Bridj

Soon-to-be-expat monthly Bridj user here, feeling compelled to do a topical text dump this morning from a user perspective. I first started using it last winter when it was better by a magnitude than suffering through the Green Line, and often cheaper in the off peak too (prices have since been adjusted up, from $3 off peak to up to $5 during peak). I only signed up for a pass a few months ago, but was really hoping to ride it through this coming winter at least. Very disappointed that they killed off the program. The fact that it's undercutting the price of a LinkPass is insane -- it should probably be priced more like an inner express bus pass for the value it adds, but I assume they thought that a dramatic price hike would be worse for their customers to swallow than simply cutting the program and starting fresh.

That said, the pass system is far from ideal or user friendly, and I get why they're killing it off in its current form. The guaranteed comfortable seat and Wi-Fi are huge, but here are a few things they should be trying to fix. One sentence summary of the long post below: Their app experience has a long way to go for a company billing itself as an innovator and pulling down decent VC money to boot. To be fair, they've had the "beta" tag on the monthly thing this whole time. But in several key respects including fare payment and (oddly enough) technology, their experience is actually inferior to a comparable MBTA experience.

1) A monthly pass buys you an app promo code that you have to type in before you book each of the rides you take every day, morning and night (max of 40 rides per month). The app saves your credit card info, and if you hit the button five pixels lower or just have a brain fart instead of typing in your hopefully memorized, case-sensitive promo code, you end up inadvertently paying for a one-time pass on top of your monthly membership. It's only happened to me once or twice, and they're good about refunding, but the app should just register the user as a monthly pass holder and dispense with other payment methods during the membership period.

2) Instead of saving your home and office locations, the app presents you with a street map and asks you to manually select your departure and arrival points every time you book, before presenting you with a list of possible departure times and locations. It's home to office, office to home 95% of the time for me, but I can't just tell the app that. Even the T's commuter rail app remembers the recent trips you take, but not Bridj.

3) I've also come to find out that if I select the street corner diagonally across from where I live, I get access to twice as much frequency by way of routes that Bridj otherwise thinks are too far away for me to comfortably reach on foot (still less than 15 minutes, though). You'd think that on a given morning you would be able to pull up a system map and find out where Bridj will be leaving/arriving, but no such thing exists. I just have to assume that Bridj knows best when I tell them where I am, but it turns out they don't always know best.

4) Bridj has GPS tracking on its fleet, but it's not as easy as opening the app and seeing where the vehicles are. You don't get access to GPS tracking until after you book your trip. So if a vehicle happens to be running late, you might not find out until after you book. They are starting to include alerts about late buses, but as far as I can tell these are being added manually by Bridj staffers and are far neither comprehensive nor all that reliable. Again, with the T, I can download any of 100 apps to my phone and find out exactly where everything is, even the blessed Green Line at this point.

5) In the press coverage, they've been talking about how the monthly passholders have been causing rush hour runs to fill up, and that's true in my experience on the morning rush. If you don't book by 6:30 a.m. these days, you're not getting on board between 8 and 8:45. But recall above that I said a monthly pass entitles you to 40 rides per month. I and others I know who ride it frequently stay late enough at the office that Bridj doesn't work for us coming home. So monthly passholders often have rides to burn, and there's nothing disincentivizing us from booking an 8:30 a.m. when we leave the house, then an 8:40 a.m. if we stop to get coffee, then an 8:50 a.m. if there's a long line at the Starbucks. I don't have a good answer for this one, except that it's frustrating that their evening runs come to a hard stop by 7:30 p.m. when lots of us are still stuck in the office. A half-hourly or even hourly route scheduled between 8 and 10 p.m. would make a killing.

Anyway, despite all of the above, I'm bummed that I'll be back to a monthly T pass next month, but I'm simply not interested in paying what it would cost to go completely a la carte with both Bridj and the T (easily $150). I'll probably treat myself to an inbound ride once a week or so -- you almost can't put a price on zooming by packed Green Line platforms on a cold/rainy/snowy morning. Then hopefully I'll jump back aboard full time on a reinstated pass program. That's assuming they are actually getting somewhere on the business side and not just burning cash, an impression you can certainly get on their empty shoulder runs. Fingers crossed.
 
Thanks for the write-up, Ernie. It sounds to me like Bridj would be better off dropping all the "real-time routing optimization" nonsense, pre-bookings, etc, and run like an actual bus route with scheduled pick-up times to known destinations. First-in-line, first-on. Demand in that case will self-regulate to supply, but they can still scale up supply to meet demand the old-fashioned way by sending more buses and making scheduling adjustments (rather that the VC-$-backed GEEWHIZZ APP COOLNESS DATA ANALYTICS DATA WHOA method).
 
Kiva Systems robots operate in a total "walled garden". Safe inside a warehouse, with people segregated from the allowed robot pathways.

Their capabilities are nowhere close to the safety systems required for open road operations in mixed vehicle, people, bike traffic.

Exactly. Kiva warehouses have "human-free zones", require you to put QR code stickers on the ground which they use for navigation, mostly operate independently of each other, but use algorithms that allow them to cooperate since they have the same algorithm, and they do have a few actors centrally controlling. They also primarily optimize moves for (a) parking of inventory and order pods and (b) occupying workers, so their plans for movements have completely different objective functions than for a commuter network.

Literally, the only similarity to open road operations are that one might use algorithms for optimization.
 
The market probably does exist for a dumb bus like you describe, Shepard, but the whole "innovation!" piece is Bridj's raison d'être. Just look at what a sore thumb they are (or would be if listed) in the portfolio of reported backer Atlas Venture. VCs like that literally want to cure cancer. They don't want to back just some other flavor of a Boston Coach -- that's stuffy old Fidelity's job. I think Bridj will have to innovate their way out of the proverbial paper bag in order to keep doing their thing. The stuff that they brag about re: data, optimization, etc. turns out to be very man-behind-the-curtain once you experience it every day, but I do think it can be improved upon. In a nutshell, they need to do a better job of giving front end users the same kind of access to the locational and scheduling data that supposedly powers their back end. The balance of my post didn't make it seem this way, but I wonder more about their ability to survive financially than their ability to improve the product, to the extent those concerns are not intertwined.
 
The market probably does exist for a dumb bus like you describe, Shepard, but the whole "innovation!" piece is Bridj's raison d'être. Just look at what a sore thumb they are (or would be if listed) in the portfolio of reported backer Atlas Venture. VCs like that literally want to cure cancer. They don't want to back just some other flavor of a Boston Coach -- that's stuffy old Fidelity's job. I think Bridj will have to innovate their way out of the proverbial paper bag in order to keep doing their thing. The stuff that they brag about re: data, optimization, etc. turns out to be very man-behind-the-curtain once you experience it every day, but I do think it can be improved upon. In a nutshell, they need to do a better job of giving front end users the same kind of access to the locational and scheduling data that supposedly powers their back end. The balance of my post didn't make it seem this way, but I wonder more about their ability to survive financially than their ability to improve the product, to the extent those concerns are not intertwined.

I think them not being listed on Atlas's website has more to do with Atlas splitting into Atlas for life science and Accomplice for technology investments.
 
I have not used Bridj, but based on my limited experience with the app, I found it very confusing and not really that convenient. I want to see the routes and the times. When I put in a start and a destination, it picked out a route that wasn't that close to me and not around the time I wanted to travel. And it is my understanding that the routes change over time. Without periodically going and checking to see if Bridj now serves a need of mine, how would I know? It just seems like a case where technology is making it MORE complicated for the end user, rather than simpler.
 
I have not used Bridj, but based on my limited experience with the app, I found it very confusing and not really that convenient. I want to see the routes and the times. When I put in a start and a destination, it picked out a route that wasn't that close to me and not around the time I wanted to travel. And it is my understanding that the routes change over time. Without periodically going and checking to see if Bridj now serves a need of mine, how would I know? It just seems like a case where technology is making it MORE complicated for the end user, rather than simpler.

Yes, the ideal of Bridj is that with a big enough market you get dynamic and optimal route planning for time and value. So somewhat like Uber you move towards a more flexible system where supply adapts to demand dynamically. So instead of a political process via the MBTA to establish routes you get a customer driven one. Just type in where you want to go and when you want to get there and see the cost of various options.

But it is hard to bootstrap that kind of system. Uber is really the leader for on demand transportation. Bridj needs to get bigger to have enough route options to make sense as something more people use. And if people are going to use it for end to end transportation planning then whichever system emerges should incorporate all the options and not just their own vehicles.

Really Google Maps or Apple Maps would be the best way. Just enter a destination and have a button that includes Uber or Bridj options and allows a hybrid of alternatives
 
Has Bridj ever considered offering different routes on weekends? Particularly on Sundays, MBTA bus headways are absolutely miserable. And people have very different travel patterns than during the week. Bridj could make a ton of money just offering service in these corridors.
 
Exactly. Kiva warehouses have "human-free zones", require you to put QR code stickers on the ground which they use for navigation, mostly operate independently of each other, but use algorithms that allow them to cooperate since they have the same algorithm, and they do have a few actors centrally controlling. They also primarily optimize moves for (a) parking of inventory and order pods and (b) occupying workers, so their plans for movements have completely different objective functions than for a commuter network.

Literally, the only similarity to open road operations are that one might use algorithms for optimization.

Dwash -- each of the features you noted were optimizations for the first gen of such a system

The key points that are relevant are the way the units are scheduled for the pickups to deliver the goods to the "picker" and the schedule and routing is optimized in real-time as the units are tracked

A Kiva-Like Jitney would know where all the customers were located [from their phones] and where they wanted to go entered through the Ap [sorry No "follow that car"]

Schedules and routing could be computed using real-time traffic information {Waze-like} and ETA's developed with options presented to both empty cars and also cars carrying a customer who might be presented with an option stay on the current route or get a discount to pick-up another customer
 
Bridj is winding down.

I'm a bit surprised; it seemed they found some good routes that are definitely underserved by the MBTA. Pour one out for them, their demise seems more the result of the fundraising environment and demand for growth than the economics of their service.
 
I be the more than anything uberpool killed them. Probably slightly more expensive than an expensive bus but door to door service.

Edit: I wonder if they looked into more low income, high employment routes. They seemed to go from indirect pockets of wealthy enclaves, but what about like Dudley to Logan. That's 60 min and two silver lines. Chelsea to longwood. I bet employers would've given general workforce home area, plus cheaper marketing, and maybe some tax deductible employee benefit fund to help costs.
 
Just another periodic reminder that all of these "new, innovative, future-of-transportation" platforms have worse farebox recovery ratios than the T.
 

Back
Top