28th July 2017

Buying insurance has a reputation for being lengthy, confusing and, at times, something of a chore. The industry is turning to InsurTech to shed these perceptions, with data analytics, artificial intelligence (AI) and mobile technology trending as the top areas for investment in insurance innovation.

Combining aspects of all of these trends, image recognition is one technology that has huge potential to transform the text-heavy and time-consuming insurance shopping process.

Image recognition enables AIs to identify and understand the content of static digital images. The technology relies on machine learning to be effective and is 'trained' with massive amounts of labelled data to be able to classify the content of an image.

Research by tech giants such as Google and Facebook has seen image recognition technology advance in leaps and bounds over the past few years, as companies use the billions of images submitted to their platforms as training material for their image recognition AIs.

Now, image recognition AI is able to decipher the content of an image with greater accuracy than a human being. While human beings are able to correctly label an image 95% of the time, an AI surpassed this in 2015 by identifying images correctly 96% of the time.

Image recognition is already here and part of our everyday lives. Facebook uses the technology to identify your friends and recommend tags in uploaded photographs. If you are on holiday and cannot read any signs, image recognition enables you to use a smartphone camera to translate the text in real time. The technology can even recognise your mouse doodles and transform them into more polished clipart drawings.

The industry is already able to provide customers with tailored premiums, but image recognition technology enables insurance retailers to further streamline the quotation journey while encouraging customers to engage with insurance products.

For example, CDL's Jay chatbot uses image recognition to draw information from a photograph taken by a mobile device, whether of a driving licence, a car or even the family pet.

This can also be integrated with call outs to other data source in order to authenticate user details and confirm any convictions or accidents, as well as vehicle make and model.

As the old adage goes, a picture is worth a thousand words, and Jay's automatic completion of input fields via image recognition significantly cuts down on the time spent applying for an insurance quote and eliminates scope for human error.

An image-led journey transforms what was once a chore into a personalised experience unique to each user. Using image recognition, insurance retailers can breathe life back into the insurance shopping process while simultaneously offering a customer journey that is fast, simple and elegant.