One-to-one personalization is the key to crushing the digital experience

In this series, professionals at Shoptalk discuss the most pressing issues facing their industries today. Write your own #Shoptalk16 post here.

You can’t have a conversation with the retail professionals responsible for the digital experience in e-commerce without turning to the subject of online personalization and where it’s headed. So, here’s my prediction:

The future of personalization is that digital experiences are actually going to be personalized — as in personalized for an audience of one. Genuine personalization will be a must-have — and soon. The era of one-to-one personalization is inevitable for three reasons:

• Consumers expect it. In the era of Google and Amazon, consumers have become accustomed to online experiences that are highly relevant and tailored to them. They expect websites and other digital sites to know them — whether they are logged in or not — and to understand their intent and the context in which they’re seeking products and content.
• Cloud computing and the ability to process massive amounts of data mean that the technology is available for retailers to move beyond sorting consumers into broad buckets and providing generic experiences for all. Constant innovation in both areas means it is now possible for those offering products and content on the web to provide the kind of experiences that consumers demand. Doing so will become a minimum requirement for those who want to remain competitive.
• Amazon. The Seattle retailer is increasing its dominance in e-commerce sales. On Black Friday 2015 alone, Amazon’s sales accounted for 36 percent of all e-commerce sales, according to Macquarie Capital. Retailers that want to compete with Amazon are going to have to establish ways to provide consumers with a memorable online experience — the sort of experience that will not only have them buying today, but coming back to buy tomorrow.

You might argue that digital retailers are already personalizing recommendations. And to some extent they are. But what generally passes for personalization today is not genuine personalization.

Segmentation and recommendations aren't enough

Many retailers today are personalizing by placing consumers into broad categories, or segments. Or they rely on personalization by association, the familiar “people who bought ‘x,’ also bought ‘y’” type of personalization.

And that’s good, as far as it goes. But it turns out, it doesn’t go very far. First, it’s difficult to reasonably manage more than a handful of segments, which means consumers are divided into categories that are not narrow enough to constitute real personalization. And providing personalization through segmentation yields disappointing results, with only 25 percent of such visitors returning within 30 days, according to our research.

When it comes to product recommendations the news isn’t much better. Recommendations focus on products, not individuals; and recommendation systems tend to continually promote best-selling products. So, best-selling products sell best, because they do. Adding new products to the mix requires constant attention. Furthermore, only a small portion of visitors engage with product recommendations widgets, with the overwhelming majority of visitors blowing right past them without even noticing they’re there.

It’s not that retailers aren’t aware of where they’re falling short. It’s that they haven’t had the staffing or technology to personalize the way they want to. To achieve one-to-one personalization manually would take an army of workers tagging products and writing rules. It would never pay for itself, no matter how much real personalization improved conversions.

Real personalization requires machine learning

Instead, the path to personalization requires powerful technology that taps machine learning and natural language processing to match every single consumer’s intent with website owners’ content. It requires cognitive computing.

Turning to such a personalization platform provides the potential to change everything. Rather than engaging a small percentage of your customers, a machine-learning system that personalizes search and the product mix in the categories that a shopper sees, touches nearly everyone who visits your site. When you think about consumer spending on your site, paths to purchase that include site search and/or category pages drive about 90 percent of your revenue.

Of course, technology alone is not the answer. The best results come from combining the power of machines to provide relevant, personalized experiences at scale with humans’ creativity, business savvy and intuition in understanding markets.

Ideally, at the center of this powerful human+machine relationship sits a digital experience manager, who is able to take a holistic view of data detailing consumer behavior and product performance. The role is vital given the findings in a BloomReach-commissioned Forrester survey that found that digital marketers undervalue site merchandisers’ input because they don’t believe merchandisers have data that can increase customer acquisition and repeat visits.   

A digital experience manager can help bring marketing and merchandising together to complement each other and improve the overall consumer experience. Improving that experience will be an imperative in the very near future.

At the core of that vital change will be providing consumers with one-to-one personalization. And given that Gartner has said that within two years, companies that have invested fully in online personalization will be outselling the competition by 30 percent, it’s a prediction that you can take to the bank.

Photo by Iskandar Shah published under Creative Commons license.

Raj De Datta is co-founder and CEO of BloomReach, winner of the 2016 Edison Award for Innovative Services. He will participate on the “New Approaches to Personalization” panel at ShopTalk on May 17, from 2:25 p.m. to 3:05 p.m.