Kwoller, A Tinder For Fashion Price-Tracking, Launches At Disrupt NY

Kwoller, a new mobile commerce startup launching its product today at TechCrunch Disrupt NY, wants to speed up your next impulse buy. Got an idle two minutes in a queue for latte? Then whip out your phone and prepare to swipe.

Kwoller is taking the swipe-to-judge gesture that’s been popularised by dating app du jour Tinder and applying it (initially) to women’s fashion. Swipe left if you hate the shoes, right if you love them enough to buy them. Rinse, repeat.

The main competitors to Kwoller are not just other shopping apps, say co-founder Tim Bernal — rather it’s competing for attention with the likes of Instagram and Twitter and Facebook, hence its focus on speed and sticky simplicity.

If you swipe right an item is added to your ‘love list’ and the Kwoller koala mascot (selected because koalas are famously choosy about what they consume) will also keep its beady eye on the item’s price-tag — sending a push notification of any price drops so you can swoop in and snag that bargain.

So, again like Tinder with its game of matching, Kwoller has something more sticky than just the prospect of rating stuff to keep you coming back to its app.

Simple is better on the mobile phone.

The koala is the cute face of Kwoller’s price-tracking service. “This koala is determined, his bow tie means he’s there to serve you,” says Bernal, getting into his stride on the topic of the app’s mascot.

That price-tracking feature is the app’s “value add” — the thing the co-founders reckon can make Kwoller go the distance vs other mobile commerce startups trying to make swipeable or shuffleable shopping app interfaces stick. (Mobile shopping app rivals trying the same swipe-to-like gesture interface include Mallzee and Stylect, to name just two.)

“The koala is the guy who stays up late at night tracking your items tirelessly, and gets no better benefit in the world from sending you a notification that the item you love is now on sale,” adds Bernal.

Kwoller is also determined to get real good, real quick at knowing what you like — thanks to all that swiping — so will be aiming to have its algorithms increasingly push the sorts of fashion items you actually want to buy under your thumb.

Swipe left and an item not only disappears forever but Kwoller’s algorithm takes note of your dislike. The more you tell it, the better it gets is the claim.

How can you build an algorithm that predicts all the fleeting fancy of fashion? After all, just because someone really loves one lurid green sweater today does not mean they will always love every lurid green sweater…

“Without tipping the cards too much on how our algorithm works, we’ve taken a long deep look at why fashion might be different than other recommendation engines,” says Bernal. “A lot of people will spend time codifying features or attributes  and then building a model to predict based on that — so if you think about the music genome project by Pandora, that’s one approach.

“There’s components of that approach that are within our algorithm, but it’s not solely there. And from our tests so far we’re very pleased. And… as we get more data, these algorithms will get more robust.”

“We have an amazing data scientist on board right now. So I think this thing will definitely delight the users,” he adds.

[gallery ids="996909,996910,996906,996905,996907,997968,997971,997967,997969,997970,995995"]

The app is launching with a handful of U.S. fashion retailers signed up, and lets users select from a particular category of clothes or accessories to browse and then start swiping through a pipeline of potential purchases.

Kwoller keeps the shopping experience clutter-free by displaying only a single product at a time, asking the user to make an isolated value judgement on that one thing, and thus stripping away the complexity that Bernel and his co-founder, Brian Louko, argue continues to dog other mobile shopping apps.

“It’s that kind of super fast shopping that I think lends itself very well to mobile,” says Bernal.

“We think mobile shopping now just sucks,” adds Louko. “Every company tries to make it more and more complex, with a search functionality, filters, QR codes, this idea of covering the ‘story’ around the shopping experience. But we go on the mindset of atomisation or ‘sipification’, which you’ve kind of seen with how YouTube has transferred to Vine, blogging and Twitter, and then most recently online dating into Tinder. So simple is better on the mobile phone.”

That ‘keep it simple, stupid’ mobile commerce philosophy is perfectly summed up by the minimal effort required to make a thumb swipe. Small effort, big potential reward when a notification about a price reduction on that coveted pair of pants gets automatically pushed your way.

Women’s fashion is the first focus for Kwoller because, say the co-founders, the “aesthetic snap judgement” a buyer makes on a particular dress or handbag is akin to the snap judgement that powers Tinder’s never-ending quest to discern who is hot and who is not.

Bernal also has a background in fashion retail which he drew on when he and Louko were kicking ideas around at Columbia Business school — where they met in 2012 — for a startup business. Kwoller was the result of their whiteboarding.

[gallery ids="998885,998897,998890,998889,998888,998887,998886"]

Beyond women’s fashion, the pair say they see definite potential to expand Kwoller into men’s fashion, and even other retail categories — such as home decor. But they’re starting off with female fashion to keep the business focused.

At launch, payment is not yet integrated into Kwoller — if you click to buy an item, the app will open up the mobile page of the corresponding retailer’s website. Which is obviously sub-optimal for the friction-free demographic Kwoller is targeting. But the pair say integrating a native purchase option within Kwoller is high on their priority list.

The business model for Kwoller is not just the obvious affiliates and commissions on swipes that lead to purchases. The co-founders are aiming to develop what’s an MVP at launch today with relatively few products being pushed into its pipeline, into a fully fledged platform for native advertising down the line — once they have enough app users on board and enough swipe data racked up to know what those users really like.

The platform they envisage will have sponsored posts incorporated into its mix — tailored and targeted with all the preference data it will be speedily amassing via its users thumbs.

Another idea they are cooking up is to sell sponsored swipes — so retailers could pay to have a series of items strung together, for instance, so the user sees them in a particular sequence/progression.

They’re also planning to create ephemeral discounts/exploding offers which would give the user an incentive to swipe right on a retailer’s item to get a discount if they purchase it in the next 10 minutes — bringing a whole new meaning to the phrase ‘fast fashion’.

In addition, they reckon there is future potential to integrate a Kwoller platform of user preference data with Apple’s iBeacon or other physical retail store technology, linking a bricks and mortar store with the stuff an individual Kwoller user loves the moment they walk through the door.

“Kwoller’s real true potential comes when looking outside the four boundaries of your smartphone screen, now if you’re able to walk into one of our retail partners’ stores — there’s iBeacons, there’s Bluetooth Low Energy — and as soon as you walk in the retailers know which products you’re interested in, and then they can start providing a better customer experience there,” says Bernal. “And you can get those deals on the products that you’ve selected that we know you like.”

“Kwoller’s going to be an amazing mobile commerce app but I think it’s also going to be amazing to enable some omni-channel experiences that are not out there right now,” adds Louko.

To further facilitate this platform play they are also leaving room for an SDK to allow developers to build ‘applets’ that make use of the down swipe gesture within the app — giving third parties the opportunity to expand the Kwoller shopping experience in multiple ways. By, for instance, building something which shows the users which bricks and mortars stores in their immediate location have the item in stock. Or using the down swipe to locate items that are similar to the one shown but which cost less. And so on.

The up swipe gesture is currently being reserved for social sharing — letting users swipe up to share an item to Facebook, for instance. That’s an area the co-founders say they intend to monitor closely. If the social sharing element proves popular they intend to use it to try to build greater engagement by expanding to offer more social features — such as the ability to ask a friend for a recommendation directly within the app.

While Kwoller’s initial focus at launch today is the U.S., it is aiming to expand internationally down the line, with Bernal and Louko saying they see an especially interesting opportunity for their low friction mobile commerce in developing markets where smartphones may not be competing with a large installed base of PCs at all.

The future of fast fashion they envisage is going to be even faster if it leapfrogs the hoary old PC all together, and inserts itself straight under its target shoppers’ thumbs.

Q&A

Q: Are you guys planning on owning the cart or just do affiliates/commissions?
A: Today we are affiliates but we know that the best experience for customers will be a unified cart

Q: I like the simplicity of what you’re trying to accomplish. But how are you coming up with what you present to the users? What prioritisation are you doing?
A: We’re using a host of collaborative filtering techniques.  Whenever you create these systems there’s a few problems you can solve for. So when it first comes online you don’t know anything about your users and your items are brand new. So what we’re using is called item to item collaborative filtering.

We basically cluster items together, and at first users interact with those items and you start learning a little bit about them, and then as uses use the platform a little bit more we can start seeing correlations and relationships between users to users, and again we can start using that to pinpoint fashion tastes.

It’s very difficult to track every single aesthetic feature within our product – because a big button or a small button could be the difference between making someone like something or not, so we’re using a host of methods between item to item and user to user to help solve that problem.

Q: From a product perspective, can you talk us a little bit through the onboarding because I think that’s going to the hardest thing you’re going to solve… And the second thing… I think there’s a super critical piece missing here – and that’s community. So what are your thoughts on that because a that’s really a piece of discovery that’s immense.

A: The first question… Right now you sign up with Facebook  so we can already get some baseline of information about you. But the great thing about how we’ve set up our algorithm is based on what we know we give you our best shot, and as we learn more we’re able to kind of change gears in which algorithm we’re using.

On the second point – the social aspect, yes we do have that so much in mind… When we were talking to customers early on just validating the idea they said oh my god, this would mean I could see my friends’ love-list and then when it’s her birthday I could find the perfect thing to purchase for her? So we’re totally on board with that. And I think it could also be a very powerful platform for brands to tell their story to customers as well. Right now there’s some – either one to one, or two way communications within the different social media apps. And I think our platform can start solving some of the problems that they currently have.

Q:  Congrats with your pitch. Many people come on stage and they’ve raised a lot more than $10k so you guys did a lot on a shoestring so that’s impressive… Your presentation of tinder for shopping is very intuitive for the younger generation so I think that’s a good thing to go after. Can you talk a little bit about… customer acquisition. You’ve launched a great product, it’s obviously a noisy space – I don’t know if you have a few examples of how you hope to get some customers early on?

A: We have some amazing ideas in our marketing tank… A little peek for that is this summer we’re going to release offline browsing mode and then in New York people hopefully will call it subway mode.

We’re using a lot of digital media and social media strategies. And one of the members of our team that is what he does for his life. So we’re super confident about that.

The captive audience for something like this is on campuses so we want to have some campus ambassador programs and then we’re going to try and work with fashion and lifestyle publications to get some great stories about us.

Q: Can you take a minute to just give us some background on the team and then also I’m just curious about the name.

A: We’re basically all from Columbia business school. We met our engineers from Columbia as well. I [Tim Bernal] was doing retail before school – consulting for retailers.

Prior to Columbia [Brian] was at a YC backed startup.

We have a PhD from Columbia as well, focused on machine learning. And we have a guy that works with us who’s a creative director at an ad agency right now.

And how did we come up with Kwoller? The beautiful thing about koalas is that they are super cute, A, and B they are super super picky so there’s something like 800 species of eucalyptus but they only eat like a handful. So what Kwoller’s going to do, is there’s a billion different clothes out there but the Kwoller koala will help you find those perfect matches just like they find the perfect eucalyptus.

Q: Is there any ability for people to filter and tell you things they specifically don’t want to see – in terms of a specific category? …Obviously the breath of categories that you could be showing a person is enormous.

A: Retailers only right now, but we’ll be adding categories and search functions very soon.