If you wish to develop your analytics strategy the right way, I suggest following these 4 steps:
Start by asking yourself a set of questions and answering them honestly.
How do you monetize? How do you deliver value to the user? What’s the crossover? What information do you need to grow your product?
Usually, this will be a metric that captures two things at once: the moment you create value and how frequently it happens.
Examples here may include: Daily Active Users, Weekly Active Buyers, or Monthly Active Subscribers.
You should also develop a set of more granular metrics, each corresponding to a specific lifecycle stage.
A typical funnel looks like this: reach ⇒ activation ⇒ engagement ⇒ retention.
Mixpanel is a great tool for doing just that. Take a look at their framework:
Mixpanel’s Guide to Product Metrics
What are the steps a user needs to take to perform key, value-driving actions? You could have a registration flow, a sign-in flow, a trial upgrade flow (if you’re a SaaS), and so on.
Look at all the steps in the product flows. These are your events.
Listen to our live session for more details—I promise you’re gonna like it!
Now that we’ve got product strategy covered, let’s jump to the core of this article: the tools.
A quick look at G2Crowd will tell you that there are about 35 product analytics tools on the market. In reality, though? You should double that number.
And if you also include web analytics tools, which are very often grouped in the category—triple it.
Most of the tools out there are subscription-based with fairly similar features and reporting abilities. But it is possible for you to group them. There are a couple ways to go about it:
You’ll pay each time a user visits your software product or website. Depending on the nature of your product and frequency of use, sessions can stack up quickly. A very popular tool that bills by sessions is Heap Analytics.
During a session, users can perform many actions. Each tracked action is an event. So while sessions don’t matter here, the intensity of product use can make or break your wallet.
Amplitude is perhaps the most notable tool that uses pricing-by-event. It also comes with a really generous free plan of 10,000,000 (!) monthly events.
This way of billing is perfect for products with a smaller user base and an intense pattern of use.
You’ll pay for each user who’s accessed your product or website (anonymous and identified). But in this case, you won’t need to bother with the volume of sessions or events that users generate.
A popular tool that uses this kind of billing is Mixpanel.
When it comes to data collection, there are 2 schools of thought:
In this case, you need to pre-define the data points, a.k.a. the events you want to track. You hardcode those events into a product to start sending out information to the tool.
You can add more events at any time. But you’ll start collecting the data only from the time of implementation.
Mixpanel and Amplitude are both classic analytics tools in this sense. They require a developer to properly set them up and maintain—especially if your product changes a lot.
Personally, I think this approach is very cool and really useful. Tools like Heap will track all events happening from the moment you place a snippet of code in your product. You then define the events yourself. In-tool, without a developer.
And to make the deal even sweeter, you’ll be able to define new events anytime and get all the historical data about those events.
Now you get why it’s retroactive!
Okay, this isn’t necessarily a separate grouping like the two previous ones listed. But for the sake of comparing the tools, it’s good to know what features to look at.
Sessions, engagement, events, retention, goals, conversions, social engagements, ecommerce, etc.
Geolocation, demographics, stitching.
Real-time, cohorts, device unification, trends, funnels, dashboards.
A/B testing, notifications, integrations, alerts, levels of access, quality assurance.
All of those will come in handy in the next section.
I’ll be honest with you, I don’t necessarily have a framework for choosing the best product analytics tool. It’s more of a structured Google spreadsheet. A modified version of Matthew Brandt’s, in fact.
It gathers business information about the tools you’re considering and your business requirements—pretty straightforward.
I think you might find it quite useful for your own analytics needs, so I’d like to share it with you. Head over to the end of this article to get the link to my spreadsheet!
To make things easier for you, I’ve already filled out the details of 3 products we mentioned earlier: Mixpanel, Amplitude, and Heap Analytics.
And here’s how this spreadsheet looks:
Product analytics tools spreadsheet
Now that you have a nice spreadsheet ready to go, you’ve got to do your part and start collecting your business requirements.
What exactly do I mean by that?
In short, you’ve got to go around and talk to each one of your stakeholders. Ask them what they want and be ready to end up with a massive wishlist.
This is where the more difficult part comes in: you’ll have to strip that list down to the absolute must-haves. Features, reports, and integrations they can’t live without.
What has helped me with getting to the essence of our stakeholders’ needs was using the “5 whys.” Rings a bell? It’s part of the Toyota Production System (TPS). They were Agile before it was cool!
The “5 whys” is exactly what it seems—you ask “why” 5 times. Doing it may feel a little weird, but I promise you, it actually works. If you don’t believe me, check out Buffer’s great piece on the “5 whys.”
It’s a really cool city, but that’s not what we’re talking about here. Or Metro 2033, for that matter.
Once you’ve collected all the requirements in the previous step, you should try and fit them into MoSCoW.
Let me explain…
MoSCoW is a pretty popular prioritization framework used in management. I’m not going to give you a lengthy definition (you can look it up on Wikipedia). Simply put, it’s an acronym that stands for: “Must have,” “Should have,” “Could have,” and “Won’t have.”
These are your categories for requirements.
Make sure that the product analytics tools you’re considering check all the must-have boxes. Ideally, they should also include most of the should-haves and a couple of could-haves.
That’s the long and short of it. Thanks for reading my article on the product analytics landscape!
Remember that data is the key to making educated business decisions. Without it, you’re lost and you’ll never be able to realize your product’s full potential.
I’ve shared with you what I think are the best analytics tools for the job, but the rest is up to you. I’d say this article can serve as a good roadmap for you to start with, though.
I hope it’ll serve you as well as it has me. Good luck with your analytics and data-driven decisions!