E-retailers need to keep searching beyond key words

The issue surrounding artificial intelligence is that whilst it inspires a to of shock and awe (and for the most part, incomprehension), there is not much intelligence that is going on in the background. Even as e-retailers struggle to adapt to the changing marketplace and consumer patterns, the technological front perhaps leaves them the most terrified, because it is in its infancy.

As strides are made on this front, online retailing and the way people shop will be in many ways almost unidentifiable from what we are used to currently.

First, it is critical to understand what is happening currently. Ads online target the consumer based on what he/she is doing on the internet. If we are looking for beauty products, we get targeted ads for beauty products. This gets increasingly sophisticated: if I am looking at a particular genre of books over time, then ads pop up for that specific genre, in a sense “guiding” me to my next purchase.

What we are being told is that this is artificial intelligence, some sort of “deep learning”. In point of fact, this is nothing but advanced key word searching, and there is not much science that is going on behind this. Anything that we type is being tracked by a computer. It’s that simple.

If we however, start conducting thought experiments, and imagine where this might take us, things start to look a little bit more interesting. Imagine if you will, a shopping assistant that knows how you cook, what you cook, and what utensils you have at home.

Imagine if this “assistant” via an app were able to assist you while you cooked, making suggestions in response to the changing ingredients that you were putting into the pot. In order for this to happen, the “assistant” would not only need to know your goals and your style, but would need to respond via the camera to the changing decisions that you were making (it would need to know bit of physics as well!).

This may sound both futuristic and somewhat disturbing, but the reality is that we are not far away from it in a technological sense. The problem is that currently, retailers and artificial intelligence practitioners focus on key word searches.

But the reality is that these key word searches are not the solution. What is required is a sense of personalisation; computers need to model what we are doing — presumably through speech and vision — what our goals are, and what assistance on a dynamic level can be done In order to get there.

This level of personalisation has been very intuitive for a while at the traditional retail end, especially in luxury. It then takes us back to the first principles of e-commerce; a level of customisation that moves beyond targeted demographics and facilitates individualised solutions.

It is when these technologies are in place does the arc of eCommerce truly start to thrive. Instead of then focusing on traditional issues such as logistics and income and/or ethnic based markets, the front-end of the platform returns to its significance. Similar to the best salesmen in a store who identify with their customer and make it their goal to determine at what is optimal for the customer rather than merely closing the next sale.

In various parts of the world, where the efforts are not just being focused on keywords, there are investments being made in developing exactly these kinds of platforms. They offer a glimpse into an exciting future, one which is moving at an even faster pace, as traditional retailers (and malls) struggle to make sense of what is happening even in this key word search universe.

It is breakthroughs such as these that will move eCommerce beyond the current point-and-click fuelled by targeted words and truly change consumer behaviour.

— The writer is head of digital operations at GCP.