How to kickstart data-informed innovation
Stop me if you’ve heard this before. You’re in a product planning meeting and everyone is throwing out new things to add to the roadmap; no not the normal extending or fixing of existing feature stuff. I mean, what big rock do we deliver next type of questions. It seemed like everyone — sales, customer success, whoever that guy was at that meeting — had an opinion, or a hunch, or a “gut feeling”. But, no one had any tangible facts.
We just wanted to build software products that people love and love to use. But there were lots of questions. What does the business need? What do customers really want? How can we help users get the most out of our product? Where is the best information to help us make a decision?
HiPPOs aren’t always the best product leaders
In an article written by three Microsoft product leaders in 2007, they stated, “[Their] experience indicates that significant learning and return-on investment (ROI) is seen when development teams listen to their customers, not to the Highest Paid Person’s Opinion (HiPPO).” Let’s face it, we’ve all been in planning meetings when the HiPPO in the room had a “gut feeling” about what to go build next.
Development teams should listen to their customers, not to the Highest Paid Person’s Opinion (HiPPO)
Now, let’s wander back to what we learned in school about how to make decisions when the answer isn’t immediately clear. Our professors worked to instill a more experimental framework for helping us get to the right answers. One that includes more scientific principles and allowing us to make the best possible decision with the evidence we had, testing it and then deciding empirically if we were right, or needed to make a change, or abandon the thesis we put forth altogether.
We were forced to make observations about the things around us and what we were learning from them. This was primarily done to help us form a hypothesis to explain what we were experiencing in our observations. We then tested our hypothesis by forecasting an outcome that was largely coupled to our general thesis. Then we created controlled experiments on these theories and, if the outcome we expected was achieved, then we had proven our theory.
Creating a culture of data-informed innovation
So how does one create a culture of data-informed product innovation practices in their organization? How does one even begin to make the organizational change required to make data-informed product innovation commonplace?
First, what we need to do is alter our thinking to pose our solution in the form of a problem statement to arrive at a hypothesis. This is especially critical to big rock items such as a new module or capability in the product. We also need to understand the segment of users we would like to target and finally the criteria for arriving at an answer to our hypothesis. Bottom line, we don’t “execute projects” anymore. Instead we execute experiments against our hypotheses.
And the key outcome of any experiment should be collect measurable evidence and learnings. To learn is to gather information from the experiment that you will use to prove or disprove your hypothesis. And to learn, we do what our professors taught us, we apply the scientific method for investigating, obtaining knowledge, adjusting to that knowledge and interrogating the previous thinking we put forth.
The basic scientific method says to:
- Ask a question
- Perform background research
- Construct a hypothesis
- Test the hypothesis with an experiment
- Analyze the data and draw a conclusion
- Communicate the results and interrogate your previous assertions
Once a product leader is prepared for the switch to data-informed innovation, there are three key aspects to consider.
- You will need a platform to target and execute your experiments for segmented cohorts, have individual conversations with your users, capture usage information and collect feedback.
- Connect this platform to your application and to the most critical external data sources, such as user profiles and CRM data.
- The ability to visualize and interrogate a rich set of analytics that are pertinent to your product. And, to share key metrics through executive dashboards with key stakeholders.
It’s important to understand that as data volume, variability, and rate increases, so does a product leader’s ability to utilize it. This data should help you interrogate your hypotheses to empirically prove you are meeting the customer’s expectations, which ultimately leads to increased adoption and retention of customers.
Summary
As I mentioned in my previous blog post, continuous innovation in software development is evolving and maturing. As it continues to mature, we have an opportunity to fundamentally change how software is created, by utilizing a more data-driven approach. Product leaders will require tools to construct, segment and run experiments for their application by engaging their users. Moreover, these tools need to help gather and measure the varying results of each cohort, whether it’s increased adoption, how a feature is perceived or used and how we discover what new features various cohorts need/want. These experiments will lead to new learnings and we can use these learnings to build the most perfect software possible for customers. With a goal toward optimizing our effectiveness in solving the right problems for customers, versus simply implementing the next HiPPO-driven solution.
At Aptrinsic® we believe your product is your best sales tool and marketing channel. We also believe using a product-led GTM strategy will help companies acquire, retain, and grow customers by creating realtime and personalized experiences driven by product behavioral data.
