How Product Managers Can Dissect Data
Katie van Zyl is currently a Senior Director of Product Management at Godaddy, where she leads product management for godaddy.com. She is passionate about GoDaddy's vision to empower the everyday entrepreneur by giving them the tools, insights, and people to transform their ideas and personal projects into success. Katie has spent 20+ years in the software industry-leading product management teams across various products and platforms. She is also very passionate about supporting her fellow women in tech, volunteering with national and local organizations as well as leading the GoDaddy Women in Tech employee resource group.
She recently spoke at a Product That Count hosted webinar and discussed dissecting data for PMs. She shared a case study from her team and some strategies that you can apply with yours.
The presentation from Katie van Zyl focused less on numbers crunching and was more focused on how to set yourself up to be successful with your product data and using it to forward your business. You can watch the full webinar above. Or you can check out the highlights which are detailed below:
On where to begin when dissecting your product data
Your data might contain the answers, but you need to make sure you know what the question is in the first place.
“So step one, know what question you want to answer. It’s always good to have a framework that you measure back against and that you leverage when you are looking at the data that you want to get the experimentation. The research data that you want to get. So, one thing that I think is super important is to be really clear about that question that you want to ask and get answered. Be clear, define, write it down.
It’s really important to understand the altitude of that question that you want answered. Zero the data in around that and what you’re going to do with that data. You need to be able to answer the question and know what you’re going to do with that answer.”
The next step when dissecting your data is to ensure you understand your metrics
Truly knowing what the data you’re looking at represents is the next step in dissecting product data and being a great product manager.
“Step two is identifying the data you need, and really understanding the metrics. You need to know what data you need to answer a question very specifically so that you know what you’ll do with the answer. And I find sometimes if you have very ambiguous data points, it’ll be really hard to make decisions to change your experimentation to win or fail to change the next set of things, your backlog and your roadmap.”
The third step is all about how you approach your research
Even before the experiment, the research you do can make or break you.
“Setting up your experiment in your research is so critical to then getting the data that you need. You have to make sure you don’t bias your results. You need to make sure you’ve asked all the right questions and that you’re set up to get the data that you need. Be really mindful and very thoughtful about how you approach the experiment or the research setup.”
It’s important to know what your data represents
With product data, you need to make sure that you understand what you’re looking at.
“Step four is to really understand the data. I see a couple of really common mistakes that people tend to make. One can be blindly accepting whatever is given to them and then executing on it. The other can be forcing an interpretation of the results to fit that hypothesis or what direction they want to see it go.
One piece of advice I always give is to make sure you’re looking at the data daily, weekly. Make sure you understand trends. And make sure you can spot, as you get more experienced with it, things that seem out of the norm. So, when you get experiment results back and it’s spiking way up, you can say, hey, that’s a great success. Next time I declare victory. Or you can say, hey, well, that’s actually interesting. Is there something different here?”
Dissecting product data isn’t a one and done exercise
You need to know when to stop looking at your product data and when to take action.
“The last step, step five, iterate as long as it makes sense. This is one that I think is tricky. You have to kind of figure out do I win, declare victory, roll to a winner, and then try a new treatment of something completely different? Do I need to iterate on it? Do I need to do more? Or do I need to just feel it? And so I think it’s important to ask that question.
Sometimes we just get caught up in the individual experiment and the data set that we need and then we move on. Often, you can extrapolate or get additional data that you can work together to make a decision on a roadmap if you just keep doing a little bit more testing around it a little bit more.”