Jack Mardack’s 4 Quadrants of User Data

“With software, not only do we have to measure data and succeed, we have to invent how to do it each time.” -Jack Mardack

Jack was employee #2 at Eventbrite where he helped grow the company from 3 people sharing a desk, to $130 million in revenue.

After, he led growth at Prezi where he and his team had to build their own analytics software. The purpose was to measure the ratio of user information & behaviors together to encompass the whole picture.

He’s currently leading growth at Chartcube, where they transform excel files into beautiful presentations. Jack’s systems thinking process allows him to use data to understand the product in the early stages.

The four quadrants of user data

The grid above shows what marketing person is usually involved in that quadrant. The is relevant to both small and large teams, the only variable is the number of resources you able to throw at each quadrant.

1. How did you get here?

The how did you get here quadrant generally comes from the mechanics through which users are delivered to your software. This is mostly by links that contain identifiers. When someone clicks a link to your site the identifier fires. These links can be stored in google analytics or HubSpot.

In Prezi’s unique case, they were stored in a database they built themselves. (You can watch more about here.) The key thing to know about data that comes out of this quadrant is that it can be relatable to things you know about the user as they pass through the other quadrants.

Once we know more about our users, we can circle back to this quadrant and optimize or automate channels that have proved to be successful for finding users.

2. Who are you?

Once we set up analytics to collect user data. We can find out the “who are you” based on what the data tells us. This data is the information is given to our analytics from the user’s IP address. These identity attributes include the user’s language, browser settings, geo-location.

The challenge with using third parties like HubSpot and google analytics is bringing this information into relation with other data after the user gives an email or opt’s in via a landing page to get more info.

3. What have you done?

The data that comes out of this segment comes from the events you instrument. This is the one you have to invite each time because each product starts out as a blank canvas.

You need to have a conscious idea of what those events are based on what you’re trying to get people to do in the software. Different from the previous two segments, you can’t expect data to guide you here.

What’s more important is the relationship between collections of data. If you never install the measurement, you’ll have a hard time finding out that it’s important when staring at a blank canvas.

4. What do you think?

This is exciting because you are trying to arrive at ideas of the user’s state of mind. A few ways to do this manually would be to install surveys or by collecting NPS.

Jack used NPS score to collect data at Eventbrite and Prezi. In each instance, this was a significant KPI for this quadrant of data. Matching users NPS score with geography or number of Prezi’s made helped influence business decisions.

The utility of any data is contingent on its relatability to other data that you collect from other quadrants.


Given the mandate of a growth team, you have to install the perspective that includes all the factors that influence a system and appropriately measure.

That has a tendency to get bigger and bigger and bigger. For that reason, the way the growth mindset comes at analytics is very good because it’s inclusive and tends to see the right relationships among things rather than being siloed and myopic missing the bigger picture.