4 Ways Machine Learning Boosts The Customer Experience

The cost of making predictions has dropped thanks to the vast amounts of data from multiple digital sources. Now is the time to use intelligent machines to enhance your customers’ experiences.

4 Ways Machine Learning Boosts The Customer Experience

In the 1950s, scientists began working to build machines capable of imitating intelligent human behaviour. Progress accelerated recently when we entered a new phase of machine learning—one that has led to a dramatic decrease in the cost of prediction.

Moore’s Law (the 18-month doubling of transistor intensity on microprocessors) was the driver of the previous phase—a revolution in technology hardware that advanced mobile innovation and smartphone adoption. Increased connectivity and scalable cloud-based storage catalysed a step change in the amount of data we collected and consumed.

Information taken from sensors, images, videos, and other digital sources is being used to generate a more accurate view of real-time context. This is happening in the following ways:

This deployment of machine learning is where marketers should be looking for AI to add value to their businesses today.

Current types of execution include personalisation, computer vision, natural language, and decision support.

1. Personalisation

Machine learning can help create highly relevant communication—much needed, given over 90% of online users in the U.S. and Europe feel advertising is more intrusive today compared to two years ago, according to a HubSpot and AdBlock Plus study.

The most recent Current Account Switch Service campaign from BACS used machine learning technology in partnership with programmatic ad buying to match audience profiles with highly targeted video content. The campaign saw an 870% uplift in acquisition click through for potential customers.

2. Computer Vision

This technology can detect everything, from objects and people to complex scenes within both images and videos.

Twitter’s Magic Pony technology makes pixelated images sharper and enhances the quality of video captured on mobile phones in poor lighting conditions. This enables Twitter to improve its streaming abilities by lowering data usage and supporting its new global Progressive Web App—itself a predictive mobile experience.

GumGum tracks logo impressions across broadcast TV, online streaming, and social media. Using a Media Value Percentage (MVP) methodology, it measures what it would have cost to buy an equivalent amount of reach. During the 2017 Emirates FA Cup Final, the platform predicted that Arsenal sponsor Puma generated $4.73 million of MVP versus $0.9 million for Chelsea’s sponsor Adidas.

3. Natural Language

Speech recognition performs today with error rates far lower than humans, and Google’s Cloud Speech API recognises over 80 languages and variants, supporting a global user base. By leveraging advanced linguistic data and cognitive technologies, marketers can confidently create the most engaging content for any digital channel.

U.K. startup Relative Insight turns language into data and helps brands predict the words that will emotionally connect with specific audiences to increase short-term engagement and build lasting, long-term relationships.

4. Decision Support

Machine learning enables the prediction of a customer’s likely next action. Digital tools and services increasingly contain features that provide advanced recommendations and help users make decisions faster.

The latest advances in mobile operating systems offer customers decision support between different user interfaces. The coming Android O release will reportedly include a Chrome feature called ‘copyless paste’ that will use machine learning to save users time by proactively offering to share information between apps.

Making It Happen

Organisations will need to adapt to handle changes in two major areas—technology and people. Software has changed, and people skills will need to change too.

Machine intelligence is coded in a different way to traditional software. Instead of optimising a long list of conditional if-then-else statements, the new method predicts what humans would do given a specific set of inputs.

As the machine intelligence tech stack is maturing, it’s becoming increasingly accessible. Huawei announced earlier this year that they’re building AI functions into a new chip that will launch later this year. Their EMUI interface supports deep machine learning and smart computing capability.

The new methods used by machine learning return open-ended results, meaning mistakes are likely. As this technology rolls out, organisations will need staff with strong people skills who can manage and communicate unexpected outcomes.

As we’re at the beginning of this journey, today’s intelligent machines need training by humans, and work best in collaboration with real people.

Smart personal assistant x.ai schedules meetings on your behalf via email. Training the system involves allowing it to observe human interactions so it can learn and predict human responses.

Facebook’s M virtual assistant is positioned as a mixture of AI and human answers, and Cisco recently acquired conversational AI startup MindMeld, around which it will build the company’s cognitive collaboration team.

The Next Step

The current collaboration between humans and machines is greatly reducing the cost of making predictions—and it’s all powered by data. This is helping brands generate a better picture of their customers’ true context and serve them an enhanced experience. And while we’re only in the foothills of AI right now, brands that have been bold enough to experiment have seen real results.