Predictive UX
Picture this, a coffee machine that makes your coffee just when you begin to think it’s a good time for one, lights that automatically dim when they sense you have a headache (self-induced or otherwise ;), a music app that knows your mood a plays that perfect song to go with it, or your car predicting that traffic jam before it happens & steers you clear.
Predictability is the key to a sustainable business model. In the connected world prediction will revolutionize the future of interaction. Ai approaches to engagement, with the colossal volumes of data, will be the only way to stop us from being overwhelmed, technically & anthropologically.
Prediction is coming, it will be scary & beautiful at the same time. Will humanity become the guinea pigs to Ai or will we use it to make out next evolutionary leap?
What Is Predictive Modeling?
Predictive modelling uses statistical techniques to predict future user behaviours. A understanding of the intricacies in designing predictive analytics, requires you dive deep and comprehend what a predictive model is.
A predictive model uses historical data from multiple sources. You need to initially normalise the raw data by cleansing it of anomalies, then pre-process it to fit a suitable format to facilitate analysis, then applying statistical models to the data to draw inferences. Each model comprises various indicators—that is, factors that would likely impact future outcomes – called independent variables, or predictor variables.
The use of a predictive-analytics algorithm to UX design, doesn’t result in changes to a user interface. Instead, the algorithm presents users with the relevant information needed.
A simple illustration of this capability from the eCommerce perspective: A user that has recently purchased an expensive mobile phone would likely need to purchase a cover to protect it. As a result, that user would receive a recommendation to buy a cover. The site might also suggest other accessories such as headphones, memory cards etc.
Other examples of predictive modelling could include. Spam filters use predictive modelling to identify the probability that a given message is a spam. Facebook uses Deep Text, a form of unsupervised machine earning, to interpret the meaning of users’ posts & comments. In CRM, predictive modelling targets messaging to those customers who are most likely to make a purchase.
more to come
Continually updated & living article
Coming Soon
I will be looking at areas such as Predictive modeling, Decision Trees, Linear Regression, Data-Driven UX Research, Measuring Cognitive Changes & How to Leverage Predictive Models in UX Design.
I will also open up the opportunity for input, collaboration & points of view in the future.
