In the four years that we’ve been running Transit, we’ve learnt that people do not mess around when it comes to their transit info. Tell somebody that their bus is on time? They’ll love you forever. Tell somebody that their bus is on time — but no bus shows up? They’ll tie piano wire between their doorposts and invite you over for dinner.
Having to deal with angry tweets and one-star reviews is pretty much the definition of a “first world problem”. However, transportation apps like ours get an unusual amount of flak from our users, and for good reason: if the information we provide is inaccurate, then we have likely seriously jammed up your plans.
We aren’t the only ones getting this kind of criticism, of course. Any app that displays transit data (like Google Maps or official transit agency apps) are often badgered for showing bad data.
Initially, this feedback made us angry. Why? Because we were getting criticized for something we had absolutely no control over: We’re a transit app. We’re not a transit agency. We display the best information that we can find, which is supplied by the transit agencies we’re partnered with. People were shooting the messenger — and then throwing him under the proverbial bus. “Would you get mad at Twitter because a transit agency made a mistake in a tweet?” we wondered.
But you don’t want excuses. You want to know when your bus is coming. When we launched Transit 4.0, we promised that we had plans to improve the precision of our transit data. And as it turns out, you’re the answer to your prayers.
Today, we’re proud to announce the first ever (not totally terrible) crowdsourced transit tracker. We’re rolling out our crowdsourcing feature gradually, starting in two cities: Montreal and Victoria, BC, but with plans to expand our offering to NYC, SF, Boston, Chicago, LA, London, Paris, Rome, Toronto and many others.
So how is crowdsourcing going to affect your transit commute?
- In some cities, there’s simply no real-time transit information. The only departure times that transit apps can show you are based on official transit agency timetables (and, as our Twitter stream will attest, those aren’t always the most reliable).
- In other cities, there is real-time information provided by transit agencies — but predictions can be wildly inaccurate. (What’s the use of “real-time” data if vehicle positions are only updated every 5 minutes?)
Clearly, there’s room for improvement — and we had to be the improvers.
You see, people like Transit because we have a beautiful interface, an incredibly speedy app, and features so extensive they’ll charm even the nerdiest transit nerds. But if we don’t have the best data it’s all for nought. Why use a transit app if its departure times are off by 5 minutes 50% of the time?
So instead of waiting for transit agencies to develop/improve their real-time systems, we came up with a crowdsourced solution. Crowdsourcing transit data can overcome a lot of the pitfalls of bad/inaccurate/jerky agency data, but it’s hard to get right. Many tech companies like to hype up their fancy “crowdsourcing” projects, only to realize, whoops! We don’t actually have enough users to harvest the data from. So what are we doing differently?
If you’ve been paying attention over the last few months, you know we recently released a powerful transit navigator called GO that gives you real-time updates along your trip. GO reminds you when to leave for the bus, when to get off, and if you need to hurry to make your connection — so if you’re a chronic procrastinator, narcoleptic, or just somebody without a clairvoyant sense for where your bus is: GO is a game changer.
Now, in addition to telling you when to leave for your bus, it’ll also track your bus location, broadcast it to our server, and predict accurate departure times all the way down the line.
Dat’s right: with just one person running GO on each vehicle, we’re able to produce accurate real-time predictions for the entire transit network.
Better yet? We don’t need you to do anything extra. You don’t have to take training or wear a lanyard that says Transit Data Specialist™. You just have to keep on using GO, which tens of thousands of you already do every day. When you use GO, we receive your exact bus location every second, which means ETAs get recalculated in real-time instead of every 1–5 minutes.
See how smoothly the vehicle icons move on the map in contrast to agency-provided real-time? It’s like tracking your Uber. No big lags. No big jumps.
We’ve designed our crowdsourcing feature to appeal to your best, benevolent self. As you board a vehicle with GO running, we show you exactly how many people you’re currently helping. If you try to quit before the end of your trip, we’ll let you know how many people you’re letting down. (And yes, we’ll make you feel like a stiff if you don’t help your fellow riders.) But you’ll also feel like a freakin hero every time you push GO.
Density density density. User density is key. We have more riders than any other transit app in North America, so we’re confident that our crowdsourcing initiative will work. After all — we’re not trying to build a dense community around the concept of crowdsourcing. We already are the community.
Montreal, Victoria, then the world…
Crowdsourcing goes live today in Montreal and Victoria, BC. We’ll be ironing out the kinks, flapping out the creases, and exterminating all the inevitable bugs before launching crowdsourcing in every city. But why these cities to start?
- Montreal: it’s our hometown, we have good rider density here, and it is the largest city in North America without any real-time data. (In the land of the data blind, the one-eyed transit app is king.)
- Victoria: because there is also no real-time data here, our rider density is obscenely high. On a given day, more than 7K of the 35K unique daily riders will open our app (or 6 people on every bus) — and rider density is the #1 determinant of whether a crowdsourced data initiative will succeed or fail
Trust us: we know crowdsourcing real-time transit data is hard — and we know a chest-thumping blog post won’t change that. But we feel we’ve put together a solid recipe: a killer navigation feature that enables passive crowdsourcing, a thoughtfully designed experience that incentivizes altruism, and millions and millions of active users in concentrated locations.
So today we switch on crowdsourcing in Montreal and Victoria. We’ll be closely monitoring how it’s used, and how accurate it is. Then, once we’ve nailed down GO crowdsourcing in our pilot cities, we’ll start rolling it out in other real-time deficient places. And then in the places where real-time sucks. And then everywhere else.
Welcome to the future. And thank you for choosing Transit. Now, get ready to launch GO, because you’ll probably make someone’s day. Then, on your way home, someone else is gonna make yours.