Analytics is not just for Christmas

Emer Kirrane has had the benefit of working in everything from startups to big enterprises, but mostly she has experience with web analytics, and reasonably clear-cut funnels. Having worked on an analytics product, she had the luxury of using the tool she was optimising for her customers, as well as direct access to the developers who could give her any data she desired!

When she joined Prezi, she found she had to think about people as much as the software to understand how the whole product is used. Of course, on top of the usage data her team has direct access to, there's also social data, support data,UX and product research, marketing campaign data, and sales data!

What Do You Do With Data

How do you organise all that data? It's a challenge, not least because that data is inter-related. In order to make sense of it, you need to have cross-functional teams (dev, UX, PM, marketing), and a culture of data will help you achieve that. To create that culture, you have to "decorate for data", ensuring that key aspects of the data you're gathering, and the spectrum of that data, is out in the open and visible to everyone.

As well as making your data visible, it's essential that you build strong relationships across teams to share information – not all data will come to you, sometimes you have to go and find it. One way to achieve this is to try having a cross-functional "daily data standup", where you talk about what you've learned, and what you're doing to shift your KPIs.

Emer wraps up with two critical insights:

• Test your data capturing infrastructure, NOT just your product. Expect your data capture to fail at some point, and build in checks to be ready.
• The best way to fight a HIPPO (HIghest Paid Persons Opinion) is with data – data that proves that their opinion will not help the bottom line. Make sure you know your users and your KPIs.