Why Google Analytics Doesn't Help To Improve Your Product
How many times have you opened Google Analytics, looked at one of the reports, and as a result, got an insight that led you to an idea on how to actually improve your product? It’s probably close to zero.
The reason is that Google Analytics is a marketing, not a product analytics tool. Many people who work on products forget that, and are applying the wrong tool to their product questions.
Marketing analytics is concerned with answering questions around where your traffic is coming from. For example, what countries are my visitors from? What campaigns, referring websites, or search traffic send me the most users?
While marketing analytics is necessary to get some idea about the acquisition part of who is coming to your website, it doesn’t relate to the product goals that matter.
Most of the metrics available in marketing analytics tools are “vanity metrics”, when compared to product analytics metrics. They show absolute counts of visits and clicks, but that doesn’t necessarily relate to how well your product is delivering on its promise of providing value to the users.
Take, for example, a product that has an increasing number of users from referring sites or organic search channels. While it may give you an uplifting feeling ito see those numbers climb, it doesn’t tell you much about what you can expect in the long run.
Your users might be ending up in a leaky bucket. Low retention and low engagement can cause those new users to be lost forever, because they don’t experience the product’s value proposition early in the funnel. Therefore, they will never experience an aha-moment.
On the other hand, a competitor that has a lot less traffic, can do much better in the long term, if it has much higher retention and user engagement.
This is where product analytics comes in. It’s the right tool for the job when you’re looking for ways to improve your product’s key metrics.
Product analytics is the process of understanding how to improve on product metrics that matter. Companies like Facebook, Uber and Netflix invest millions into data analysis in order for them to analyze their users’ key metrics. This includes factors such as engagement, virality or retention in order to improve the product metrics.
This benefits both customers as well as company revenue streams. Ultimately, it translates into better business decisions being made on behalf of those who benefit from them.
But finding out which product analytics metrics are relevant for your product or service, is not as easy as you might think. It involves a lot of what-if work based on analytics reports, creating complicated machine learning models, and experimentation.
All together it’s a very time consuming and costly process, usually involving big data teams, large budgets and many months of work.
After a lot of experimentation with their data, Facebook, for example, found out that if people add 7 friends within 10 days, they’re much more likely to stick around and keep coming back to the platform.
Doing this kind of aha-moment analysis properly costs much more work than you might think. It’s because it is easy to make mistakes and difficult to use proper data models that are unbiased.
On the other hand, take the struggle of deciding which product features you should remove, and which ones to nurture. By doing this properly, you make sure that users get to experience it’s value proposition as early in their journey as possible. But making mistakes here can be catastrophic for the product experience.
Another aspect product analytics uncovers is measuring feature usage. It does it by looking at power users. This metric is based on the L30 and L7 metrics that originated from Facebook’s growth team. The L30 will tell you how many days out of the last 30, your users were active. If you apply the same analysis to features usage, this will tell you which features are more or less important. It will also help you decide on the most suitable pricing model.
To show you an example, let’s look at the pattern in the chart below. We can see that this feature has been used for many days in the last 30 days.
This feature seems thus crucial enough that a large portion of active users need it on a daily basis. So it seems like a good candidate to nurture and make sure the user gets to experience this feature as early as possible in the onboarding journey.
But if we look at the pattern below for another feature, we can see that it has mostly one time usage. That doesn’t mean we should immediately remove this feature, as it may be a very crucial one. Such as exporting user data to make a backup.
But if our expectation for this feature is to have daily usage, we can clearly see that it doesn’t match reality, and we should reconsider and dive deeper to see if this feature is really necessary.
As you can see, the possibilities are endless when it comes to the types of insights you can get from your product data. This article scratches just a small amount.
Where things start getting more complicated, is once we start looking at extracting behavioral patterns and utilizing them for an improved user experience. And as a result, improving the product.
This is an area that most popular product analytics platforms don’t cover or simply fail at. That’s because although they show specific metrics, they don’t give you a way to predict what behavior leads to conversion. Therefore, those metrics still leave you in the dark in terms of making actionable steps to improve the product.
Luckily, Stormly developed a way to turn your data into actionable product insights without having to spend hours on experimentation and guesswork. All the metrics we mentioned can now be discovered within just a few clicks. And then easily added into your personalized dashboard!
Stormy even goes a step further and lets you pick your product goal, such as increasing retention rates or ARPU. This will then help uncover behavior that leads to reaching those milestones. And to make sure that they are creating best-in class products for their customers!
What are you waiting for? It’s never too late to try out the best analytics software that uses the same techniques as tech unicorns. Sign up today!