The new Google Maps traffic notifications: are beautifully implemented but practically useless

Some time last week Google asked me for permission to send me traffic notifications. Magnanimously, I agreed, and now, from what I can tell, Google sends me alerts about traffic where I am a few times in the morning and a few times in the afternoon. I get notifications when traffic is bad for the first time, when it continues to be bad, and when it starts becoming better. The area includes the expressways and highways around my current location.


From notification…

The feature is well designed: the notification tells me everything I need to know. This one, from yesterday afternoon, says that traffic is slower than usual with delays of up to 8 minutes. The notification takes me to a maps page where the currently delays are marked with heavy red highlighting, with the extra minutes added for emphasis.

...yes, it's bad.

…to traffic mess.

Yet, though in some ways I feel we need all the traffic help we can get here in the Bay Area during rush hour, these new notifications don’t seem helpful. Let’s start with this definition of a quality notification:

  • What is the purpose of the notification?  Its purpose seems clear: to let users know about traffic congestion that might impact them. It does that.
  • How is the notification going to be useful to users? Here’s where I’m not sure. A notification of congestion can only be useful if the user was planning to go somewhere and might delay because of unexpected congestion. But A, from what I’ve seen, the congestion behaves in a similar manner around every rush hour, and after a week of notifying me about traffic in the same spots every day, it is no longer useful. B, users check traffic anyway when heading out and Maps already considers traffic when suggesting routes. So why the notification? 
  • Will the notification be relevant to all users or just a certain user segment? The traffic notifications are highly personalized and focused on a time and location, yet aren’t tied to a scheduled meeting or a planned event so they’re irrelevant.
  • Do you really need notifications in your app at all? Does Maps need notifications at all is an excellent question. Do users even need to be reminded to check traffic before setting out to their destination or is that an ingrained habit already? Notifications should be (as they are) part of a calendar notification, telling users when they need to start driving based on their current location to reach their appointment on time, not just a report on local traffic conditions.
  • What action(s) is the user expected to take after receiving the notification? The flow works well from notification to viewing current local traffic problems. There’s also a button to quickly calculate a route, so if a user found the notification helpful, the next steps they can take are clear.   

Here’s the thing: notifications are supposed to provide value and increase users engagement with an app. I’m not sure that the traffic notifications does that. Even after only a week of them turned on, I find myself close to declaring them useless as there are simply too many of them. If I need to go somewhere, I check traffic and plan accordingly. If I don’t, why do I need to know what traffic is like?

There are two situations where I would find a notification like this useful: if there were a special and unique reason traffic was worse than usual. For example, a game at the 50,424 seat Stanford Football stadium creates unusual traffic in my area or traffic when I wasn’t expecting it, such as a Black Friday rush at the mall. Another beneficial use case is to find situations where users don’t usually plan a route because it’s a local drive, one that they make a few times a week, say to the gym (based in calendar entries.) Those are cases when knowing where traffic is congested en route is useful.

It’s not helpful to get an identical notification every workday at about the same time. Yes, traffic sucks at 5pm on a weekday on the 101. Tell me something I don’t know.

Music’s new frontier: the home assistant

On October 4th, Google reintroduced Google Home, its answer to Amazon’s Echo after announcing it at I/O in May. In it they listed the top four intended uses for Home: listening to music, getting answers from Google, managing tasks and controlling IoT devices. Putting music first is pretty significant, and I assume that Google believes that that will be one of Home’s main usages. This theory can also be supported by the emphasis they put on the speaker build during the presentation.

Introducing Google Home

Introducing Google Home

Google talked about more than just sound quality. They also touted how its search capabilities would be used to play a song without knowing the song’s title, and how users could link four non-Google music streaming services (Pandora, Spotify, TuneIn Radio and iHeart Radio) to their Home device (not sure how and if they would all work together.) Google, while it doesn’t quite have the product admiration Spotify has, has had a music streaming product since 2011 and also bought Songza, one of my favorite music streaming apps that was one of the first to tie music playlists to activities. Today Google Home has the ability to play “music from popular music services by artist, song, genre, album, playlist, mood or activity” and deems music as one of Home’s most important capabilities.

So it makes sense that Amazon, creator of the smart home assistant category, would start making music a priority on Echo. Before this week, Amazon offered Echo users all the four non-Google services that Home supports, along with Amazon Prime Music (on demand streaming)  and Amazon Music (purchased songs.)  Prime Music was only available to Prime members ($99 per year) and offered a limited catalog: 2 million songs, compared to the ~30 million that Apple, Spotify and Google offer. Unlimited offers the same 30 million song catalog and supposedly works well with Amazon’s virtual assistant, Alexa, and claims to be able to find songs, learn personal preferences, and create playlists.

It’s interesting to note that Amazon Unlimited costs $4 per month. Compared to all the other services that cost $10 per month, it seems like a bargain, yet it is limited to one Echo and only that Echo, no other device. This means you cannot listen to Unlimited on your phone, laptop or in your car, which the more expensive services offer.

Predictably, the big three, Apple, Google and Amazon, have all chosen to build services that are available exclusively on their devices. To listen to your Apple Music subscription on your Echo means connecting via Bluetooth and using the Echo as a speaker. I’m not sure, since I don’t have an Echo, how smoothly this works. It certainly isn’t as smooth as just asking Alexa to play a song from Unlimited.

What’s disappointing with the recent flurry of releases is the direction the big three are taking to lock users into their hardware. These wonderful, intelligent music products can automatically create playlists based on current demands, historical preferences, and what similar users have listened to. With a large library of songs at their disposal, ever-improving machine learning capabilities and speech comprehension, listening to music can be a delight. Additionally, these services improve over time based on user preferences with personalization of playlists and prefered artists. Yet by not supporting the other services on their hardware, Amazon, Google, and Apple are limiting users to buying only Echo, Home, and whatever-Siri-will-power respectively.

What this means for consumers that, for example, they may need to compromise on sound quality (“the Echo’s sound quality is uneven at times, with weak bass at high volumes”) to continue listening to their playlists and accumulated preferences. It also means that Spotify might start becoming the service of choice for more listeners as it is currently supported by both Echo and Home, as well as across all phones and devices. It could also mean that users start subscribing to multiple services, whether that’s e result of wanting to have the best performance on their chosen hardware or just for accessing exclusive music releases.

Finally, it also serves to underscore just how much listening to music has changed in the last decade: from owning the media and having the ability to play it on any hardware that could support the format, be it CD, vinyl or tape, or MP3, to on-demand music, vaguely defined by listeners current moods and activities but influenced by their previous choices and accessed by voice, but only on selected devices.


Moovit vs Citymapper: making public transportation easy

Earlier this week, Techcrunch reported that Moovit, a public transportation service, has revamped and relaunched its app. As a big fan of Citymapper, one of its competitors, I wanted to see how an already really good product could be improved.

Right off the bat, Moovit is beautifully designed and doesn’t introduce new forms of UI, which I appreciate. Sometimes it’s nice to open a new app and have the flow be instantly familiar. There’s also some “general” information about local public transportation, which doesn’t have to do with planning a route. Yet, it’s the route planning that I use the most, and comparing that part of Moovit to Citymapper was my first task.


Moovit displays only four route options.

Citymapper displays 6 route options and 3 alternatives, though travel time would be easier to read in hours:minutes format.

Citymapper displays 6 route options and 3 alternatives on the first screen.

Both apps do a good job at finding routes at figuring out and presenting route options. I asked to go from my location to AT&T Park in San Francisco and to be there at 6pm, when the Giants game starts. Citymapper wins this round.

First, it presents more options, including walking, biking, and getting a cab/Uber. This may be less useful to get from Palo Alto to San Francisco, but necessary in a city where walking can be a good alternative to public transportation. Citymapper showed six suggested routes, two rail only and two rail/Uber combos, while Moovit showed three suggested route and one bus only. Again, probably more useful to have more suggestions in a city with multiple public transportation options, such as New York, than the Bay Area, where we have one rail line.

Second, Citymapper has laid out the route data better, including more information in less horizontal space, which allows for more options to be shown on the screen at once. It’s an interesting lesson in UI to note what information each app has decided to present and how to present that information visually. Moovit emphasizes travel time, and makes it easy to see which option requires more walking, but makes it hard to figure out when to leave. Citymapper has travel time, time to leave, and the cost of each option. I choose routes based on travel times, and I appreciate seeing more options on the same screen, without scrolling, before making my selection. It would help if both stated the travel time in a hours:minute format as opposed to just minutes, but that’s just my preference.

The one feature I liked on Moovit that Citymapper doesn’t offer is to plan a route for the “last lines for today” which is useful when public transport ends early (according to Moovit, to come back from AT&T Park today, I’d have to leave at 8:07pm.) Moovit also offers route options such as “least walking” and “least transfers” which are great, but are on another screen. I’ve used Citymapper when it was raining when it automatically added “rainsafe” routes in the main route selection screen. Again, faster route choice is always appreciated.

So, in cities where both operate, I’d choose Citymapper, but on my last trip, to Berlin and Tel Aviv, Citymapper only had maps available for the former. Citymapper covers 37 cities, while Moovit claims to have 1,200! If it has in any way made it easier for local public transportation authorities to interface with it, that is a huge advantage.

Flooding at 14 Street station.

Flooding at 14 Street station.

What I am really intrigued by is Moovit’s attempt to use crowdsourcing in addition to official sources to provide routes and other information to its users. Intrigued because despite my love for Citymapper, it has failed me two times, both in New York. The first was having me walk 10 minutes to a Subway station (and to climb three flights of stairs) only to find that it was closed for renovations. The second was directing me to a station that was flooded by a burst pipe, with trains shunted to alternate tracks and not stopping. In both incidents Citymapper had no idea there was a problem and I ended up wasting a lot of time finding alternate routes.

Reporting conditions for a line on Moovit.

Reporting conditions for a line on Moovit.

By allowing users to report problems, users can avoid certain stations, and lower everyone’s stress levels. I didn’t want to actually report a problem but it looks like Moovit has made it easy by choosing a line or station on my chosen route and then letting me choose what type of problem to report like crowdedness, an incident, cleanliness, and so on.

Techcrunch ended its post by questioning Moovit’s necessity as a standalone app quoting its VP of Product: “Meydad’s hope is that by distilling the service into a few key experiences and making them easily accessible, Moovit will provide some kind of differentiated experience from Google Maps that merits being a completely separate app.” I don’t doubt for a second that a standalone app is necessary. Anyone who has traveled to an unfamiliar city and just wants to get from one point to another quickly and efficiently needs a public transport app. Using Citymapper on my last two trips completely changed my experience for the better: instead of taking cabs because I didn’t want to open a map, I could take a quick look at an app that instantly gave me the best route. Google Maps takes more time just to come up with less options and less route detail. Having a dedicated app is definitely the way to go.

Can LinkedIn do more to help women returning to the workforce?

This morning, at a program for women returning to work after a long break, a representative from LinkedIn talked about the importance of having a LinkedIn profile and instructed attendees on how to create and optimize one.

The audience was extremely receptive but as the speaker was going through the steps to build a profile, it was clear that LinkedIn’s current structure created a conflict for some of these women. The challenge, which was reflected in the numerous questions about the different parts of the profile, focused on how to optimize the profile to appeal to recruiters and overcome their biases for age and gender without omitting too much of what returners are passionate about and without lying. The questions focused on everything from how a profile photo dates candidates to whether to omit graduation dates and even positions held to appear younger.

The tone in the room was positive, with a reassurance that women returning to the workforce do bring necessary skills to the table. There are companies that appreciate those so-called “soft skills” that appreciate the maturity and experience returners can bring to a role. The rich variety of strengths and wisdom that they have gained outside the workforce, in volunteer positions, parenthood, and caring for sick family members, are an advantage. Research has shown that companies with more diverse workforces perform better financially.

LinkedIn's profile advice: first step is a photo. Source: LinkedIn

LinkedIn’s profile advice: start with a photo.
Source: LinkedIn

Yet even though companies, and by extension recruiters, may be interested in skills that returners have, LinkedIn is really focused on the more traditional, climb-up-the-ladder career. LinkedIn isn’t about soft skills, or on how caregivers juggled multiple responsibilities that weren’t professional. LinkedIn is about a single career path, and not a windy road. This difference between what the women in the room needed and what LinkedIn, as a product, offered was obvious throughout the talk.

The two biggest product differences between what the LinkedIn profile offered and what the women in the room need were:

  • Functional vs chronological format. This was a big obstacle to many women. Consider the advice LinkedIn gives users to create a complete profile: “An up-to-date current position (with a description) and two past positions.” This is easy for someone currently working and that has been working for the past few years. This is challenging for returners because most often there is a gap that they have to fill with some creative twists such as self-owned consulting companies and exaggerating minor positions.Yet, a solution exists: the functional resume, that allows job-seekers to present their experience in a better light. LinkedIn could offer seekers different profile formats and allow them to choose which they wanted to use.
  • Multiple objectives. Several women stated that they would like to target multiple fields and positions yet LinkedIn allows users to present one face to the world. The analogy is a detailed cover letter that job seekers would tailor to specific positions, that highlights relevant skills. This request came up mostly when discussing the headline and summary. This could be solved if LinkedIn could allow users to pick what summary/headline to show for specific queries and job titles by recruiters.

Two other minor issues were brought up:

  • LinkedIn really spotlights companies worked at, and recruiters can even search for experience at larger, well-known corporations. Many women who have taken a career break do not necessarily have that on their resume, sometimes because they have been out of the workforce longer than these companies have been industry giants.
  • Volunteer experiences and causes are not part of the timeline and are tacked on at the end of the profile and seems like an afterthought.

Perhaps some effort can be done on the side of recruiter search tools. As part of the talk, we got to see what they can filter for, mostly based on experience. Yet, if there is an industry demand for certain soft skills, perhaps it is time to allow filtering for them? One thing I did like is that for experience, the current threshold is “over 10 years.” Many returners have way more than that in terms of experience but this option will bring up their profile even when the recruiter is looking for “only” around 12 years. Yet there is also so much of an emphasis on the current position and company that is sure to lock many returners out.

Finally, there was talk of both gender bias and age discrimination, specifically in the tech industry. A woman asked if it is even helpful to add photo that shows recruiters just how old she is. The response was that since profiles with a profile photo are 14 times more likely to be viewed that profiles without one, that outweighs the hypothetical disadvantage of showing an older face. Just today the Wall Street Journal posted an op-ed piece that women in tech should definitely not use their photo online not only on LinkedIn but on all social media platforms. In fact, women shouldn’t even use their own name!! The advice given: “in your LinkedIn profile, Twitter account, email address and online correspondence use your initials (or a unisex name) and eliminate photos.” Photos are used to determine not just age , but also gender and race, and give recruiters a very easy way to eliminate candidates. 

Could LinkedIn fight that?

A few weeks ago Airbnb announced it would “experiment with reducing the prominence of guest photos during the booking process” to prevent discrimination. Experts weighed in and said that more should be done to anonymize the process to combat discrimination: “profile pictures should either be eliminated or only shared after the booking is confirmed [and] that names may need to be treated the same way.” In all honestly, on LinkedIn, are names and photos even necessary? After all, skills and experience are what recruiters are really looking for, not looks. If no one had profile photos, would that level the playing field? If LinkedIn is uneasy at removing both of those, how about just for the search results? Let users optimize headlines, leaving photos and names for the full profile.

Update: one day later and a few responses have been written condemning the op-ed.”The point is, Greathouse advises women to appear more like the tech entrepreneurs we know now. But that makes absurd demands of women. And it ignores a fairer solution: asking funders to change what they think an entrepreneur looks like.” 

Google Allo: yet another messaging app

Google announced Allo, its new messaging app, at I/O back in May, and launched it last week. I didn’t rush to download it because, frankly, I don’t need another messaging app. Right now on my phone I have Hangouts, which works well with my Gmail contacts, which are the majority of the people I want to communicate with, and WhatsApp, which I was basically bullied into using. I’ll get back to that in a moment. I also have Messenger for text and Google Voice for even more texts. Bottom line: I can already message anyone I want with the apps currently installed on my phone. 

But in the spirit of messaging adventure, I downloaded Allo. First it asked for permission to access Contacts, SMS, Storage, Camera, and Media, and wouldn’t continue until I granted the first three. By ‘wouldn’t continue’ I mean throwing me out to the app setting screen without a clear indication of what to do (i.e. go to permissions and allow everything.) I gather that research has indicated that users just ok every permission asked for but for those who don’t, this isn’t the best experience. Anyway, continuing through an easy phone-number verification, Allo asked me to either message a friend, start a group, or chat with the assistant.

Chatting opens up a list of my Google Contacts with everyone I ever entered, from my local Costco to my dentist, to my friends and husband. Surely, Google, you know better than to present these contacts alphabetically? Surely you know who I chat with most often on Hangouts and email most often? Allo adds to the scrolling misery by listing every number for every contact separately, so that it takes even longer to find the right contact. What I like about this interaction, though, is that Allo doesn’t care if my chosen recipient is on Allo or not, it just sends a message, routing it by SMS if necessary. The Verge says that it’s supposed to do something way more interesting and invoke Google’s new “app preview notification,” which should “shoot a notification directly to your Android device instead of going through SMS. Your friend will get a notification that looks and acts almost as if they had the app installed in the first place, message content and all. It means they won’t incur any SMS fees, either. Your recipient can reply within the notification, or tap on it to install the app.” This didn’t work to the friends I sent an Allo message to, but it sounds promising and a great way to help build an Allo network.

Discussing the weather with Google Assistant. I love the follow up bubbles.

Discussing the weather with Google Assistant. I love the follow up bubbles.

What I do love on Allo is Google Assistant, and I have a feeling we are going to be good friends. It’s easy to ask questions, replies have been spot on, and the assistant (shouldn’t have a catchier, shorter name?) suggests good follow up questions for some queries. For example, I asked about the weather, the assistant said to wear sunscreen as it is 84F and I was given the option to ask a few follow up questions, such as “what about this weekend”  It will need to get smarter, through, because as I ask “what time is the debate” I get the top few news links referring to the debate instead of the answer I needed. But this kind of machine learning and searching for information is what Google is great at and I’m sure it will improve greatly over the next few months. I also like that it can be invoked in group chats, especially useful when talking about where and when to meet and future, mutual events.

When is the debate?

When is the debate? I don’t need news, I need a time.

Finally, getting back to why I installed WhatsApp a few months ago. While I felt that I could message anyone I needed, it turns out that a majority of conversations between friends and family were happening on WhatsApp, which I had not yet installed. In Israel, where many of them live, literally everyone uses WhatsApp, to the point where schools, PTAs, public events, and government- and municipal- sponsored activities are all coordinated via WhatsApp. While I could still message my friends on the other platforms, WhatsApp groups is where everything is happening, which is why I finally started using it. Just like Allo, WhatsApp identities are based on a mobile phone number and while WhatsApp doesn’t come with Allo’s friendly invites, WhatsApp already has all my (international) friends and family. There is no argument I can make to get them to switch to Allo, where the only benefit is having an assistant on tap. The inertia (and network) is strong with WhatsApp. I also don’t see my Hangouts or Facebook Messenger pals switching to Allo. Whether it can gain enough of a user base to become a messaging app of choice for any community or group remains to be seen.   

A few things need to change to put smart devices in more homes

A few weeks ago Gartner released its updated Emerging Technology Hype Cycle, where it places various technology along the early adoption curve and estimates how long technologies have before they reach mass adoption. Internet of Things devices have been on the Cycle for a few years. Last year they were at the Peak of Inflated Expectations, meaning a lot of fluff  and talk (aka hype) but not a lot of great products. This year Gartner moved them off the chart, perhaps signifying that IoT has indeed reached mass adoption. With hardware, appliance, and tech stores full of connected, “smart” devices, maybe IoT really is mainstream.

But is it?

Looking at the abundance of devices we could be tempted to say that yes, there’s a “smart” everything now (even a t a connected candle!) but ask a non-techie friend what smart device they have and I bet they’ll only show you a fitness tracker. I think this boils down to a few issues:

  1. Apple's HomeKit looks like a promising IoT hub with simple addition of devices and the ability to define rooms and scenes for easier and more automated control. Source: Apple

    Apple’s HomeKit looks promising, with simple addition of devices and the ability to define rooms and scenes for easier and more automated control.
    Source: Apple

    Central control is not there yet and every smart device comes with its own app. It’s fun for the first two, three devices but then it becomes tedious to have a different app for your thermostat, blinds, lights, and coffee machine. The act of getting up in the morning now requires opening four different apps or automating everything independently, making it a hassle every time you want to wake up one hour later. The big hope for this space, Apple, just released a version of Home Kit with centralized device control and Google has the “works with nest” platform and is also planning to launch Google Home soon, with the intent of controlling all home IoT devices. Amazon’s Alexa also works with some devices. The question, given that these solutions are relatively new, is whether they are ready for wider adoption and how easy it will be for other devices to work with them. Also, how difficult will it be for device manufacturers to be compatible with all systems?

  2. Setup of each device is unique and it’s not always easy Many manufacturers realize that requiring electricity for their devices is too high a barrier for adoption and have opted for battery powered devices with wireless connectivity (of course) but every device requires its own software setup and each is unique. Perhaps once the IoT platforms are more robust a device will only have to be turned on to be added to the system.
  3. Prices are still way too high. A Nest thermostat costs $249, programmable ones under $50. Light bulbs cost between $20 and $60 each, and some require an additional hub, whereas non-connected bulbs are a fraction of that. The Nest fire alarm costs $99 while a three-pack of regular alarms are less than $50. All six smart locks in this review cost over $150 each where a simple lock costs under $15. A smart doorbell? $200. Granted, all the smart devices offer much more functionality than their unconnected equivalents but those are premium prices, intended for customers who are willing to pay extra, aka early adopters.
  4. Bugs are treated nonchalantly. IoT malfunctions shouldn’t be an acceptable consequence of a software upgrade. It’s not as if a game crashes, it’s unlocked doors or doors that don’t open, burst pipes and cold babies. These stories gather much attention and serve as a deterrent for new adopters.
  5. Privacy of collected data and the ownership of that data. Every smart device stores data “in the cloud” whether it’s steps taken or the content of the fridge or when people enter and leave your house. There are three issues here customers need to be reassured of: whether that data is safe, who has access to it, and whether you need to pay to get that data.
  6. Less ownership restrictions. There are many strange uses of DRM by companies to protect their products such as Keurig with its “use only ours” pod and, more recently, HP blocking third party ink cartridges with a software upgrade. Even IoT devices aren’t immune: Phillips, for a time, blocked third-party lights from its Hue hub. Connected devices are continuously open to such upgrades. Manufacturers can also change the terms of use when they please. Tile just sent me an email notifying me of an update this week of an update to their Privacy Policy without giving me an option to opt out or clearly mentioning what had changed. Manufacturers can also choose not to continue to support older products, leaving important devices, such as a front door lock, inoperable.

Early adopters, those willing to pay a premium, risk security, and spend lots of time setting up devices and getting them to work with one, central hub have already bought the first and second wave of IoT products. Gartner estimates that the number of installed consumer devices will more than triple by 2020. For that to happen, products need to be simpler, safer and cheaper.

Why we’re still struggling to buy clothes online… at least women are

Last week, Tim Gunn (you know him from Project Runway) wrote an interesting op-ed about how fashion designers were ignoring a large section of American women. “There are 100 million plus-size women in America, and, for the past three years, they have increased their spending on clothes faster than their straight-size counterparts. There is money to be made here ($20.4 billion, up 17 percent from 2013). But many designers — dripping with disdain, lacking imagination or simply too cowardly to take a risk — still refuse to make clothes for them.” Mr Gunn also added that “most designers max out at size 12” yet “the average American woman now wears between a size 16 and a size 18.”

So why is this in a tech blog? About a year ago I wrote about challenges in online clothes buying unique to women and how little fashion retailers and brands were doing to personalize the experience. One year later, beyond the personalized shoppers, most sites still go for the standard display: a grid of clothes, mostly modeled on a tiny (Size 4? Size 2?) model or laid out on the floor. The better sites offer a wide range of filtering options but only a few offer any kind of ability to see different sizes modeled.

Elizabeth Suzann's wonderful size lineup.

Elizabeth Suzann’s wonderful size lineup.

Yesterday, a Racked post talked about a small brand in Nashville, Tennessee, making the effort to show their clothes on different body types: “Elizabeth Suzann will now show every item of clothing in their Signature Collection on models who represent every size in which the item is available.” See, for example, this beautiful LBD shown first on the customary waifish model and then on six different sized models, along with their height and weight. Having this information makes the buying decision much easier, reduces post-delivery dissatisfaction, and might help prevent a return.

Detailed review on Rent the Runway.  Includes size worn, overall fit, occasion rented for, size usually worn, height, age, bust, body type, and weight.

Detailed review on Rent the Runway.
Includes size worn, overall fit, occasion rented for, size usually worn, height, age, bust, body type, and weight.

Before Elizabeth Suzann I had only seen Rent the Runway’s solution: for each outfit, different women shared of how it looked them, how it fit, and some personal stats. Most share height, usual size, chosen size, and fit; some shared age, weight, bust size, body type, and occasion. Most importantly, all share a photo of them wearing the dress. For someone who needs to choose a dress to wear to a special occasion, such information is invaluable. A random dress I picked out in the “popular” section had 167 reviews with 126 (!) photos. That is an incredible number of reviews and offers users a chance to see what a dress looks like on different body types. Rent the Runway also offers a way to filter reviews and photos to show only “Women Like Me” (by size, height, bust size and age.) Viewing this information can make the rental decision easier and more satisfied customers.

So what will it take to see a shift, if not in the range of sizes offered women, at least in the way clothing is represented online? If Elizabeth Suzann manages to significantly increase revenue, it might prompt other brands to show different sized models. If not, as Elizabeth Suzann is a relatively small brand, (making this effort exceptionally impressive) it might take a larger brand to influence others. The other alternative for online retailers is what Rent the Runway does with its detailed reviews, which, for huge collections, might be a bit more feasible. Either way, it’s high time for a change.