In part 2 of our Segmentation & Insights and Onsite Content Personalization blog series, we dive into the root of what powers retailers to deliver personalization: leveraging data. As we explored when we first revealed our product offering, data – the gathering, analyzing and actioning of it – is a fundamental part of the personalization journey. But to better understand how retailers use data to increase personalization, we must address the role of Artificial Intelligence in retail and the ways data powers AI.

 

Personalization used to be a “nice-to-have” functionality for an ecommerce business. But with the rise of a more intelligent and automated retail ecosystem, it has become a “must-have”. It is now expected by consumers who are focusing on the quality of a shopping experience as a whole rather than individual factors like product availability or pricing.

This all sounds great in a perfect world – but from a retailer’s perspective, it can be pretty daunting at first. The true path to achieving personalization means that it’s important to address the very term on the mind of any retailer set on success: AI.

 

Defining Artificial Intelligence

If we look at the ecommerce market, AI refers to solutions that use data to automatically adjust its output towards a given set of goals.

Common goals AI optimizes for include:

  • Conversion rate
  • Average order value
  • Return visits
  • Reduction in bounce rates
  • Repeat purchases

All of these KPIs ultimately drive the retailer’s revenue. In fact, according to recent research by Retail Week, AI is projected to boost profitability for retailers by 59% by the year 2035.

 

Transactional vs. Behavioral Data

Now, it’s easy to talk about AI on the high level but what does it mean in practice?

Let’s start with onsite real-time browsing behavior in combination with transactional data.

Many retailers still use CRM-based customer segmentation tools to address the need of personalization. However, transactional data is typically old data that only reflects a customer’s past purchases. This data becomes available and actionable only at the end of the browsing session if the shopper converts.

Research over the past 7 years has shown that transactional data only accounts for 1.6% of the data captured in an online store.

Our own research has also revealed:

  • On average, it takes 3 visits for a shopper to make a purchase (and 5 visits from the first purchase to make a second one)
  • Only 20% of repeat customers order products from the same product category of their initial order.
  • This means using transactional data to deliver personalization is relevant to only one fifth of customers.

So where is that crucial 98.4%, you ask?

It’s in behavioral data, which reflects a shopper’s intentions and buying pattern as they interact with a store. Many retailers think they already track onsite browsing behavior; but is this data tracked in real-time?

Data made available by tracking browsing behavior typically becomes available and actionable only at the end of the browsing session, after the shopper has exited the site. At this point, it is already too late to affect the current shopping experience because the shopper is long gone.

The true value in leveraging data onsite is to do it in ‘real-time’, using AI to track the onsite behavior and then update the pages that the shopper sees during that same live session to deliver the most relevant shopping experience possible.

 

AI in Practice

In the below videos, we can see all of the browsing behavior being taken into account including:

  • Interaction with a welcome pop-up that includes a shipping reassurance bar
  • Viewing a page of different brands
  • Viewing a brand-specific page – in this case: Adidas

  • Browsing a specific Product Detail Page
  • Checking out some recommended products and
  • Adding an item to the cart

 

 

By not collecting this information, all you are left with is personalizing a shopping journey based on transactional data which may not even exist.

As data on many shoppers is collected and browsing behavior is analyzed, a deeper understanding of the store is formed. In this visual representation of a product graph, we can see things like which products are most commonly bought together or what products shoppers actually go on to buy most after viewing a particular product.

Once the data has been analyzed and prepared like this, it is time for AI to go to work. We can see that in practice with brands like Function18 who leverage both transactional and behavioral data in real-time, enabling AI to deliver completely unique shopping experiences.

 

Personalization Comparison: A Look at Unique vs. Standard Shopping Experiences

In part 3 of our Segmentation & Insights and Onsite Content Personalization series, we’ll explore a side-by-side comparison of the kind of shopping experience that a shopper expects versus an unfortunate reality and how it affects retailers.

Eager to start delivering completely unique experiences to your customers? Get in touch with us today for a free demo!