Setting Metrics

“What gets measured gets managed.” ~Peter Drucker

As a Product Manager, your entire role revolves around metrics. Metrics are measurements that inform you about your product and ensure you are covering the customer lifecycle for each user.

Of course, it’s crucial for you to first define what a successful metric would look like. It’s also important that you keep your measurements consistent.

Types of Metrics

Growth & Activation

Typical examples include total new users per week/month, new users by source, and activated users.
As a PM, you want to know where new users are coming from (SEO? the app store? a blog?) and if they actually engage with your product (i.e., send their first tweet).

Engagement

This will be one of the most common metrics tracked at consumer product companies. These are usually tailored, per company, to encourage a certain type of behaviour, depending on the goals of the company or industry. There is no consistent, quantifiable definition of engagement across different products.

For example, a to-do app might define an engaged user as someone who logs in every day to add and complete items. For an invoicing app, an engaged user might only log in once a month.

Retention

This is tied closely to the growth and activation metrics. It’s a way to find out who is coming back. This can measure returning users or resurrected users (users you convinced to start using your product again after a period of non-use).

For example, let’s say 100 users downloaded your app last month and got activated. This month, only 60 are back. User retention rate is 60%. Let’s say for the 40 who are gone or ‘dead,’ you send them a push notification reminding them how awesome you app is and all 40 come back next month. That would be a 100% resurrection rate.

User Happiness

Pretty self-explanatory. This quantitatively tracks how happy users are. This could be net promoter score (NPS) scores or the number of customers who have written a complaint to customer service.

These are some of the most difficult metrics to get but very important, especially for companies or industries where there is little competition or where people don’t have other options (i.e., your cable company).

Revenue

These are very important metrics at companies earning revenue from customers or at consumer companies where consumers are being advertised to.

Lifetime value (LTV) and cost of acquisition of customer (CAC) or cost of customer acquisition (CCA) are common metrics. Business-to-business companies put a lot of value into tracking monthly recurring revenue (MRR) and its counterpart annual recurring revenue (ARR).

Metrics at Large Companies

Project managers at larger companies will be more concerned with certain metrics and less sensitive to certain risks and costs.

Metrics at Startups

Project Managers at startups are looking for validation metrics. As you might have guessed, a validation metric something that demonstrates real interest from your potential customers. The questions you are solving for are: Is my product a good idea? Does it meet a user need? Does my identified problem really exist? Some metrics might be: Percentage of people who sign up; percentage of people who share your posts; average purchase price; number of people who open your email.

You are also more concerned with the economic viability of the project. Lifetime value and gross margins

Exploratory Metrics

These are metrics you’re not always tracking or telling your boss about. They’re there to explore and see what user behaviour is like.

Reporting Metrics

These are metrics you track over long periods of time to ensure your product is doing well.

How to Set Good Metrics

Good metrics should be:
1. Understandable and simple.
2. A rate or ratio. Example: Monthly active users / Total users
3. Changeable. Example: You run an e-commerce business and sell an item people only buy once a month. Increasing their visits would be a marketing (not a product team) issue. Therefore, you might try and get them to spend more money once per month.

Note: Watch out for false correlations in your data.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.