The world’s most innovative companies use data to design their products. This is how they do it.

By using both quantitative and qualitative data to discover, validate, decide and design.

Use data analytics for quantitative research (quant).

Use design research for qualitative research (qual).

There are 2 ways to do this.
A. Discover with quant, validate with qual.
B. Discover with qual, validate with quant.

A. Lead with quant research and back up with qual research.

Discover anomalous patterns with data analytics. Then validate the why and how behind the what with design research.

Example 1

A data analysis of the first 100 users for an MVP revealed that only 20% of new user verified their account, which leads to dismal 10% of total users onboarded. There’s also nothing wrong with email deliverability. You identified some of the users who did not verify their account, did an outreach and speak with them for quick 15-30 mins to understand what’s stopping them from verifying their emails and using the app. You learned that 15 users that you talked have not received enough value from your MVP before verifying their emails. Therefore you decided to eliminate the friction of having to verify emails, let the users experience value and prompt them to verify emails before performing critical events - which drastically boosts the email verification rate to 90%.

Example 2

Your product analytics report discovered that 75% of users are navigating the same page repeatedly (5x median) with at least 5 mins session time each before clicking to submit. You identified and quickly interviewed to 5-10 of these users to ask what’s going through their mind before clicking submit. You learned that the users are interested with your product but is confused about the requirements and how to navigate. Therefore you decided to simplify the requirements and navigation to enable operational transparency to boost time-to-submission rate within 5 mins and 2 session repeats.

B. Lead with qual research and back up with quant research.

Discover behavioral insights with design research, then validate emerging patterns exist with data analytics.

Example 3

Your team is tasked to develop new solutions for the elderly market in Singapore to age in place. You conducted a user research with 13 elderly above 65 years old who are aging in place; and discovered that 80% have enough food and do not need rice donation, but are having difficulties maintaining their houses (eg. Change lightbulb, fixing door knobs). You decided to run a survey with 1,000 elderly with a local non-profit and discovered that indeed a whopping 95% have difficulties in maintaining their house that’s hindering them from aging in place. Your team ended up creating a home improvement solutions knowing that there is a large TAM.

Example 4

In your initial user research for your new product with 10 users, you discovered that 3 of them couldn’t achieve their outcomes claimed using your product. If left alone, the churn rate will widen. You analyzed the behavioral data of the users and discovered that users that took A, B, C actions but not X, Y, Z actions will generate the desired output but not the desired outcomes. Knowing this, you decided to to make X, Y, Z actions as recommended to users right after completing A, B, C actions. As a result, this boosted the positive outcomes from 75% to 98%.

Combining data analytics and design research is a powerful and proven approach for teams to continuously generate insights and make informed design decisions.

But it’s not uncommon to find these 2 disciplines operating as silos.

Design research is owned by product professionals (user researcher, product designers) and typically sit within the product or design function.

Data analytics is owned by data professionals (data analysts and scientists) and typically sit within the analytics or technology or digital function.

If your goal is to enable your teams to continuously generate insights, these 2 disciplines shouldn’t just be sitting inside their domain functions, but be  working closely in tandem as an Insights function to collect data from various sources and mine insights with quant and qual research.