The moment you open Screenius, you’ll see this app is visually distinctive, not looking like any other mobile app you own – what, no pastel colors? no big flat buttons with san serif fonts? no lists? no tables? Hopefully, you’re someone like us, looking for the unique and the different.
As a feat of engineering, Screenius is intricate and finely made, like a good Swiss watch. If you know software design, you may appreciate that this craftsmanship is all client-side code (so as far as we are concerned, your data remains private to you and your iPad).
The aim of Screenius is to take the work out of finding interesting videos to watch. We want to minimize your typing, searching, scrolling, etc., automate whatever we can, to make your iPad as easy to use as a TV, but a TV with a brain.
We learn your taste by watching what video you consume (and don’t), and automatically fetch videos, two at a time, to present to you. In History, you can go back and scan through any video without ever bookmarking or tagging – we keep track of all videos you don’t give a thumbs down to.2Why do I pick from only 2 videos?
Why can’t I see an endless list I can pick from? Trust us, there is an infinite supply of video ready and waiting for you – but why should you have to page through umpteen entries in an infinite list of video descriptions or wander through a never-ending grid of video thumbnails when all you want is to see something interesting?What could be simpler than an immediate choice – left or right? A or B? We use this simple, quick choice to hone our personalization algorithm – every set of choices we present is an automated A/B test to help us learn more about your taste, and only your taste.3Why does Screenius shuffle randomly between my selected topics?
This is not North Korea – we assume you are interested in more than one topic. Most people have a broad set of interests, and different topics may hold different appeal depending on your mood or time of day or day of week. So we let you select any number of topics you enjoy, then we find videos from those topics for you automatically, filter them by what we know of your personal taste, and you always have the option to change your mind (change your topic) with a tap of a finger.4Why can’t I save videos into a “favorites” folder?
Why should you? Memory is cheap, so why don’t we just preserve everything for you? Less work for you. If you really don’t want to ever see this video again, just rate it negatively and we’ll trash it. Otherwise, don’t worry, be happy.5Why do I have a score?
That’s groundwork for future features. You can brag about your score or ignore it for now. As you wish.
Click here for a recent interview that captures some of our start-up experience.
How user interface and user experience can save us from data overload
Talk given at Silicon Valley New Tech Meetup (thanks be to Joe, Nisha, and Vincent, who made a guest appearance after wandering around the world for some time).
Normally, here in Silicon Valley a graph that is up and to the right is a happy graph, a graph of good tidings – sales, customer growth, etc. But it can also be a graph of horror – population growth, global warming, or other data plots leading to the end of civilization as we know it. This graph here is a graph of terror. This is Cisco’s prediction (and they should know) of data growth over the public Internet – they have looked at the bits flowing through their equipment on a global scale, noted the rate of change, and plotted the trends out. An exponential rise in information overload is headed our way in the next 4 years. Most of this flood is going to be video – 90% of the zetabyte per month data expected in 2015 will be video bits. If you did nothing but watch video every second straight for 5 years, you would be able to watch the amount of video that will be coursing every second over IP in 2015.
What’s scary is that we humans are not expecting an exponential improvement in our ability to absorb information in the next 4 years. Nor is anyone predicting an exponential increase in the hours of the day nor an exponential improvement in search technology in the next 4 years. So how are we going to cope? Three approaches offer hope:
We are naturally social creatures – language probably was invented when one ape wanted to tell another ape what to do (‘I’m only trying to help here’). So any technology that lets us leverage the wisdom of others is a good thing.
It’s not that personalization is a bad idea – it is a great idea and essential if we are to survive the coming information tsunami. But this first generation of ‘personalization’ has been implemented in a mostly evil way – it has been implemented as personal data collection so the data can be packaged and resold to third parties, and not built for the primary benefit of the customer. This anti-consumer behavior will not stand. The marketplace has proven unforgiving to anti-consumer behavior once a consumer-friendly alternative arises. The future demands personalization that is designed to service the consumer, not the advertiser.
Our portals to video content today suffer from sheer ugliness. The jumping off point for products on the market today is a static page (Web) or a table grid (TV) filled with text, whereas video is inherently visual and dynamic – perpetual visual motion. Also, there is tremendous groupthink going on, where companies spend most of their time looking over their shoulder at their nearest competitor and copying features from one another rather than thinking in fresh ways. You can see patterns in this groupthink by examining the food chain for video discovery competitors.
Big fish, the size of an Amazon or a Google, have optimized for fast browser load times, all else be damned – white pages, lots of text, blue links, and small thumbnails scattered throughout thousands of pages to search through.
Up-and-coming but not insignificant players like Netflix or Walmart.com, share a similar look: less text than the bigger fish, DVD cover thumbnails arranged in endless grids, stars instead of a description, their logo/brand color featured prominently, etc.
Smaller still, recently funded start-ups (Plizy, Boxee, Deja, ShowYou shown here) have also copied one another: black replaces whitespace, thumbnails are larger and smushed together in an endless mosaic, filmstrip, or page layout.
And then there’s Screenius:
As you can see, we don’t look like anyone else. Our interface is unique – no one is copying us (yet) and we didn’t copy anyone else. On the one hand, this is just a testament as to how insignificant we are in today’s food chain – we are just 3 guys bootstrapping our way to first product. But we believe, that in the face of infinite content, the simplest choice will be the best for consumers – left or right, A or B.
Our UI is automated A/B testing. Rather than present 50% of the population with one option and the other half with another, we present 100% of our customers with only 2 options at a time. Depending on what action they take (or don’t take), provides data we use to adjust weights in our personalization engine for that customer. The interface is dynamic, visual, and basic.
And it is just the start. Screenius delivers a novel approach to video discovery, one that scales gracefully to infinite data, personalizes to every single individual, and leverages the wisdom of others in a fun way. Please download our current iPad app, send us feedback, and stay tuned – amazing things are in the works.