Startup of the day (July 2017)


    Continuing the series of Startup of the Day digests, today I present the most interesting projects for July. If you want to get acquainted with the rest, then I ask in my blog. Entries are available on Facebook , ICQ and Telegram .


    Twiggle


    About eight years ago, there were many such projects, but it seems that the quality of the product was not enough for anyone, everyone died. Maybe now the time has come and now it will work out. Twiggle sells e-commerce services intelligent search technology with a natural language understanding.


    image


    As a demonstration example, they use the query “Open back button dress for a wedding under 200 not blue” - try to drive its analog into any Lamoda, in the best case it will turn out blue dresses. Or, for example, “under 200” - the most stupid consultant in the store will perfectly understand that this is about the price, but some online store - no. Even Amazon does not understand! The query “tv under 200 dollars” easily falls out in offers there, it is popular, living people want to look like that, but look what the “store of everything” finds in response.


    Twiggle's demos, of course, work. And “not blue”, and “under 200”, and “dress shirt” with “without” in other tricky examples - the startup understands all this correctly, and the results of the delivery correspond to human expectations. Unfortunately, it is impossible to look at a project in real life, but not in presentations - no one has fully implemented it yet, although "some clients already have A / B tests."


    Technically, Twiggle works as a SaaS, providing partners with essentially two API methods: “load the database” and “get the search results”. Integration, accordingly, is simple, the design does not need to be changed, it is easy to use. The Russian language, unfortunately, is not supported, it will not work to introduce miracle weapons into local projects, but you can try the domestic counterpart - Detectum


    Cardlytics


    image


    Cardlytics elegantly trades banking secrets. Indeed, in fact, any bank knows so much about its customers that no search engine has dreamed of, and they want to sell targeted advertising. Cardlytics implements one of the possible scenarios.


    Part of the startup’s software is installed inside the perimeter of the bank, has access to all transactions, and based on these data it somehow segments users. At the beginning of an advertising campaign aimed, for example, at wealthy lovers of beer restaurants, selected customers are transferred to the system of the bank itself, and it sends out SMS or other notifications about the super-discount and increased cash back. An unsuspecting person rejoices at a successful gift (“you have to, as you guessed it”) and sets off to get acquainted with the new institution. Cardlytics and the bank receive their commission, and the bank also receives the additional loyalty of a happy customer.


    Personal data is protected on two levels. True user names can not be transferred to Cardlytics, it is enough for him to exchange completely conditional identifiers with the bank. In addition, the part of the software located in the perimeter, in the perimeter, should store everything that it has accumulated, only general statistics should be sent out.


    Theoretically, everything that Cardlytics does can be done by the bank itself: Smart Algorithms on BigData are very arbitrary there, and you can use straight-line filters with access to card spending. In practice, small banks have neither development resources nor the ability to reach out to large advertisers, and a startup gives them a single entry point.


    Now 1,500 American banks are working with Cardlytics, together they see about 10% of all American spending on cards and checks. For comparison, in Russia such a share would correspond to position number two after Sberbank. The volume of investments corresponds to the scale: for almost 10 years of life, the company "mastered" $ 180 million - try cheaper to connect 1,500 small banks to your platform. Serious revenue has gone only recently: 80 million for the year 2016 with rapid growth.


    Tile


    image


    A new gadget for forgetful hipsters offers Tile startup. He sells small electronic cards with a Bluetooth chip that users put in their wallet, hang them on a keychain with keys, attach dogs to a dog’s collar, or somehow connect with something they’re afraid of losing. Now, if fears come true, just press a button in a mobile application, Tile will ring and a thing will be found. A small wow-effect from product designers: if the phone is suddenly lost, everything works absolutely symmetrically - you need to press the button on Tile and the smartphone will ring.


    This approach works great if the keys are rolled up somewhere within the apartment. But when they flew out of their pocket somewhere on the road, Bluetooth, most likely, would not finish. In this case, Tile offers two mechanisms. The application can show the point on the map at which it last managed to contact the card. If this is, say, a job or a favorite cafe, then the problem is most likely resolved. If the wallet fell out on the bus and the geographical coordinate of this place is completely meaningless, then the community enters the fray. When the card loses sight of the host phone, it starts sending an SOS signal to everyone. Any smartphone with the Tile application that is near the loss will receive a signal and forward it to the cloud, from where information about the things’s movements will be sent to its owner. Of course, no one gives an absolute guarantee,


    One card costs twenty-five bucks, when you buy several at the same time, you can get a significant discount. The battery lasts a year, after this period the company offers to exchange cards for new ones for an extra charge - in fact, a hidden subscription model with an annual contract, I do not quite understand why it is not drawn up explicitly. Now the startup has reached a turnover of $ 100 million a year, and in the entire history it has sold 10 million devices, of which more than half over the past 12 months. Investments are growing in about the same way: of the 59 million received, 25 came this May.


    Levelup


    image


    American LevelUp offers users pre-order kiosks - as in McDonalds, only through the application, and not in the iron box. The office employee goes down for lunch, enters the elevator in an nearby eatery, and by the time he approaches, takeaway food will be waiting for him. The user does not have to waste time, the restaurant does not need to pay the cashier, the startup collects its $ 69 per month for a subscription, even other, less advanced guests are satisfied - the queue has somehow become shorter.


    This one function and subscribing to it is already a big business, LevelUp talks about 300 thousand connected points. Of course, it’s impossible to estimate revenue by multiplying them by $ 69 and by 12, the networks have huge discounts, but there is still a lot of money. And subscription is not everything, another source of income is advertising and promotion. A person launches the application when he goes somewhere for food, and his location is known - the perfect moment to advertise some new burger from a cafe nearby.


    The most controversial way to monetize is CRM. Orders and payments go exclusively through LevelUp, and they come to the catering establishment already depersonized. Knowledge about the habits of users is a startup and trades. The cafe can segment its guests, make them special offers and even send pushies.


    It is interesting that the product is young, and the startup is very old, but before it was doing completely different things - it did a kind of Appe Pay through QR codes. LevelUp changed its business model about a year ago, but in all the articles in the press, and even in its own Help section, an old version is described. If you count from the moment of the pivot, then the investment received 42 million dollars, 37 of them more recently.


    Meet For Charity


    image


    Sometimes a startup can be just a page on a social network. The Russian startup Meet For Charity has no IT platform, but this does not prevent it from achieving success.


    The idea of ​​the project is very simple - it is a regular charity auction for the right to meet an interesting person. Every day a new guest (“lot”) is announced, and the page subscribers directly in the comments make their bets - who is willing to pay what amount for the opportunity to chat with this person. After the end of bidding, the winner transfers his bid to the charity fund, and he shares 15% of the commission with Meet For Charity. As a result, absolutely everything is won: the world gets more good, the startup makes money, and someone gets an important meeting for themselves.


    Guests "in the lots" are very different: businessmen, show stars, athletes. At the time of writing the original post, Kortnev was “playing out” from the “Accident”, and a record sum of 1.2 million rubles was obtained two months ago for a dinner with Dmitry Grishin. For comparison, similar auctions for a meeting with Warren Buffett at the beginning of the 2000s ended at a rate of 20-25 thousand dollars.


    The current average final price of the lot is 90,000 rubles (a year ago it was 20,000), 15% of the commission over the past six months gave more than one and a half million revenue, with almost zero expenses and organic growth. Inorganic founders are looking for investments to grow - they want to leave the Facebook page to a mobile application and go out with their model to other countries. I think we should at least wish them good luck!


    Also popular now: