What’s next for Citymapper? A bus?!

Loyal readers know by now that Citymapper is one of my all time favorite apps. It offers the best public transport route planning I’ve seen with delightful features such as a recommended car to be best situated at the end station and what exit to use from the station that’s closest to the final destination.

Citymapper GO: accompanying users on every step of their ride, including a “get off” alert.

Last month, using New York’s Subway, I discovered a new feature, the GO button, that breaks down the route into easy-to-follow steps. Each step has its own detailed description, map, and a notification card, visible from the lock screen. The part of the journey that includes the ride also has the option to receive a “get off” alert, which is a great feature for tourists. Citymapper also added a GO dashboard to keep a tally of calories burned and trees and money saved per trip vs driving a car. The gamification was a bit redundant as a feature but cute nevertheless.

Citymapper GO dashboard: the complete journey from point A to point B.

Little did I realize that one reason for the new feature to enable data collection based on real travel. By knowing full routes, not just when a rider got on a bus or train, Citymapper can better understand actual transportation needs, not just the compromises riders made due to route availability. With this more accurate data, Citymapper hopes to optimize public transportation routes when data shows there is a need. It can also show what new routes might work. To test that hypothesis, it launched a short bus service in London today and said “while the service will not get passengers very far, it is seen as a test of the technology that could lead to something much bigger.”

The implications of such a service are interesting. First, the flexibility and dynamic capabilities: transportation authorities could decide spontaneously that they want to add a route based on an event that they didn’t know about or a problem that they didn’t anticipate. They don’t have to reprint the transit map and hang signs, they only have to integrate with Citymapper so that the app can suggest the new route to travelers in need. For users, the trust in Citymapper is such that if the app suggests a route, it must exist. It’s a much quicker way to set up alternative and temporary transportation when things go wrong.

Citymapper said it best: “We’ve helped people figure out which bus to take. When it arrives. How long it takes. When to get off. Now it’s inevitable that we help make them work better. We don’t have to do it all ourselves, we’re glad to partner with others. We built an easy to use app by being users ourselves. So we feel the best way to build software for buses is to run buses ourselves. And learn from running some public experiments.” The first part I agree with, they’ve built a better app for existing transport systems than any of those transport systems (I’ve tried New York, Bay Area and Berlin.) Whether they need to be the ones to run buses remains to be seen, though I appreciate that they are running these London ones to prove a point.

Bus routes are based on historical data and political decisions. While transport authorities might have current usage data for existing routes, they don’t know much about what routes outside their system riders actually need. That’s why we see companies running private shuttles to and from their offices – those routes were needed and were not served by existing public transportation necessitating private vehicles. Just as a point of comparison, our local transit agency, the VTA, is also planning a route overhaul. They’ve asked for input from college and high school students, commuters and other community members.  They’re having community town halls and trying to get feedback from as many entities as possible about the changes. This is honorable but falls short of getting actual usage and, better yet, desired usage data such as Citymapper has.

Can Citymapper leverage its fantastic app to make a real difference in how city transportation works? Can they anticipate need and help cities plan accordingly? Will they tie their system into private transportation such as the carpool options on ridesharing apps and help everyone travel faster? Regardless, it’s a very interesting first step and I’m waiting to see how it works in London and other cities around the world and what cities will embrace Citymapper’s help.

AR for the rest of us

“Making the camera the first AR platform”
Source: Facebook

Augmented reality (AR) touted at Facebook’s F8 with a place of honor on Facebook’s 10 year roadmap as one of three technology areas Facebook is going to focus on. Mark Zuckerberg stated in his short keynote that Facebook is “making the camera the first augmented reality platform” which, with their strengths in machine learning and the social graph, might make for some very powerful tools. That said, there are many smaller, more focused applications that are better suited for widespread adoption of AR, even on today’s hardware.

Take, for example, the Vivino app. I know very little about wine, but I do appreciate a good glass every once in a while. Vivino gives instant access to wine rankings for almost 11 million wines from a community of 23 million users who care about wine enough to rank and write reviews. Their wine list implementation is extremely helpful. All users need to do is take a photo of the list and the app provides rankings for each wine and a handy color scale.

This is one of the first times I have seen a useful implementation of AR, aside from Google’s Translate app, and I started thinking what makes an AR app practical?

  1. Immediate: does it use the phone camera, with either a live view or a photo or does it require dedicated hardware? A phone is much easier to access. Also, how many steps are needed before the information added?
  2. Saves time: does the app replace a search or, even better, several searches? Does it automate data entry by recognizing the text? Does it replace a task often done on the go? For me, that’s usually searching undecipherable menu items in hipstery restaurants.
  3. Visually simple: is the additional information is presented in a way that isn’t too complex to understand at a glance? It shouldn’t be a complicated infographic, but a few additional data points.
  4. Adds value: does the added layer of data add value? Does it provide actionable data? Too much data or irrelevant data can be a nuisance, especially for AR apps on the go.
  5. Doing the math: can the additional data be manipulated in a way to provide more value? For Vivino. there was a suggestion to calculate points per dollar/euro so that users can quickly choose the best wine their money can buy. For other applications, say a grocery app, a photo of products can provide initial value with a health ranking (similar to Fooducate) and a cost per serving.

Finally, one of the cooler uses of AR discussed at F8 was the mesh between facial recognition and the social graph, where, via a pair of AR glasses, names pop up above people relevant to you in a crowd. For people such as me, who are better at faces than at names, that would be amazing. Till then, I’m hoping to discover a few more practical ones.


Facebook Spaces: what it is now and what it could be

This week at Facebook’s F8 conference I had a chance to try out Facebook Spaces, their new virtual meeting place for groups. As a non-gamer and someone who has only tried VR in passive, sit-back-and-enjoy-the-show applications, I was curious to see what a “VR for everyone” product looked and felt like.

My avatar taking a selfie in a virtual bazaar in Facebook Spaces.

At F8, Facebook set up a two-station approach to trying out Spaces. The first one was one to set up a Spaces avatar and link it to my Facebook account. This was done via an Oculus Rift headset and Touch controllers. The setup was about as clunky as any other avatar setup with menus of everything from eye and hair color to hair style and glasses shape. Facebook did add a nice shortcut by asking me which of my Facebook photos of myself I’d like to use as a base, so that right off the bat my avatar started with brown hair, glasses, and more or less my skin tone. On to station two, the actual “space.”

Three participants trying out Spaces at F8.

For demo purposes, Facebook put together stations with two guests and a Facebook “host” to walk us through the features. Again we donned the Rift headgear and Touch controllers and once in the VR space, our host walked us through some places she had 360 VR videos of, which were stunning, and we were able to take selfies, scribble in the air, and talk to each other. It was fun and in some ways I can see it becoming some sort of place to meet.

That said, Spaces doesn’t compare favorably to its main “bringing people together” competitor, the video call. Video brings an instant immediacy and intimacy to communication that Spaces doesn’t match, not that it tries to. It’s also much easier to access. Spaces is supposed to be fun, a place for friends. Yet it’s exactly this coolness factor that has the danger of becoming boring and mundane after a few sessions.

Can it catch on? Video calls a decade ago required users to add a camera, microphone and speakers to their desktop and even many laptops did not have them built in. Today they are on our phones, part of almost every messaging app, so that a high quality video call is relatively accessible with a mid-market phone and good connectivity. That said, the total cost video call hardware was never as expensive as an Oculus Rift at close to $600. Spaces also has the disadvantage of being a more expensive platform and, hence, will be more exclusive, with less users able to participate. The question is will our fascination with VR hold on until the tech becomes cheaper and more intuitive to use.

Finally, I see a different potential application for Spaces: education. The VR immersion in different worldwide locations was beautiful. Imagine that in an application  with a teacher/guide taking students around the world, in space, inside the human body, to places they have never been. Then imagine that for people without mobility, going places that they couldn’t travel to in life. That’s where Spaces has the opportunity to do something amazing and become more than a cool game.

Toll Roads – adding another dimension to Google Maps

Yes, I’m going to talk about Google Maps again. I know, I’ve been writing a lot about travel apps, but they are one of the most popular mobile apps category and a road trip in new places always surfaces new needs.

Our road trip took place in the approximately 250 mile distance between New York City and Washington DC. Our way had multiple toll roads, bridges, tunnels, and pay-to-use express lanes. All were conveniently paid for by an in-car transponder, the EZPass which we had to load with funds every few days. Our tolls ranged from a high of $11.52 to cross the Verrazano–Narrows Bridge to $0.50 for a section of the New Jersey turnpike. That said, it’s not always obvious on the road how much a certain highway, express lane, bridge, or tunnel actually costs and I was quite surprised at the variation when the fees appeared on my EZPass bill. All our routes were decided on by Google Maps, where the goal is clearly to minimize travel time.

The system of toll roads in the mid-Atlantic states present a different challenge to route planning. It’s not just about the fastest way to get from point A to point B, but could also be about the cheapest way. The only way Maps deals with the cost issue is by allowing users to avoid highways, toll roads or ferries when planning a route, which unfortunately ends up being too crude as completely avoiding toll roads for a longer journey creates unnecessary and long detours. When the way between point A and B has multiple, crisscrossing freeways connecting it, where some are toll roads and others not, some have pay-to-use express lanes and some just regular HOV lanes, finding the best route shouldn’t be just a question of time. Also consider that that the toll highways and express lanes will almost always be less congested than a parallel highway, so perhaps the Maps algorithm prefers them right off the bat.

So, how can Maps help users with toll roads?

  • Google Map, three alternate routs from JFK Airport to Philadelphia, all with similar times, all on some toll roads. Labels currently include travel time and toll road icon.

    Step one: include cost per route. There are already tags on routes to denote travel time on the selected route and additional/less time on alternate routes, so users are already used to looking for that additional information. The tags currently have a “this road includes tolls” icon, but no cost breakdown.Also, it’s not always obvious on the road how much a certain road, bridge, or tunnel actually costs and I was quite surprised at the variation when the costs appeared on my EZPass bill. 

  • Step two: offer more route options. Currently Maps offers two or three but that doesn’t seem like it would be enough to offer users the combination of time and cost that make sense to them. Citymapper, for example, offered 17 different subway and bus combinations to get from Bryant Park to Hamilton Grange in New York under groupings that include 5 routes under the Suggested header, 4 under the Bus Only header, 4 under Subway Only, and 3 under Rain Safe. All include travel time and cost. So while it is easy to just pick the top suggested route and get going when in a rush, it’s also easy to take a quick glance at all options when there are additional concerns.
  • Step 3: be smarter. The overarching directive to find the fastest route doesn’t consider possible compromises when a user is willing to, say, spend 5 more minutes on the road but avoid a $10 charge. Because that tradeoff can be individual, Maps will need to offer a list routes, not just two or three “best” ones. It also needs to start learning from user behaviour to start prioritizing the routes that are a better fit. These are exactly the on-the-fly tradeoffs in cost and time that machine learning can do best: solving problems that are “easy for machines and hard for humans

The interesting question is why hasn’t Maps done this up to know? Is it a product blindness due to lack of exposure? Up until about a year ago, the San Francisco Bay Area had tolls only on the bridges and recently added pay to use express lanes, so I wouldn’t have thought to add cost to routes because not taking a bridge in the Bay Area is a significant detour. Or maybe cost just isn’t an issue for most users who just care about time? Though even without adding route options and balancing speed and cost, Maps should at least provide the expected tolls. Just that would help tourists immensely (and I bet a few locals, too!)

What’s your favorite app?

It seems like an innocent question, right? One that you could answer after a second or two of thought? After all, you use your phone every hour, every day. Surely you know what your favorite app is, right?

I gave that question some thought this week as a product management exercise. As exercises go, it provoked more questions than answers. Picking a favorite app is not so simple because there several factors that don’t always align. Here are those I considered:

  1. How often do you use it? There are apps that I use every day, several times a day such as Google Maps. I couldn’t live without it and need it, especially on my phone but it’s not my favorite app.
  2. Is it a way to access another service? I use the Twitter, Facebook and Slack apps every day to see what’s going on with my friends and catch up with current events. Yet those are benefits of the service, not the app. Lyft might be a necessary app on your phone, but it’s not about the app, it’s about the service.
  3. Does it provide access to utilities? Our phones allow us to answer email, manage our calendar, and call and text our contacts. Yet these, too, are a service and don’t often become favorite apps. We need them on our phones because need to access them when we are away from a computer.
  4. Does it enable access to content? Media apps allow easier access to their sometimes paywalled content than a mobile browser. Some are better than others and do a great job aggregating and presenting their content, but the app is still defined by that content.

The Nazareth Iris photographed in the wild by my mom and shared on WhatsApp

So what’s left? Maybe it’s time to channel Marie Kondo and ask what apps spark joy? Alternatively, what apps make our lives easier and, in doing that, cause delight in unexpected ways, be it because of efficiency or the information or service they provide. Finally, do we really use them or are they a sort of a critic’s choice: fancy, well designed, but don’t provide a clear benefit. So after a lot of thinking, these are my favorite apps:

  1. Citymapper, the public transportation app. Citymapper provides a service that other apps do, such as Google Maps and Moovit. It tells you how to get from point A to point B using public transportation, but it does so in a way that outshines the competition. Citymapper has become indispensable on trips as a source of information, but, perhaps more importantly, it completely changed my behavior while on a trip. Instead of begrudgingly taking out a map and trying to figure out stops, or, more likely, switching between a map app to find stations and a public transport app to find train routes, or hailing a taxi just to get there, I now embrace subways, metros and undergrounds. It’s easy to find a station, to find the right train and platform at the station, and to get out at the best exit at the destination. Unlike Google Maps, which offers one or two different transportation options, Citymapper offers 8 or 10 (in New York, at least) along with weather-safe routes, walking, biking and ridesharing options.
  2. WhatsApp, Facebook’s messaging app. Similarly to Citymapper, WhatsApp provides a service many, many other apps do: a platform for sharing content with your social circle. Like Citymapper, it manages to outshine the competition in one, unexpected way: it got many of friends, who had never shared anything on any other social platform before, share photos and text and participate in conversations.  I credit this newfound sharing capability to WhatsApp’s instinctive UI, which makes it extremely easy to post content, coupled with the security of posting to a closed and known group. There is no unintended sharing to larger groups on WhatsApp or a chance of the content being reshared and I think it is this limitation that has prompted my social circle to share more often.

So what’s your favorite app?

Maps: a little feature that makes a big difference

My San Francisco, with way too many yellow dots to be useful.

Google Maps has launched a new, smallish feature has been on my list of feature requests for a long time. It consists of giving users more options than a yellow star when saving a place. I love this feature because, as you will see on the right, I save a lot of places. The problem is that saving so many places renders this feature practically useless – who can see the forest for the trees?

Place saving options

Google decided that the new categories to mark special locations, other than a yellow star, are Want to Go and Favorites. Other than those three, users can add their own, custom titles for new categories, called lists.

What I liked:

  1. Different colors for different categories allow users to better identify spots on a map without clicking each one for more information.
  2. Custom lists can save different types of places such as Bakeries, Brunch Spots, Good Hikes, etc. This flexibility is welcome.
  3. There are also three other types of locations with special marks:
    • Events and appointments on the map along with their time and date, which disappear after the event is over.
    • Addresses of select contacts in gray.
    • Spots of interest including photo-worthy tourist stops (marked with a brown camera) and landmarks like museums and important buildings (marked with a brown castle/tower.)
    • Spots that I’ve searched before without saving, though I’m not sure what the criteria for these places are yet.
    • Stores are occasionally marked with a red shopping bag but like other spots, it’s not clear what the rules are for marking them.

All in all, it makes it much easier to see what’s nearby or what’s close to my planned route. That said, I do have a few minor suggestions.

  1. “Want to go” and “Favorites” are two sides of the same coin: the former is before you go, the latter, hopefully, after. Are users supposed to go back and switch the marks after a positive visit? Perhaps Maps can prompt this change after a user has visited the place?
  2. No matter how many extra categories a user defines, they will always be marked with turquoise on the map. This makes the customization useless after one category. Can we have more colors please?
  3. Google can read my appointments from Calendar and email but (understandably) cannot get them from other, non-Google sources such as Slack and Messenger. It would be great to have a way to “send” appointments to Maps.

Finally,  what I love about Maps is that even though it’s extremely efficient in its most basic use case, finding place A and B and getting from one to the other, it never stops adding features that improve the product in different ways for, perhaps, different groups of users. Also, Maps is bucking the trend to spin out features from multi-functional apps into their own standalone app which makes for more efficient interaction but another split moment to decide what app to open. 

Maps is an app I use every day, sometimes several times a day. I rely on it to get places. Any other feature needs to enhance that basic case but also not get in its way. The new place markers do that well.

Super Bowl 2017 advertising: social media is on the sidelines

My favorite Super Bowl moment: the Schuyler sisters, Renee Elise Goldsberry, Phillipa Soo, and Jasmine Cephas Jones, sing America the Beautiful before the game.

My favorite Super Bowl moment: the Schuyler sisters, Renee Elise Goldsberry, Phillipa Soo, and Jasmine Cephas Jones, sing America the Beautiful before the game.

Even when the actual football is as exciting as it was Sunday, I still watch the Super Bowl for the ads. It never fails to bring out the most interesting, creative, and funny spots, where most brands just want to stand out. This year a 30-second ad spot cost a bit over $5 million and over 113 million Americans watched the game. Last year 112.6 million people watched, easily making it the most watched show in the US for 2016. As a comparison, in second place for 2016 was the post-game show, which drew around 70 million viewers. In third place was the exciting 7th World Series game, which drew a bit over 40 million people. There is nothing like the Super Bowl for getting a lot of Americans watching the TV at the same time.   

To figure out what the advertising world thinks of the web and the second screen viewers are holding in their hands as they watch the game, I noted social media, hashtags, and URLs for all 109 ads shown from right after the national anthem and right up to the beginning of overtime. I noticed a few interesting trends:

  • Social media handles are (almost) nowhere to be found, with only 3 brands including them in the final end-slate. T-Mobile was the only brand that mentioned Facebook, Instagram and Twitter, and they ran four different ads. Other than T-Mobile, Pistachios and Positive Coaching Alliance included Twitter handles. No one included Snapchat.
  • Hashtags are a bit more important but also down in popularity. I counted hashtags in 29 ads, about 27%. Danny Sullivan at MarketingLand says that according to his count, hashtag use is down from a high three years ago of 57% to 45% last year and 30% this year (a note below on why his count differs from mine). Some hashtags were specially created to match the Super Bowl ad, such Febreeze with #BathroomBreak, Airbnb’s #WeAccept, Tide’s #BradshawStain and all four of T-Mobile’s, but many were the “generic” brand hashtag, such as Skittle’s #TasteTheRainbow, Sprite’s #WannaSprite, and all those for movies and TV shows. Yet it is interesting to note that brands aren’t interested in driving the conversation to social media, especially Twitter, as much as they had in previous years.  
  • This isn’t to say that social media didn’t play a part. T-Mobile and Verizon started a weird Twitter war some time in the 4th quarter and there were “5.1 million tweets about #Gaga’s performance, and Lady Gaga was mentioned on Twitter 2.1 million times between 7:50 and 8:40 p.m. EST.” So don’t discount the second screen quite yet.
  • URLs were added to 33 ads, around 30%, slightly higher than hashtags.

I’m surprised that the number of hashtags was significantly down and the number of handles was almost nonexistent. Sure, social media has mostly been a tacked-on afterthought on television ads, but for the advertising world to say that it has lost importance over the past few years when both Twitter and Facebook played a huge part in the 2016 elections is astounding.

Also, a note about the count. Danny Sullivan, who has been tracking social media and web mentions for a few years now, counted only 66 ads. He says that “we tried to count only ads that were nationally shown, as best we could, viewing from Los Angeles. Promos for shows on FOX or from the NFL were not included, nor were very short 15-second ads for “The Walking Dead” and Fiji Water.” I omitted promos for my local Fox affiliate but there were some ads that Mr Sullivan didn’t count that I’m pretty sure were national, such as Fitbit, Tide, and Samsung Gear VR. He also counted ads from the kickoff and I counted them from the anthem. By Mr Sullivan’s count, 30% of the ads had hashtags and 41% had URLs. He saw 5 Twitter mentions and 4 each of Facebook and Instagram (he counted every one of T-Mobile’s 4 spots as different mentions). These are still very low numbers.

Finally, the most intriguing and unexpected ad of the game might be a portent of things to come. The ad, for 84 Lumber, ran right before halftime and chronicled the journey of a mother and daughter from Mexico to US. The ad ended before the pair reached the US and called for viewers to continue watching online. The reason for the suspense: “Fox rejected the ad for being too overtly political” as the pair’s journey ended at a big wall. Unfortunately, 84 Lumber, a building supply company, certainly a non-traditional Super Bowl advertiser, wasn’t prepared for viewers to do exactly what they wanted them to do: watch the full ad on their website.

The onslaught of viewers just goes to show that even though they’re watching the big game, viewers are more than willing to look at content on their phones. It will be interesting how ads will look in two years. Will brands push the conversation to social media or will they create more opportunities to continue stories online?