The Five Whys of Data-Informed Decision Making

When it comes to data, we are all enamored. We track our steps and our packages. We let algorithms suggest our playlists. And IBM will tell you that it’s allowing us to create a smarter planet. But is it creating stronger enterprises?

Snapshots of data don’t always suggest a path forward. It often takes some digging beyond your initial questions to unearth the deeper insights that inform better decision making.

Back when I was a product designer at IDEO, we often got to the heart of a challenge by asking ourselves a series of carefully considered questions – sometimes called “five whys,” based on the notion that it can take five successive rounds of questioning to finally surface an underlying cause. In many ways, this reflects how business intelligence analysts approach the massive datasets they interrogate. But they’ve only recently gained the software tools necessary to follow a nuanced line of inquiry in minutes instead of months.

Before, even at most large companies, the data were limited. The tools were unwieldy. The few available analytics experts juggled competing demands on their time. Answers took too long to turn around, and if you wanted a quicker response, you had to ask your question in a way that was derivative from something you’d asked before. In the end, it was often better to avoid a meaningful shift in your line of questioning – even if it meant giving up on a promising lead.

Today, that’s no longer the case. Data are plentiful, thanks to mobile devices, social networks, and user-generated content. And storage is cheap, thanks to the metronomic march of Moore’s law.

Meanwhile, smart emerging tech companies are doing more than simply saving everything to the cloud. They are also weaving together their various datasets in the cloud – payment transactions, marketing automation data, you name it. And most important, they’re adopting cloud-based data analysis platforms that make it possible for a broad swath of users to access that all-encompassing dataset.

As a result, analysts no longer need to settle for fixed dashboarding tools or rely on IT teams to spend days or weeks generating costly models. In fact, the best of the new cloud-based analytics tools provide unfettered access to raw data in order to give analysts the flexibility and precision they need to get fast results.

As others have indicated, this has given rise to a whole new group of analysts: the citizen data scientist. These are creative, inquisitive people who often have no formal database training. But, when exposed to methods alongside analysis, they begin to tinker and are able to learn the basics. And when working with analysts, they can combine their quantitative skills with deep domain expertise to investigate and improve everything from the product or service to the business operations. They’re the type of people who ask “five whys” – or as many “whys” as it takes.

My firm, Foundation Capital, led the Series A of one company, Mode, that created its collaborative SQL analytics platform with these citizen data scientists in mind. Recently, I was reminded of the “five whys” when Mode’s co-founder Benn Stancil told the story of finding Watsi’s growth engine.

Watsi, which crowdfunds healthcare for people around the world who can’t afford it, had experienced a sudden spike in donations and an increase in its fund’s growth rate. What caused it? Could they replicate it? Could they improve it? Benn used Mode to delve into a series of more specific questions.

1. What led to increased donations?

A $20 Watsi gift card giveaway, sponsored by a partner.

2. What factor made gift card recipients most likely to donate?

Showing pictures of patients.

3. Is there a downside to offering $10 gift cards instead of $20?

No. When the gift card is of lesser value, people give reoccurring donations at the same rate – and you reach more people.

4. Which type of gift card yielded the most donations?

Cards given by Watsi’s partner, Segment.

5. What makes Segment cards special?

They required donors to provide personal data. The process effectively weeded out people who were less committed to Watsi’s mission.

At the end of his investigation, Benn was able to offer concrete suggestions to help Watsi improve its growth engine: show more photos of patients, offer twice as many $10 gift cards, and make the cards harder to claim. He calculated Watsi’s return on investment would go from 74 percent to more like 150 to 200 percent.

It’s easy to see why customers love the product and love the team. Mode is enabling everyone from scientists to CEOs to ask five whys – and then some. In the process, they’re using data not only to take better-focused snapshots and tell more powerful stories, but also to draw a more detailed map and plot the best way forward.