Analytics for Product Managers

Product Managers at growing startups need to define clear targets that relate to the product vision. Data-driven decision making has worked well for Amazon, Facebook, Google, et al., so there’s little doubt about its effectiveness at least at certain points in a project lifecycle.

It’s easy to feel over-whelmed by the task of analyzing scattered data. So following some established methodologies and patterns is useful.

Here are a few terms you need to understand, before integrating product analytics into your plan:

  • Vanity Metrics
  • Metrics and KPIs
  • One Metric That Matters (OMTM)
  • Segmentation
  • Funnels OR User Journey
  • Retention
  • Data Points

1. Vanity Metrics

Metrics such as page views or historical total downloads don’t always convey actionable measurements. They might make you feel positive about growth but think hard whether you want to base your strategy on these. Every product is different! These are often, unfortunately used by startup founders to show big numbers that aren’t truly meaningful.

2. Metrics and KPIs

“A Metric is a quantifiable measure that is used to track the status of a specific process.”
“A KPI is a specific metric that your organization or your department has identified as one that closely tracks or predicts performance.”

For instance, your advertising spend is a metric, not a KPI. Advertising might help you get new users but the money spent is just a measure. A KPI could associate the sales increase with the advertising spend, i.e. cost of customer acquisition could be a KPI. Likewise, page view alone won’t give you a clear idea of engagement. You need to be evaluating user actions to know the actual engagement. To sum up, all metrics are not KPIs.

3. “One Metric That Matters”

The most important metric for your startup. So you put all your resources to move this forward. Helps keep the operations focused and gives everyone a single theme to chase. Of course your goals will change with time, so your OMTM (One Metric That Matters) will change.

4. Segmenting Analytics based on Attributes

Segmenting divides analytics based on attributes. This helps in observing patterns more minutely instead of being lost due to averaging. If properly planned, as your product grows, you would gain increasing ability to ask complex questions from your data. For instance, what browser or device your users are using more, which region or countries see more visitors, what is the age group of your most frequent visitors, etc.

5. Funnels OR User Journey

Users perform a series of actions when using a product — measuring the key events at each step of the user journey could help you figure out important usage patterns for your product. For example, at what particular point a user leaves the signup process etc.

6. Retention

Retention is how often your customers return and use your application. It’s important to know how the changes to the product affect the usage. This could involve Cohorts Analysis: “Cohorts are simply groups of customers that started using your app within a defined period of time”. So If you can report the number of app users between the age of 20-30 today compared to last week or last month, that would give you a good idea of growth.

7. Data Points

A Data point is a discrete unit of information and is roughly equivalent to datum. Expressed in another way, it’s a measurement that can be represented numerically or graphically. Here’s the more detailed wikipedia definition.

Setting up Analytics for a Web/Mobile Product

There are three broad steps you need to perform to setup analytics for your product:

Step 1: Plan

  1. Product Vision – Start with defining the vision for the product that you are building. What problem are you going to solve? Which KPIs will indicate performance?
  2. KPIs – Define the measurements that predict critical areas of performance.
  3. Metrics – List all the metrics that might be useful for the product.
  4. Data Points – Further drill down on where and how you’ll collect the metrics.
  5. Funnels – Plan out how you’ll track the user journey through your product.

Step 2: Measure

Avoid the temptation to build analytics and BI tools internally at least at an early stage. Keep your focus on the core product instead of being penny-wise and trying to save money on external tools.

  1. Set up the tools to collect the metrics you have planned. Checkout Google Analytics, Kissmetric, Mixpanel, Flurry etc.
  2. You might want to set up separate databases to run regular queries.

Step 3: Report

Presenting your analysis properly is as important. Few things that you can consider generating reports for (this would depend on your product):

  1. Comparisons with competing products
  2. Growth in number of users, revenue, average revenue per user etc.
  3. Demographics, Maps, trends in your products state etc.

Of course, there’s product aesthetics and functionality and a lot more that you need to care off in a new project, but this would give you a good start on analytics planning. 






Leave a Reply