En-Route Search @ Bing
In 2012, I worked for Microsoft Bing. There, I used user-research and competitive analysis to identify an opportunity to increase usage of their maps service and develop an advantage over a competitor.
The feature was called En-Route Search. I pitched the idea to top-level managers at Bing and two years later it was released to the public. Soon after, it was quickly mimicked by Google Maps. I think the feature makes a lot of sense for users, and is representative of the product management process at larger companies. Here's how it came to be.
Most good ideas come from user research. I started this one by interviewing 20-30 google and bing maps users and sending out a survey to a randomized sample of 200 participants via Amazon's Mechanical Turk service.
- User Research: Questions about scenarios in how people currently look for businesses, use maps providers, and plan their trips
- Mechanical Turk: Used to quickly and cheaply compare interfaces between Google and Bing and test mock-ups of new designs
Some of the unmet use cases that came out of the user research were:
- Ability to combine multiple modes of transport: e.g. drive to a bus stop, and take a bus. (analysis: not a big enough pain point)
- Ability for public transit users to optimize for running errands along transit lines. (analysis: this is a big friction point: public transit users are constrained to transit lines, but the target audience/impact is small)
- Ability for users to find places to stop on road trips, weekend trips, and longer drives: e.g. amusement parks, gas, fast food. (analysis: infrequent use case)
- Ability for users to combine multiple errands into one flow: go to bank, get groceries, pick up kids, etc. (analysis: this is the most common, and high potential case)
- ..and more
Comparing the user flows for opportunity size and tying them back to company goals, it was clear that the ability to find extra stops along the route would be a winner for both maps users and the company. It would allow users to search for more businesses while planning their trips and provide a key competitive advantage for Bing Maps over other providers.
- # of businesses visited per day/month (growing user and business value)
- # of active users on Bing Maps/Local (competitive dominance)
Inevitably, good ideas face setbacks in getting them built. Here are the ones I faced and how I overcame them:
Technology. It's very computationally-intensive to figure out what relevant businesses exist along a user's route. Maps queries are usually performed with a given radius, and depending on the length of the route, you would have to execute 10-100 (or more) queries to find the relevant stops along a route.
- I tackled this one by working with an engineer on the maps team to develop a better heuristic/algorithm for querying the maps data. We balanced the constraints of the existing API, and made engineering-product trade-offs to find something that worked for the user, and made it computationally feasible.
Vision. It's one thing to explain an idea to someone, and another to show them a prototype. I knew that this idea would be a hit with the executives at Bing if I could give them a tangible prototype to play with. The more fleshed out and tangible it was, the higher our chances of success
- I teamed up with the same engineer, Emmanouil, to build a prototype. I got my hands dirty and made the front-end work while he did all the heavy-lifting and implemented the back-end algorithm.
- Then I set up a pitch to Bing executives. In that, I gave them access to the app and asked, "have you ever tried finding a coffee stop on your route somewhere?". With that one question and the ability to look up a Starbucks on our prototype, we got their buy-in.
Timing. This last one there wasn't much I could do about. There were many features already planned for Bing maps. Thus even though we had good background research and a working prototype, our project took a backburner until the existing items were implemented.