How to Use Data to Prioritize the Projects That Drive Real Growth

Why is using data important? How can product managers use data to prioritize initiatives that really fuel growth?

These are only a couple of the questions that Claudiu Murariu, the CEO and Co-Founder of InnerTrends, covered in his presentation at the 2022 Product Drive Summit.

Would you like to learn more? Let’s jump right in!

TL;DR

• We generally divide teams into those that use data to make decisions and those that don’t.
• These days, almost all companies collect user data with the hope of using it to improve their products. However, only a fraction of them collect the right kinds of data, and an even smaller percentage are able to extract actionable insights from the data.
• Teams that don’t use data, rely on anecdotal user feedback or feature requests, or on what their competitors do to make product decisions.
• They make decisions based on general assumptions that are not always relevant to their product or user problems.
• Teams that use data, have structured processes and frameworks in place and use them to identify issues, generate ideas, make hypotheses, and prioritize solutions.
• To use data effectively, you need to create a customer journey map that will provide a structure for data collection and analysis.
• Next, you collect data about user interactions with each of the touchpoints and display it in one place to get a high-level understanding of your challenges.
• The next step involves asking lots of questions to get a deeper understanding of the context.
• Answering the questions will help you formulate hypotheses about the causes of the issues and possible solutions. Come up with multiple hypotheses to give your team more options. Writing hypotheses is easier when you use a framework.
• When your hypotheses are ready, you need to prioritize them according to their potential impact.
• How can Userpilot help you make data-driven decisions? Book a demo to find out!

How do you decide to build?

When it comes to decisions on what features to build, teams fall into two camps: those that use data and those that don’t.

Teams that make decisions without using data, get their ideas from secondary research, like internet blogs. It’s a common practice to build features that customers request.

Very often, their decisions are also driven by the competition. This could be to build parity features or the opposite – offer users functionality that no other competitor supports.

Data-driven teams don’t jump at promising ideas straight away. Instead, they explore them to understand their context, identify cause-and-effect relationships, and prioritize the best solutions.

Which team are you a part of?

How well can you use data?

Collecting data and using it effectively to generate insights don’t always go hand-in-hand. In fact, a great majority of companies fail to take good advantage of the data.

Most companies in the SaaS space collect user data in different shapes and colors. Their intentions are good. They want to gain insights into user behaviors to build great products.

However, only a fifth or so of these companies collect the right data that has the potential to help them gain a better understanding of customer habits and needs.

Even if they get their hands on the right kinds of data, only a tiny percentage of these companies are able to actually use it effectively to get meaningful and actionable insights.

How do you run a product onboarding project?

To give you an idea of how the two approaches to building products differ, let’s take an onboarding project as an example (which is so down the Userpilot aisle!)

The ‘without using data’ way

If your team doesn’t have the culture of making data-based decisions, they probably won’t know that there is an issue with the onboarding process until customers complain or something else brings it to their attention.

Next, they pick an idea on how to solve the issue, like a product guide. The marketing team would design the guide, select a tool to get it out to their customers, and their job is done. Or is it?

The using data way

What’s the alternative?

You can use the data to make informed choices on what problems to address and how to do it so that you get your priorities right, and maximize the return on your investment.

How to use data and drive growth

Now let’s take a closer look into the “Using data” way.

Step 1: Build your Customer Journey Map

The first step involves building a customer journey map.

Thanks to the map, you will be able to see all the customer touchpoints across all the stages of their journey, from the moment they discover the product to the point when they subscribe to your product.

To make the map as comprehensive as possible, make sure to involve all members of your cross-functional team. This will help you get all the different perspectives on what each stage of the journey involves.

This is also necessary to align the team and get a shared understanding of the customer experience.

More importantly, a customer journey map enables you to navigate and analyze your data in a structured way.

Step 2: Find areas of improvement

Once you have the journey map, you need to put data on all the touchpoints of the journey.

How do you do that?

You need a dashboard for that. It will give high-level insights into how different aspects of product adoption happen.

Thanks to this, you will be able to spot the biggest issues that need addressing immediately. If you have suspected that there is an issue with product onboarding, this is where it will show right away.

However, it may as well turn out that the other areas like acquisition or retention require even more attention.

This is where you should concentrate your energy to maximize the impact of your team’s work to drive growth.

Step 3: Understand the context

To improve or optimize the product or processes, you need to understand the context.

You do it by asking the right business questions.

What are some questions that could be relevant to our onboarding case study?

Example 1: How are people converting during the onboarding process?

By looking at the onboarding funnel, you can identify where users drop off and which areas you should improve. As a result, you are able to focus on optimizing only the necessary steps.

Example 2: How long does it take to finish the onboarding process?

Investigating this aspect also helps you identify issues. For example, if an action takes much longer to complete than you had expected, it may mean there’s some friction point that you’re not aware of.

Understanding how long the most successful vs the least successful users take can also be helpful, as is identifying the correlation between the time and payments.

Example 3: What parts of your product help people onboard?

By looking at user actions between different onboarding stages, you can identify those that contribute to customer success. Next, you can drive users with in-app guidance and messages to those actions to increase the chances of completing the onboarding process.

These are just a few examples that can help you understand the context better. Again, team members with different areas of expertise can help you ask the right questions, so make sure to involve them.

Answering the questions will give you some ideas on how to address the problem you’re facing. What’s important, these are based on your product data and not what your competitors do or internet research.

Step 4: Come up with a hypothesis

Once you’ve answered the key questions and have a solid understanding of the context, it’s time to formulate some hypotheses.

Mind the plural form here. That’s because you should always come up with a number of possible explanations and solutions, not just one.

Using a framework like the one below can help you do it in a structured and consistent way.

You start by stating the status. This is the current situation or baseline. It explains why you’ve decided to take on the issue. In our example, this could be the fact that the onboarding success rate is below an industry benchmark.

Next, you depict the context. These are the findings of the data analysis you’ve carried out. That’s where you identify where exactly the problem is.

In the next step, you focus on priority. At this stage, you explain why it’s important to address the issue. You can do it by focusing on the impact the change would have on your product. This is necessary to get the buy-in from your team members or the business stakeholders.

Finally, you formulate the final hypothesis. The statement should start with ‘We believe that…’ and could go on like:

“…building a product guide on the 2nd step of onboarding will give us the highest chances to onboard more users. We expect an increase of at least 15% in the onboarding rate.”

Remember to back up any claims that you make with solid data and rationale.

Step 5: Prioritize

As you’ve developed multiple hypotheses in the previous phase, it’s time to prioritize them.

How can you do it effectively?

Start by documenting all of your hypotheses accurately. That’s where many companies that aspire to be data-driven fail, so don’t overlook it.

Next, set the impact score for each of the hypotheses. The point of this is to choose the ones that will drive the most value.

After that, invite your teammates to question your assumptions. This is to help you identify fallacious premises underpinning your hypotheses. To start with, it may be hard if they’re not used to it, but as time progresses, they will develop the habit.

As you start implementing changes, keep a record of the difference between the estimated impact and the actual impact. This will help you estimate better – and track how much better you’re getting at estimating.

Finally, make sure you establish cause-and-effect relationships when you’re testing your hypotheses. The easiest way to do this is by manipulating one variable at a time and running A/B tests to check if the needle moves.

How to use Userpilot to build and test hypotheses

Userpilot is a product adoption platform with a number of features that enable you to track product usage and run experiments to formulate and test your hypotheses.

Product usage tracking to develop hypotheses

With Userpilot, you can easily track and analyze how your users engage with your product.

And by engage, we mean not only clicks but also text inputs and even hovers. In practice, that means you can create a very detailed heatmap-like picture of user activity.

To enable that, you need to tag them, which you can do directly from the Chrome extension.

You can track each of them individually or combine them together into Custom Events to track and analyze them as one.

What’s the use of that?

For example, you can use the functionality to track how your users progress through the adoption funnel. If you see, that they drop off at a particular point, you can look at ways to help them succeed, for example by using a range of individual UI patterns or developing custom in-app onboarding flows.

A/B testing to test hypotheses

How will you know that the in-app experiences you’ve just designed work?

As mentioned, A/B testing, also known as split testing, is the easiest way to test hypotheses and establish cause-and-effect relationships.

You simply enable a new feature, or in our case, an onboarding experience, to half of your users from a relevant user segment, and watch whether it makes any difference to the conversion or adoption rate.

Userpilot enables you to create and design such tests easily, and completely code-free.

To get the best result, test one variable at a time to ensure you capture causality, starting with those that are likely to have the greatest impact.

Conclusion

Instead of relying on luck, hunches, or anecdotal information, using data enables teams to identify issues reliably and generate solutions and ideas in a structured way.

If you would like to see Userpilot data analysis features in action, book the demo!