By Dave Shuman Data whisperer, Consumer Products & Retail SME at Cloudera

Machine learning for personalisation purposes used to be limited to a certain level of industries, it’s now open to all. Traditional retailers probably aren’t leading the way when it comes to customer personalisation (especially those with physical stores). They’re aiming to emulate e-tailers – think about recommendation engines, this was cutting edge 5 years ago now it’s a must have.

Across the next 10 years, one way advanced personalisation might play out will be using a form of AI for automated check-out experiences. Practically this would be a real-time way for customers to pay for an item off a rail, their mobile will automatically process the online payment, however we’re certainly not here yet.

What is the future of personalisation in Retail?

While some e-tail brands have been slow to spend and analyse the cultural currency of user generated content (UGC) for commercial gain, others have been throwing their weight behind using it. These brands are learning to personalise marketing and the shopping experiences customers are served, and this will only grow.

GoPro, the company that made everyone want to own a camera again, mines their community video feed to deliver smarter marketing. They have created a seamless cycle of influence that inspires new purchases – beginning with a potential customer seeing a video and thinking it’s awesome, buying a GoPro and exploring the content channel, then recording and posting their own cool video on the GoPro channel that inspires others to buy a camera and sees the cycle repeat itself.

This process is enabled by generating data in the cloud, which is managed and processed by GoPro’s Data Science and Engineering (DS&E) team, and data analytics on the Hadoop-based data management platform, which includes Cloudera Enterprise on Amazon Web Services (AWS).

The future of retail will combine IoT, machine learning, AI and automation. All fitting together to personalise deals and content to individuals.

 Just some of the channels feeding into GoPro’s marketing decisions are logs tracking product analytics, social media data, web traffic, GoPro channel data, third-party systems, and internal ERP systems – as well as streaming and batch data. Data governance and access control are taken seriously and role-based permissions dictate levels of system access. The governance structure extends through Trifacta and Tableau systems for analysts to slice and serve up data.

It’s getting easier to understand and respond to customers’ needs. After all, there are so many more customer touch-points in the online world. When customers even add items to a shopping cart, they’re sharing a little bit of ‘content’ about themselves, and now retailers can act on these clues instead of ignoring them. Collecting the data is no longer the problem: there’s as much available as we care to collect, it arrives in real-time, and it’s cheap to store and process. The real question is what exactly to do with this UGC and info, and that’s where machine learning comes in. The real challenge is not whether to, but how best to use these tools to personalise experience for every user, even if it means knowing when to leave them alone!

There will still be an element of Retail employees manually deciding what offers and content is served to shoppers based on previous purchasing habits, rather than it being automated. There is a balance that will be struck.

UGC analysis and machine learning will continue to be at the heart of brands personalisation strategy as it allows for more ways to interpret data. However, as interesting as this technical advancement is, its potential can only be reached through user experiences that match user needs. Regardless of the way customers are served personalised experiences in the latest way, it’s the job of human analysts, marketers and e-commerce professionals to create user experiences that work for the user.

In the case of Cloudera customer GoPro, insights are taken from product usage patterns like which camera features are most popular with customers, which are proven to help guide research and development spend. And because GoPro is able to better understand the profile of a user who’s likely to share videos, they can serve a personalised experience in the right way.

How can retailers best serve their consumers where big data is concerned?

 For retailers, the ultimate goal will be to gain a 360 degree view of your consumer’s online and offline shopping experience, basically all touchpoints the consumer has with the brand and personal preference behaviours. Machine learning will be used increasingly by retailers to combine data already being captured with less readily available data. Everything from footfall analytics – which aisles consumer are spending the most and least time in as well as social media, demographic data, purchasing preferences and so on. In reality some retailers are currently operating a 90 or 180 degree view of their customers today.

The reason we’re not seeing this yet is because most retailers aren’t collecting real-time data in the right way and aren’t storing it for long enough to identify patterns, but this will advance over time with the use of AI and machine learning, and that’s exciting.


If you would like to have your company featured in the Irish Tech News Business Showcase, get in contact with us at [email protected] or on Twitter: @SimonCocking

Pin It on Pinterest

Share This