HotelTonight - story of a modern data-driven business.
Jan 20, 2014
Colin heads up the data and BI teams at HotelTonight, a leading mobile app for booking hotel rooms.
At HotelTonight, our big data challenge revolves around trying to get everyone in the company engaged in looking at data. We know that data helps us make better decisions as a company. The key is that we can’t be top down about it. We can’t have leadership telling everyone that looking at data is important.
People need to be motivated to use data every day. And they’ll be motivated if they know they’ll make better decisions when they’re backed by data. When they can be entrepreneurial, figure out problems, and fix them on their own. Because data is where we find the problems and where we find the answers.
For us on the data side of the house, the challenge is to make data accessible. We use Looker to build everything from dashboards to analytic services, so that every single person in the organization can pull data. What’s great about Looker is that people who don’t even know what SQL is can start digging into the data themselves and learning about what’s happening in the business.
This chart shows Looker usage inside Hotel Tonight. (The dips are weekends.) I can see that about 50 percent of the company is logging into Looker almost daily to look at different analyses. That’s something that you absolutely don’t see other places.
This is the single most important piece about Looker for me. It’s about understanding how corporations, at least the ones I’ve participated in, interact with data and why it’s different with Looker. In the classic model, you’ve got a data team. They’re defining key metrics. Those metrics get passed to engineers who build dashboards. The team members then interact with those dashboards.
At HotelTonight, our data analysts can pass key insights and models directly to team members. If we discover an important insight, we share it with business decision-makers and they can start applying it immediately. Because Looker is web‑based, it’s as simple as sharing links across the organization. It might be a link tied to an email where we say, “This is a new way of viewing things.” Or as straightforward as sharing a URL of a news article on the web.
What’s also great is that we get these new purple connections between different team members. These might be two account managers, or someone on the finance team and someone on the account management team, or an engineer and an account manager. Looker allows them to either take reports that were given to them by the data team or to develop reports on their own. They can start passing reports among themselves.
An example of this is one of our account managers in New York. Over time, she’s gotten very comfortable with Looker, and—to try to understand her city—she’s created most of the reports on her own. Now the rest of the market management team is using those same reports to understand their cities. It’s what’s enabled by a web‑based service that’s super lightweight and very intuitive to everyone, but on that can still do all the heavy data analysis.
At HotelTonight, there’s a relationship between people and data that you just don’t get in other organizations. And that’s exactly why we’re so excited about Looker.