Project Overview
Snap Fashion is the world's first cross-platform visual search engine for fashion. That means that you can see something you like, take a photo, and we'll find you similar items within seconds, searching hundreds of thousands of items for you. If you love something we direct you straight to a retailer’s website to buy it, or you can add it to your wishlist to save it for later, and we’ll even email you if it drops in price.
Organisation
Team
The team consisted of 2 in-house backend algorithm developers, who invented the visual search algorithms. We then worked with in-house designers and web developers, alongside an app development agency to create Snap Fashion and Snap ColourPop
Project Brief
Fashion is a really personal thing, and it's sometimes really hard to describe what it is that draws you towards your favourite item. You end up trying to search for it using words (receiving everything from the bizarre to the out-of-stock) or having to manually filter through thousands of items to find what you're looking for. Searching using pictures seemed like the obvious thing to do, and I'm proud to say that we were the first people to do it on the web and with a mobile app for fashion.
There are three ways to use Snap Fashion - we have two mobile apps and 1 website.
Our newest app, Snap ColourPop is really simple to use. You take a photo of a colour you like and it will find you everything in that shade from our 150 retailers. I used it the other day to find shoes to match my outfit. I also find it really fun to take photos of colours that I love, from colours trending on the catwalk to really vibrant flashes of colour in street art.
Our second app, Snap Fashion, goes one step beyond ColourPop because it lets you search by cut too. Take a photo of a specific item that you love and we'll find something similar based on the cut, colour and pattern. I use it to get inspiration from celebrity style and new catwalk trends.
Project Need
Normal search engines are obviously amazing for certain things, but fashion isn't one of them. It's really hard to articulate what you're looking for, and you end up constructing strange descriptions like "mini skirt dropped hem waist purple less than 50 pounds". You'll get a mix of results depending on how images have been tagged, rather than what they look like. If it returns results that you like then they'll be invariably out of stock, featured on a fashion blog from a couple of years ago, or not shippable to the UK. Everything in Snap's database is available in the UK and we do daily stock checks to make sure that we're all up to date.
And when it comes to shopping when you’re out and about, thanks to our phones everyone is now carrying around a pretty good camera in their pocket. There are some things in life that are so much easier to search for using a photo rather than tapping away at tiny keyboards and running into autocorrect woes, so I think it's only a matter of time that it becomes the normal thing to do.
User Experience
When designing Snap Fashion, the most important thing to us would be the ease of use. As the world's first visual search app for fashion we were conscious that we were trying to drive a new user behaviour, and as such we had to be careful not to make the process too complicated. We literally spent months running user groups and honing designs, before setting on Snap as it is today.
The process of Snapping is really simple. Tap the camera icon on the app, then select your clothing type. This produces an outline overlaid on top of the camera, to make it really clear that you should be photographing clothing, and drives users to take photos of the clothing head on.
Once the photo has been taken, we take them straight to results. One tap lets them see the product in more detail, the next tap takes them straight to the retailer to purchase it. We also implemented a custom ribbon navigation at the top of the app to let users quickly flick between their different types of results. This feature has proved popular, and is the user experience element that remains constant across our Android and iOS apps, along with Snap Fashion and Snap ColourPop.
Project Marketing
Snap Fashion's success to date has all been down to word of mouth. We have not invested in product advertising, but have driven over 280,000 downloads and web visits since launch in September 2012.
The most significant traffic driver for Snap Fashion has been PR. We have been featured in the likes of Cosmo ("It's so good you're probably downloading it as we speak"), The Times (Top 15 Fashion Apps in the World, 2013), The Guardian ("A glimpse of online shopping's future"), Grazia (Top Fashion App, 2013), Vogue ("The first of its kind"), Stylist ("An end to sales hell") and Style ("Snap Fashion will change the way we shop") to name but a few.
The reason that Snap seems to resonate so well with our users is that we're solving a true problem in a way that's easy to understand and satisfying to use. Online shopping has become crowded and frustrating - and we're cutting through that noise with a single photo.
Project Privacy
We have taken our users' privacy incredibly seriously when developing Snap Fashion.
The temptation when developing this kind of project is to use data from our users to drive content on our platform. However, we know that style is a personal thing, so never share any of our user generated content or search starting points.
We let our users pick what they want to share. All of our Wishlists are private and password protected (with our databases all hashed of course!) Users have the option to create a public Giftlist if they want to drop some serious hints around Christmas time, but our default position is keep everything private, and let people choose what they want to share.
Fashion
This category relates to fashion themed applications.
More Details