Awesome!!! You came back for more punishment to learn how product management and user analytics will make you a zillion dollars for you company.
Or did you come here to learn about an illness that affects all product managers at one time or another……Metric Obsessive Compulsive Disorder.
Wait!! What’s that? You are still confused by user metrics, analytics and KPIs? Well, since I’m feeling particularly charitable today, I’ll do a quick recap:
Want to understand corporate goals and success metrics? Read this – The Product Roadmap And Corporate Strategy: Connecting The Dots.
Forget what user metrics are? Read Product Management and User Metrics.
So much for a recap right? Consider yourself lucky I am at least pointing your lazy ass in the right direction, which is more than I do for my kids.
Now it’s time to dive in to that ball of professional yarn, user analytics.
I’m going to start off with a bit of a disclaimer here. When you you start talking about user analytics, you’ll typically hear references to analytic tools offered by a range of vendors, like Google (Notice that Google Analytics isn’t called Google Metrics) Google, Adobe, Mixpanel, Matomo and StatCounter. All of their marketing material make it sound like these tools are simple to use and offer you guaranteed success. Nothing could be further from the truth. Getting the most from user analytics is a product management art and much can go wrong:
- Sometimes the products are not implemented properly
- There is a conflict between your code and the analytics package (wait till you find out that Google page speed algorithm penalizes you for installing their analytics package on their your site)
- There are bugs in the analytics code
- You inherit someone else’s messy project.
I raise these points to illustrate there is always a learning curve with user analytics. As corporate goals change, you will need to understand how the underlying metrics impact the user analytics you are using to track your progress, as you may adjust your funnels and metrics to better reflect your new reality.
Beyond expertise, user analytics require time to aggregate a meaningful level of data for what every process it is you are testing in your product hypothesis. You are better off erring on the side of caution and letting experiment run longer.
engagement, retention, and self-perpetuating. Startups that go through all three tend to turn into multibillion-dollar companies, whereas startups that get stuck in one phase commonly fail.
The goal of the first phase is to get customers using your product and completing the core action, like posting a photo to Instagram. This is a sign they’re engaged with your product, and we could say that completing the core action is a success metric that supports an engagement goal. Pinterest’s core action is pinning something.
User Analytics Gotcha: Product Management Survival Secrets
You're so vain You probably think this song is about you
Make Sure Shit Is Installed Properly: (Do I need to have a trigger warning on the site if I swear?) The marketing material makes things look easy, but tagging the right object to capture the right metric is easier said than done. Then layer on top of this that there is sometimes a bit of a lag between when the data is collected and when its reported makes installing analytic packages easier said than done.
Understand How Metrics Are Collected: There are two factors at play here:
- Most analytic tools use samples a subset of your traffic for its reports.
- Ad and tracker blockers abound. If a browsers blocks analytic script, you get no data. Learn more here about the issue.
The end result here is that your user analytics does not represent what your customers are doing on your site,
Metric Obsessive Compulsive Disorder: A good indicator that you are dealing with someone who does not understand analytics is when they insist on capturing every…. single…. …metric.
As this is indicative of a poor understanding of analytics, you and your career would be served well by doing an audit of the relevant user funnels and processes to make sure they are reflective of your KPIs before building out and testing any product hypotheses.
