Product Analytics and its Essentiality
Why should we care about product analytics? And how can we implement it?
Working for different startups as an intern, product manager, and co-founder made me understand the importance of product analytics.
Spending money on different aspects of the product’s growth, like in marketing (ATL & BTL), sales, PR, and so on, is worthwhile. But, due to various reasons, many startups try to skip the part of implementing product Analytics in their product. This sets a significant setback to the product’s growth, as there wouldn’t be any tangible observations on users’ behavior to iterate the product (UI/UX/features).
A recent statistic shows that “80% of new users stop using the average app just three days after downloading it!”. Yes, google analytics is the best option available to understand the installation rate and uninstallation rate with other metrics. But, it doesn’t really provide us the reasons for the churn rate we see in our user base.
To address this problem, we use Product Analytics, which allows us to address some critical questions that wouldn’t be answered by the traditional approach to analytics, commonly used for marketing efficiency.
Few points that product analytics helps us in are:
- What is the usage of certain feature?
- What is the purpose of the user to use the product (Mobile / Web app)?
- Identifying marketing channels that are driving the best users.
- The bottlenecks in user onboarding.
- The users’ retention rate.
- Behaviors of the most visiting user.
In regular web analytics, the users’ actions are defined towards conversions of users’ buying a product, registering, filling a survey form or some other activity.
Similarly, analytics in marketing is used to find the # of users landed on a particular page, or more over the goal conversions, under the category of marketing activity performed.
Unlike these, product analytics is tricky and unique for every product. For instance, as a product manager, I want to understand on which category of filter most of my users’ performing actions, I have to set some custom triggers or event points to know the answer.
Before understanding a few sets of Frameworks that almost every Product team usually gauges their analytics outcome, let’s know a few parameters that help conduct this exercise holistically.
Any proposed product analysis includes three areas:
(1) Pilot view of the product state.
(2) Spread of user activities on the app ecosystem.
(3) Deep understanding of specific product feature.
The goal of analysis in each area would be to:
(a) Establish a shared understanding of usage patterns.
(b) Guide discussion towards essential focus areas and actionable improvements.
Two perspectives that are commonly considered in product analytics are:
- Thinking from the perspective of users coming in with different hopes or expectations for the product: How well does the experience match those expectations? Where does it fail?
- Thinking from the perspective of the development team: What is the mission of this app? Does the feature set (and user experience) match that mission?
There are many frameworks adapted by different startups and companies based on their industry and value proposition. Here, I have mentioned some of the most commonly followed frameworks for analytics:
HEART:
- Happiness: User satisfaction level
- Engagement: Engagement time of users’
- Adoption: The users’ adaption towards the product
- Retention: Users’ visiting the product within a week/cohort
- Task success: Efficacy and the efficiency of users’ tasks in the product.
Engines of Growth:
- Sticky Engine: Measure the stickiness of the app. Low stickiness will lead to a high churn rate.
- Virality Engine: To measure the virality coefficient of the product. This will lead to compounding user growth due to referral, word of mouth, or other ways.
- Paid Engine: The growth rate of payments in your product. Linked with stickiness and virality engines.
North Star Framework
- The North Star Framework organizes your product team around one critical metric, the “North Star Metric.” That metric should encapsulate three things:
- A measure of customer value
- A representation of your product strategy
- A leading indicator of revenue
Though Product analytics exposes the raw reality of how people use the product or even a particular feature, it is usually one-dimensional. A holistic way to build the best product possible is to mix the product analytics results with some qualitative feedback from customer surveys/interviews.
When I worked as a product manager, I, with my tech team, thought of trying to adopt Hypothesis-based analytics, commonly used in data science. The objective of this framework was to clearly understand the problem/ product questions and solve them effectively.
The structure goes like this:
- Define the problem
- Create a hypothesis
- Design an experiment
- Carry out analysis
- Develop a theory
Here’s an example of how it might apply in a fitness-based mobile app:
- Problem: many people sign up for a freemium model, but only < 14% convert to paid accounts.
- Hypothesis: Users not able to understand the value proposition of the product within the given period.
- Experiment: Set the freemium model for 14 days.
- Analysis: did >14% convert to paid after an extended trial period?
- Theory: If it’s a Yes, we understood that an extended trial period helps customers know the product and signup for a paid program. If its a No, then retest the Hypothesis with another experiment and/or develop and test a new hypothesis like providing a free nutrition plan for 15 days and the workout program.
Apart from the framework that one can choose to gauge the results on, I found a good blog written on this website that lists metrics that can be found implementing product analytics in the product are:
Installations: Users rarely give apps that struggle to get installed another chance.
App Downloads: Determining an app’s popularity is through measuring the number of times it gets downloaded.
Uninstallation: While your app may get uninstalled for varied reasons, it is prudent to collect details about the specific reason for it being uninstalled.
Subscriptions: Keeping track of all signups, subscriptions, and unsubscriptions are essential.
Registrations: As it figuratively shows that users are buying your product, registration is another factor that needs to be gauged.
Crashes and Glitches: Although apps are susceptible to glitches and crashes, these must not be frequent. Minimizing impacts and arranging customer reviews is essential to ensure better user experience while achieving a reasonable retention rate.
Growth Rate: The tracking rate of growth is essential as it helps understand how a business that uses an app grows and facilitates its growth.
Upgrades: Keeping tabs on which premium packages customers prefer and how long they take for upgrading is necessary.
Rate of Retention: Gaining back those who have once chosen to uninstall your app is necessary, but it can only be done when you can recognize why they left first.
App Session: Another indicator of how many users are following your app while describing the app’s popularity, digitally generating how many sessions users create, is essential.
Length of Session: The time users spent on your app, session length, is relational to uninstalls and the number of crashes; it is essential to keep track of session length for achieving desired actions performed by users.
Session Depth: Study of the number of interactions per session helps understand how the app content could be improved.
Session Interval: Time taken between each session is a KPI that depicts how your app fares among users.
Average Pages per Session: The higher the number of URLs or unique pages that users hit during a visit, the better it will be for the app.
Active Users: Daily Gauging the number of daily users reflects the growth of an app.
Monthly Active Users: Certain apps are required to be used less frequently by their nature- such as financial apps. It is essential to measure this KPI as well.
Brand Awareness: Analyzing the brand’s diversity is crucial, as more awareness of brands translates to profit.
Social Media Shares: Number of shares — how many times an app is being recommended or shared denotes the popularity.
Churn Rate: The rate at which users are seen to leave your app helps arrive at the appropriate updates and improvements.
Per User Revenue: Per User Revenue: The average revenue per app user is indicative of the app’s value as a whole.
Time for First Purchase: The time that a user takes to make the first purchase through your app helps understand users’ needs better.
The crucial financial App Performance metrics:
- Cost of customer acquisition
· Purchases
· Cost per acquisition
· Effective customer acquisition cost
· Lifetime value of customer
· Effective Cost per mile
· Organic conversion rate
· Paid conversion rate
· Cost per install
App Performance metrics related to User experience:
· Devices
· Loading time
· Carriers
· Operating system
· Screen resolution / dimension
· API latency
- Permissions granted
How can we implement Product Analytics in our product?
When I mention implementing product analytics in the product, there are two options that I have used in my work, which are not exhaustive but are enough to give an understanding of the scope.
- To integrate a product analytics tool like Apxor, Mixpanel, Kissmetrics, and others. Using these tools is simple and will save you a lot of time, provided you are in a stage to spend some money.
- The other option is to set up trigger points to measure user behavior at points (Buttons, icons, images). If there were any user activity at those points, the triggered data would help you analyze the user pattern.
The second option is a lengthy task, as to measure all the metrics, you have to carefully understand the different sets of activities or events that a user or a client performs in the app.
Then set up trigger points to measure the population performing/not performing the activity based on these events. Before setting up the trigger points, you have to create an event taxonomy, i.e., how you want to name and record the events in your user flow and your product analytics platform metrics. This would help capture the event types, event properties, and user properties you need to record for analysis.
Thanks for reading the lengthy content. Let me know if you think I have made any misstatements in this writing. These are a few companies providing product analytics tools. Visit their official websites to know more about how product analytics can help your product:
- Apxor
- Mixpanel
- Kissmetrics
Disclaimer
I am not associated with any of the services I use in this article.
I do not consider myself an expert. If you have the feeling that I am missing essential steps or neglected something, feel pointing it out in the comment section or get in touch with me. I am open and happy for constructive input and how to improve.
