If you are anything like me, or any other technology user in the industrialized world, you have used a predictive application before. But, chances are, you probably didn’t realize you were even using one.
So, what are Predictive Applications anyway? They are purpose-built applications that use predictive analytics, machine learning and, by proxy, big data to automate decisions within an application that can’t be made simply by using determinate decision making like computer programs did in the last 80 years of computing history.
Describing an application as “predictive” refers to one particular facet of that application and, like other categories of applications before, these descriptions are additive meaning you can have predictive web applications or predictive mobile applications.
In fact, I’m basing my initial prediction that you have most likely used a predictive application on two observations: my first observation goes without saying: you have most likely used a smartphone today and chances are you are reading this article on a mobile device. My second observation is based on statistics gathered by homescreen.is, which is a mobile and web application that collects usage statistics of mobile phone users. In fact, out of the 28 most popular applications this week, almost half are predictive applications, that use predictive analytics:
- Instagram – suggests photos in the Explore section based on people you follow or geographic location
- Facebook – ranks stories based on predictive algorithms and ranking highest what is most likely to be relevant
- Google Maps – predicts traffic conditions and travel times based on other user’s data
- Twitter – suggests tweets to read and users to follow based on your activity
- Youtube – suggests videos to watch and channels to subscribe to based on your activity
- Spotify – creates music recommendations and playlists based on your listening behavior
- Gmail – determines the importance of emails and filters them based on Content analysis and predictions
- Evernote – shows related (context) information to its Business and Premium customers
- Google – need I say more?
- Mailbox – with Auto-swipe, it learns from your activities and automates common actions
- Chrome – includes Google Now, which predicts relevant and contextual Information and delivers it in-browser
The fact that Google is behind nearly half of the apps on the list (Google Maps, Youtube, Gmail, Google and Chrome) illustrates a few important points about predictive applications:
- They are already extremely popular and will soon be everywhere
- Their existence goes mostly unnoticed because they power features quietly, behind the scenes
- Like all sufficiently advanced technology, they are sometimes indistinguishable from magic
- Like all sufficiently new technology, they can break in unexpected or entertaining ways
- They favor data-rich companies such as Google, Twitter and Facebook
- They provide a distinct competitive advantage to their creators
Just like mobile applications five years ago and web applications ten years ago, adoption of predictive applications will be a powerful transformative force over the next decade. However, unlike the web application and mobile application revolution, due to their behind-the-scenes-nature, this will generally go unnoticed by most users.
Another reason for this secret predictive revolution is the adoption of predictive applications by enterprises. When convenience food chains use predictive applications to automate their replenishment of fresh food and other perishable goods, nobody notices except for controllers and shareholders who see costs decline and revenues grow. When eCommerce retailers adjust their prices based on actual demand and availability patterns, no one notices except for category managers and shareholders who see increased revenue and margin. And when energy producers and utilities reduce turbine startup time based on integrated sensors, nobody notices except for maintenance engineers and shareholders who see costs go down.
Predictive applications are the reason corporations are investing massively in data collection and data storage infrastructure. They are also the reason for the shortage of data scientists and why the next 1,000 startup business models will take a shot at your industry and see where machine intelligence can be added to extract more value.
What do you think, where the Predictive Applications will lead us in the next ten years?