As open-minded developers, designers and marketers, we are happy to share our thoughts on tools we use every day to bring our users great experiences with Localazy. Within this post, you can read more about why we have decided to use Mixpanel as a part of our analytics stack and how we utilize it.
❓ What is Mixpanel? 🔗
Mixpanel is a pretty popular product analytics tool, and we believe that everyone involved in product management has at least heard about it. The principle is very simple; it allows you to collect user events from different sources through its API or SDKs. Once you collect enough data, you can create powerful reports inside Mixpanel.
Mixpanel constructs these reports in a way that helps you uncover the user journey and answer the most critical questions about your product. It gives you a great insight into who your users are and, most importantly, how they are using your product. Mixpanel is also extremely useful for marketing operations, especially for syncing custom audiences with additional marketing tools.
Mixpanel has five types of standard reports. They are Insights, Funnels, Flows, and Retention. Then there are Signals, which are available only on the enterprise plan, and unfortunately, we haven’t had a chance to try it out yet. All reporting functionality is based upon user cohorts you need to create beforehand. After you’ve designed some reports, you can group them into custom dashboards to track your most important KPIs.
The above screenshot is from Mixpanel Demo project
🤔 What is our experience with Mixpanel? 🔗
Since the beginning of our company, we have wanted to use some product analytics tool. That was early spring of 2020. However, because at that time, we didn’t have any users yet, there weren’t many things we could track. 😄
Also, we could get much more valuable information from different types of user analytics like Fullstory, which helped us understand our initial users and their behavior. We started to get some useful data from Mixpanel around summer 2020, and it was at that point when we began to use it actively.
Why did we decide to use Mixpanel at Localazy? 🔗
We did a bit of research before deciding on Mixpanel, and at that time, we were half the size, and we didn’t have any resources to come up with something overly complex. We needed a tool that is quick to set up and simple to use, so everyone in the team can find the answers or data quickly without any complicated onboarding needed.
Another requirement was the ability to push data through an API and export the cohorts in the same manner. Also, integrations with marketing tools such as Google Ads were necessary. Mixpanel ticked all the boxes, and we decided to try it out.
As a pleasant bonus, we’ve received a special startup deal from Mixpanel, but that is a pretty common sales technique, and all other competitors offer this type of deal as well. We also chose Mixpanel because of its transparent pricing. We simply didn’t want to spend hours with salespeople of competitors listening to countless demos only to find out their solution is too expensive for us in the end.
How we use Mixpanel at Localazy 🔗
We have a couple of different use cases.
- The most common one we use almost every day is tracking our most important metrics and KPIs. So we have a dashboard where we often check whether the metrics are moving in the right direction. We look for spikes, unusual behavior, etc.
- Another use case is very helpful for answering factual questions. Anyone can come into the Mixpanel to create a report or insight and share it with others. This way, you can make more informed decisions when designing the product or prioritizing some features.
- Last but not least, we export some pre-made cohorts of users, which we can use for marketing purposes. We do it either through API or through some of the integrations, for example, Google Ads. You can use these data to improve audience targeting, upsell some features, or retrieve lost users. However, you need to reach particular requirements on the Google Ads side to use this integration.
How Mixpanel helped us to make decisions 🔗
- When we were designing new navigation, we could check the most common paths in the app and make some informed decisions when designing the feature.
- We are continuously improving the product, and with limited resources, we cannot risk developing a feature nobody will use. So what we do, for example, right now is some market testing. We release a dummy feature or an MVP and watch whether users are interacting with it; this way, we can prioritize the development of components. We can focus on things that are in high demand by our users.
- When developing a new feature, we can export a cohort of users who might be potentially interested in it and meet some criteria and send an email to ask if they’re interested in testing it.
- We use selected Mixpanel cohorts to improve the learning curve of Google Ads campaigns, expand the reach, and optimize bids based on gathered data.
The above screenshot is from Mixpanel Demo project
What is the learning curve like? 🔗
There obviously is a learning curve, but it depends on whether you’ve used any product analytics tool before.
- If you have, I’d say that the interface is pretty self-explanatory. Anyone can come in and create the report they need without any complicated onboarding.
- As for the implementation, it definitely needs some planning before you dive into it. You need to realize what you want to track, set up the important events and metrics, and send the related events accordingly.
- You need to read the docs carefully to get it right from the start. For example, it’s necessary to keep property names consistent across events in order to create some types of reports. But the implementation itself from a developer’s point of view is then straightforward.
- If you don’t have much experience with product analytics, Mixpanel has many great resources apart from its documentation to get you on track. These include articles, webinars, white papers, etc.
How soon did we get the value we were looking for? 🔗
With a tool like this, the time to value is usually counted in months.
- First, you need to plan the implementation. What you want to track, what are your metrics etc.
- You need to implement all events in code, which also takes some time but much less than the planning phase.
- Once the implementation is ready; you still need to wait for some time to allow the tool to collect some data. And then you’d probably need to iterate, since most of the time, you’ll find out that you’ve missed something important anyway.
Only after that moment, you’ll get the value, not earlier. So for us, it was something between 2-3 months.
How did we approach the implementation? 🔗
- We started with a tracking plan and an analysis of what metrics we want to track.
- We planned it in a simple docs sheet that was the source of truth for implementation.
- Then we had to determine where each of these events happens. Whether is it possible to send it from the front end? Does the back end need to take care of that?
- Everyone from the team has access to the control panel, and they can use the data or reports whenever they need them.
- We’re constantly adding new events before the new feature release.
It was only a one-time effort to plan the initial implementation, but later on, it’s not time-consuming at all.
What are the best features? 🔗
- Mixpanel’s best feature is the intuitive and clean UI. They did a great job there, and creating complex reports feels like a breeze. I take it as the most significant benefit because anyone from the team can log in and find any data they need.
- Lexicon. It enables you to manage all events coming into Mixpanel without maintaining a separate google sheet manually. Everything is populated automatically. You can merge and rename events in case something is wrong with your implementation. You can add a description and get a great overview of what you’re tracking and why.
- Impact analysis. There are some serious statistics and ML behind these reports. It lets you check whether a concrete feature release had some statistically significant impact on some of the metrics you’re tracking without the need for A/B testing.
- Speaking Marcom, we haven’t been able to use all Mixpanel features such as messaging. But seeing that the guys at Mixpanel have decided to seal the fate of messaging, we are not too disappointed Still, on the other hand, the direct and automated integration with google Ads or Facebook makes Mixpanel a really powerful source of custom audiences. You can use it to acquire new users, support user retention, or generate valuable conversions based on your user’s needs.