A Data Science Central Community
Customers... What do they like? What do they use? What kind of customers used my precious new feature? There are many reasons that any business should use customer events. Ultimately everything leads to two points: customer satisfaction and revenue.
When I first started playing with Google Analytics it was beautiful. A new world was in front of me. With only a quick setup I was able to see what's going on with my users. Many companies cross to more detailed questions and Google Analytics are not enough. There is a large amount of tools out there such as Mixpanel and KiSSmetrics. At Blendo, we are huge fans of Mixpanel. Mixpanel is the leading web and mobile analytics platform. With minimal coding, Mixpanel helped us to easily create funnels and visualizations of how people are using Blendo. With Mixpanel, you may dive into your events, analyze them, discover trends or create weekly cohorts.
Essentially anyone on a team can run some analysis in a short amount of time. That's the beauty of Mixpanel and how it democratized access and analysis of customer event data.
At some point, you may need a Product Manager, an Analyst or Data scientist to ask that one more sophisticated question. Questions like "What is the average time between action A to action B" or "How do custom segments of users behave in my product" or "What kinds of actions are most likely to lead to a sale"?
In many of these cases, you may need data from various other sources. For example, you may need to combine specific data or reports from Mixpanel with data from sources like Mailchimp, Stripe or your transactional DB.
Using the information with specific events from Mixpanel will provide half the truth. The other half is to get the information of how to segment your users. For example, a segment could be:
This will give you some insight into how segments of your users behave, and how to stir your product development.
Ever wondered how long it took for a user to signup until to make a purchase? Or how long it took for a user to do make an action or do another? All these data are in Mixpanel, but you may need a separate layer on top or custom views to add your business logic.
From the users that actually registered and paid for your service, who came from a webinar or an ad? Who was more valuable in the end as he made an up-sell.
Using custom properties and events from Mixpanel is your friend again here. In addition, you will need data from your payment service or Google Ads. This can help in making intelligent business decisions about how to allocate resources towards your webinar or your Marketing Ads.
One way is to get the data that reside in various places (eg Mixpanel, email marketing, Ads, payment service, your transactional DB etc). Next, you will need to push everything to a data warehouse solution like PostgreSQL or Amazon Redshift or Google BigQuery.
If you have the development resources and time, you may build the data pipelines on your own. That means that you have to write, host and maintain a flexible data infrastructure. But it is not always the best solution as it comes with heavy maintenance and commitment from your dev team. That will lead you to redirect valuable development sources to something that is not your company's core value.
If you need something now without having to commit development resources you may use a self-serve product.