Lookastic

Your personal automated stylist

get it
There are no images or videos added to the gallery.
Add to gallery

Reviews

Discussion

You need to become a Contributor to join the discussion - Find out how.
5649
Sean Percival@percival
Style is emotional, topical and unique to the wearer. So I'm not sure how big data is applied here. Help me understand what's going on under the hood to drive good results.
20634
Alexander ZaytsevMaker@alexandrzaytsev · Co-founder & CEO, Lookastic
@Percival you are right about style. Our tagline has "stylist" in it because it's easier for people to understand what we are about. What we actually do is help people dress better. We believe that dressing well is basically knowing what goes with what (items, color, patterns, materials). Right now, when someone decides to start dressing well, they have to invest a lot of time into reading magazines/blogs to learn that stuff (we've done this ourselves). We save people time because we say "Hey, so you have a navy blazer. Here are things that go well with it: ..." A screenshot is worth a thousands words: Under the hood we tag items (navy blazer, khaki chinos) present in fashionable looks. We then analyze the relationships between the items to see what goes with what and calculate that rating of the combination. So if both a navy blazer and khaki chinos are present in a lot of fashionable looks, it means this combination is a good one. Then, if a user has a navy blazer in their wardrobe, we will recommend them khaki chinos (products & inspiration on how to wear it).
60824
charles lee@chug2k · Co-Founder, CoderSchool
This is cool. Do you have any idea how "accurate" the recommendations are listed? Is there a way to gauge the "quality" of the recommendations algorithm? I'm kind of clueless when it comes to dressing, mostly, which is why I was a fan of Trunk Club for a while.
20634
Alexander ZaytsevMaker@alexandrzaytsev · Co-founder & CEO, Lookastic
@chug2k yes, of course! We don't display the level of accurateness yet, but internally we have this data (it depends on many factors, such as how often two items are present in the look together, how popular said looks are, how often similar items are present in the look, etc).
60824
charles lee@chug2k · Co-Founder, CoderSchool
@alexandrzaytsev I'm just personally curious which types of clothes are the hardest to match. Khakis + Blazer makes sense, but I wonder if it's easier to screw up shoes, etc.
20634
Alexander ZaytsevMaker@alexandrzaytsev · Co-founder & CEO, Lookastic
We use big data to analyze the user's wardrobe and give personal style & shopping advice. Our team consists of 2 girls and me (we've been friends before Lookastic). I'm the tech guy, Negina runs operations and Olga takes care of content.
5649
Sean Percival@percival
What are your plans for monetization. Selling them the outfits you recommend?
20634
Alexander ZaytsevMaker@alexandrzaytsev · Co-founder & CEO, Lookastic
@Percival we don't sell the whole outfits, we have affiliate links to items present in the look. Basically, all looks on Lookastic are shoppable (http://lookastic.com/4174). We have ~80k visitors monthly. We are growing 30% month-to-month. We also have a flexible search engine, allows you to search for looks with certain items (Navy Blazer + Jeans: http://lookastic.com/men/navy-bl...). Ultimately, our goal is to become The Internet Fashion database: we want know what goes with what, where you can buy it and how you can wear it.
5649
Sean Percival@percival
Looks like you are pretty good at user onboarding. Do you have data on the conversion rate of people who visit and complete the whole process?
20634
Alexander ZaytsevMaker@alexandrzaytsev · Co-founder & CEO, Lookastic
@Percival 25-30% of people who click "Get started" end up creating real accounts later (we create temporary accounts automatically so that they can try us out without the need to create an account).