We are a small startups and want to understand the benefits of moving from a tool like Firebase Analytics over to something like Mixpanel or Amplitude.
I once heard a concept that Amplitude/Mixpanel saves you needing to hire a data scientist as it has more intelligence built into the analytics.
Based on my research, one difference I am seeing is user-centric tracking in Mixpanel / Amplitude vs. Basic Event tracking in Firebase. I am also wondering about the data storage options/requirements/costs with Mixpanel / Amplitude and the ability to lookback to a historical window.
Has anyone made the switch from Firebase? How was your journey?
The biggest advantage of a toll like Amplitude or Mixpanel or Heap is the ease of tracking user paths/funnels through an app. Secondarily, over the years they’ve incorporated fast, easy ways to easily answer all sorts of complex questions.
Any of those three tools are great, but I also like grabbing events with a platform like Segment - that makes it easy to track once and analyze in multiple places (like databases as well as user analytics)
You can answer a lot of questions with firebase, and possibly everything you need, especially if you have the time and technical understanding.
Amplitude and Mixpanel are great tools, if you already have an event funneling system, like Segment, to nearly instantly understand user life cycles, workflows, drop-off points, and help find the questions you haven’t thought of yet. Mixpanel has the ability to ingest data, but I’ve had good experiences with Segment as a data middle layer.
Heap is more of an all-in-one, removing the need for a Segment, but requiring someone to really live in heap to manage and track data directly.
I haven’t used Google analytics nearly as much as I’ve used the other tools, but whenever I do, the ability to quickly answer a question is better with Amplitude (or any of the other comparables).
Re; data storage options/requirements - my experience with an Amplitude/Mixpanel based stack was that event data went from UIs and our production database into Segment, which then had a data warehouse sink for the events, and Amplitude (then, Mixpanel) pulled event data from this warehouse. Viewing historical data was just a function of filtering based on timestamp/cohort, which is quite easy to do with these tools, and we just didn’t throw data out of the warehouse. Cloud storage is quite cheap, certainly a fraction of the subscription cost of the analytic software that you’d connect to it. At your scale you probably don’t need a large-scale DW, you could just use a vanilla relational database e.g. Postgres.
I think the nice thing about Segment is that you can use it for both your Martech and your product analytics stack.