3. Fix the database first

Time and again, we’ve heard the instruction to clean up our data before building apps that people will actually use. Likewise, we’ve been told that the back-end data model must be fixed before the front-end digital experience gets built.

While I don’t wish to discourage anyone from cleaning up data or fixing a data model, insisting on doing this first can actually prolong your problems, or even make them worse. Why? It’s the users of the software that will do the best job of scrubbing your data. With a properly designed user experience, you can actually crowdsource data management transparently, without forcing it.

As one compelling example in the B2C space, consider Waze, a GPS navigation software app owned by Google. By leveraging smartphones and tablet computers that have GPS support, plus user-submitted travel times and route details, everyone gets real-time reports on accidents, alternate routes, road hazards, and more. Waze crowdsources data input to deliver better information than a centralized solution.

In the B2B space, a full-stream oil and gas company customer created a crowdsourced human resources application that would provide up-to-the-minute information on employee and contractor reporting structures across more than 35,000 employees in 100 different countries working on hundreds of different projects. The users maintain the data as part of their normal workday, without any heavy lifting on the back end or top-down data quality reinforcement. The crowd cleansed the data in less than a week.

Transparent crowdsourcing of data cleansing has been woefully underused in the world of enterprise software. That’s because in many companies, data quality is still considered a mechanical exercise for data analysts or data entry personnel rather than a baked-in process of day-to-day engagement with an application that people actually want to use to get stuff done. When people want to use apps because they deliver high value within their role, context, or process, the data gets scrubbed almost magically.

In regards to fixing the data model first, we have seen far better results from fixing the front-end digital experience first, then working back toward the data model, only making changes that are absolutely necessary, driven by what people need to see and do in the application.

Digital transformation: Mostly about people

Digital transformation is a misnomer, because it is mostly about people. There is no such thing as a successful digital transformation that did not begin with a human-centered approach.

First, align your stakeholders around shared pain points and shared goals. Second, deeply understand the people who will actually use the software you intend to create, creating inexpensive prototypes to test your assumptions. Third, analyze how your people actually use the products you build, and iterate like mad to get the software out of their way.

When a digital experience adapts to match people’s human experience, magic happens. When digital technology feels like a natural extension of who we are and what we need to accomplish, it actually gets used, creating a transformative, positive impact for everyone.

[ Get answers to common digital transformation questions and lessons from top CIOs: What is digital transformation? A cheat sheet. ]

Pages