are testing and changing. This is why in the previous post we talked about the importance of making a good measurement plan.
Mixpanel is simpler, but enough to early stage startups ands it´s much less complex.
Before it was very good, but it has been lagging a little behind in product. The price is much lower than Amplitude for similar functionalities.
These are their plans:
Posthog is an open source analytic tool, built to capture and store data in a database similar to the previous ones.
It provides features such as user behavior tracking, event tagging, user segmentation. It also allows you to integrate with different platforms and applications.
Posthog allows you to have a high performance database and have first party data. You can have clickhouse behind and well-structured kubernetes opens doors to then take that data to a CDP.
As cons its interface is much rougher.
Product Analytics 4 / Firebase
Analytics 4 or Firebase (I think they are going to unify it in GA4) will get you out of the way in early moments, there are even people who have done cool things with Firebase, since it provides integration with Jsons and Google Cloud services.
It is a good solution for the beginning if you implement it for login or as a user database.
Yes, Hubspot can be used for some of the product analytics. In recent times they have developed an all in one suite, which can pick up your sales, marketing and product strategy.
It is a very complete product, which has been put the batteries in the part of events in app. If you don’t want to have a tool for everything, this is your solution. Although in fine product analysis it may still be limited. It is very useful to follow the journey from lead to user and even customer.
Hubspot can help you solve the attribution problem. By having the same platform from the time we collect the lead or the first user contact, it is much easier to attribute channels. You can then pull this info into a data lake or CDP and generate cohorts.
I personally spoke with Rafa Ceo Trainning Gym, a Saas for trainers and gyms, leader in its category, with great MRR, which went from using more than 20 applications, to centralize it only in one. Hubspot made a use case for it:
Hubspot is loved and hated in equal parts. You also have perks for startups of 90% reduction in the first few years.
Conclusions on top product analytics tools comparation
In conclusion, there are many options for product analysis tools. You should consider which one is best suited to your needs and your budget.
My advice, don’t get married to just one tool and separate analysis from data collection.
In short, Segment or Rudderstack cloud, for events. Don’t mess with in house or selfhost implementations for testing. Amplitude or Mixpanel free and you can use G Analytics 4 as additional backup.
And you know, if you don’t want to access several applications at the same time, you can use Next Scenario. I leave you here the link:
Thanks for reading this post and glad to know your experience with these tools and if you need help in your analytics strategy, you know you can count on us.