Hyper-personalisation in eCommerce: show customers products they actually want to buy

Retailers around the world are investing in personalisation technology, with the aim of providing a unique shopping experience, tailored to customers’ specific needs. 

In this blog, we’re going to talk about some of the challenges that come with personalisation in ecommerce, and why we believe that Hullabalook offers a better alternative.

Challenge 1: Personalisation should not be a code word for segmentation 

Let’s start with our first problem. Most retailers rely on segmentation when ‘personalising’ their experience. What do I mean by this? Let me explain…

Currently, ‘personalisation’ on ecommerce pages is achieved by leveraging pretty broad characteristics. Rather than learning about each shopper using the experience individually, retailers use large pools of data in order to make assumptions about their customers. 

It’s easier to build experiences this way, because retailer data suggests that people belonging to similar demographics (e.g of age/gender), or people who arrive at their sites on the same user journey (e.g by redirection from social media), are more likely to purchase similar items than those approaching from divergent perspectives.

Sure, this data is helpful, but can we really call it ‘personalisation’ if it’s not specific to each user? If so, we need a better word to describe truer personal shopping experiences…

This is where hyper-personalisation comes into play.

Challenge 2: “Offline” analytics doesn’t work for infrequent purchases 

The above approach is very inconsistent and retailers agree. For example, a customer recently explained that when it comes to selling clothing, they believe that they know their audience well - this is because shoppers spend a lot of time browsing, engaging and returning to fashion ecommerce pages (you can analyse a user who visits every week and have relevant results for when they come back next time). However, when it comes to furniture, the availability of data differs because these larger purchases are more infrequent.

In these situations, it’s even more important to offer a truly personalised experience based on user actions in that session (not just previous sessions), because if shoppers don’t find what they’re looking for, it’s very likely that they’ll leave and not return. 

Challenge 3: Privacy concerns are on the rise 

Consumers are highly sensitive to data usage and storage, and adhering to data privacy regulations which are becoming tougher is essential for all retailers. These regulations mean that there is less data to analyse, making recommendations far less accurate and useful than before. 

The Solution: Hullabalook’s Personalisation Engine 

Hullabalook’s personalisation engine uses the power of hyper-personalisation to ensure every shopper who engages with a Hullabalook-powered page has a unique experience. We analyse every interaction a shopper has within their session and use this data to provide a truly personalised experience. 

For those new to the concept, ‘hyper-personalisation’ is the idea of using real-time data and information provided by the customer during their browsing session, in order to tailor their shopping experience more deliberately.

The more interactions a shopper makes, the more personalised their experience will be. The good news is that there are multiple touchpoints associated with a typical commerce journey, so we have a lot of data to play with. The more a shopper engages with our technology, the greater our ability to show them products that are right for them.

How does it work? Let’s start at the very beginning.

When a shopper arrives on your site, we don’t know a lot about them. In the first instance, we’ll recommend products based on top-level information like those referenced in the opening paragraph. The magic happens once they start interacting with our technology on your page.

Most shoppers start by navigating to a category page and then clicking on some filters. For the purpose of this blog, let’s introduce some new user journeys. 

Sophia lands on a DIY retailers website. She navigates to the paint category and clicks on the ‘grey’ colour filter:

- Hullabalook’s personalisation engine knows that Sophia is looking for grey paint.

- It responds by instantly re-ranking the PLP to feature grey paint products at the top. 

- By reacting instantaneously, we’re showing Sophia products which match her in-session needs, if moments before she was looking for wood paint, we can put those at the top of the colour matched results.

Jessica is on a different site browsing stainless steel Bosch washing machines. She later lands on a dishwasher PDP:

- A standard ecommerce PDP presents Jessica with a carousel of other products available on the site which have been rendered by calling on some arbitrary but not necessarily personal or complementary metric (Dishwashers other people bought…. You may also like…. More products in this range….)

- Hullabalook’s personalisation engine uses what it knows about Jessica’s preferences to automatically generate a Visual Bundle featuring the dishwasher she’s currently viewing, plus the washing machine she previously clicked on

Clive has added several large items to his cart already (armchair, coffee table) and is looking for  items to decorate his living room:

- Hullabalook invites him to view these items in the Room Creator experience.

- The items he’s already added to cart are shown at the top of the results grid.

- Additional accessories which complement the armchair and coffee table are recommended and can be added to the room visualisation.

Hullabalook’s personalisation engine uses all of the in-session data that each shopper gives us about their preferences, at every stage of the ecommerce journey. We’re able to use this data to present them with the products they are most likely to find appealing on that given day.

By relying on data that shoppers are willingly giving us during their session, we avoid making inconsistent assumptions, and provide them with a user experience tailored to their personal needs. 

Plus, since we don’t track them between sessions, we avoid many data privacy issues.

Now that’s what we call hyper-personalisation. 

Interested in learning more about our technology?