DigiDirect uses personalised product data to drive revenue growth

Australian retailer employs AI and product recommendation capabilities to drive personalised experiences digitally for retail customers

With the desire to deliver personalised experiences crashing headlong into consumer privacy concerns, marketers are having to walk a fine line between extracting value from customer data while keeping them onside.

But Australian independent speciality photographic store, digiDirect, has found a way to increase basket size by focusing personalisation initiatives onto something it already has complete knowledge of – its products.

Using artificial intelligence (AI) technology from Sydney-based developer, Particular Audience (PA), digiDirect has boosted sell-through by improving product recommendations. According to digiDirect head of digital, Haig Kayserian, his company had traditionally created complex algorithms to surface related products, such as lenses or cases. However, this method made it difficult to implement changes.

“What this allows us to do is look at the historical purchases of all of our customers and look at what they paired products with,” Haig said. “The artificial intelligence is learning what the consumer of this camera is accessorising it with.”

This means if a customer is browsing a specific camera, they will see accessories related to that specific camera, rather than from across the entire product category, with this knowledge based on the previous browsing and purchasing behaviour of other customers.

Changing customer behaviours

Haig said digiDirect implemented PA’s technology on its site in the lead up to last November’s Black Friday sales event. While he initially believed it might take weeks or months for the algorithms to build up sufficient data to improve recommendations, he began to notice changes in customer behaviour in December.

“People were adding more products into their carts because of the more relevant related product recommendations, and we were getting more items per cart per sale,” Haig said.

This uplift of 30 per cent year-on-year has been especially pleasing as Haig said accessories tend to offer slightly higher margins.

PA senior product manager, Patrick DiLoreto, joined the company in late 2021 having been a customer in his previous role as head of ecommerce in Australia at PVH Corp, owner of the Calvin Klein, Tommy Hilfiger and Heritage Brands. He said one of the goals of PA is to give clients a source of data beyond the customer.

“There is a race to first-party data that makes a lot of sense, but who is to say how that will change with the regulation around that and the customer’s ability to choose what they do and don’t want to share?” DiLoreto asked. “Our focus is we build the most unique and relevant personalisation based on understanding deep relationships between item data, rather than on the user and who they are exactly.

“We are creating value for retailers and customers through 100 per cent anonymous personalisation - building ultimate customer experiences through ecommerce without having to rely on customer data.”

PA uses a mix of technologies to refine its results, including natural language processing and computer vision. This combination is especially useful in ‘cold start’ scenarios where new products have yet to build up a body of customer interactions.

“The technology will recognise that image for what it is and make recommendations – understanding more than the name and description, but also related products,” DiLoreto said. “It creates a better understanding of the product by allowing the machine to understand it.

“If you walked into a store, the sales associate would easily be able to pull products together and present them to you, so to be able to do that at scale online is really important. And we are proud to say we do it very, very well today.”

To date, PA has extended its AI technology into recommendations, search and merchandising, and soon will launch a retail media service.

“We see retail media as one of the fastest growing advertising channels in the industry, with really strong tail winds behind it,” DiLoreto said.

For digiDirect, these capabilities will prove important as it expands its into consumer electronics category and reaches a wider audience of new customers.

“That is where some of the advances from PA have come in – staying ahead of the curve and making sure we are the ones consolidating this market and moving into these other categories,” Haig said. “We are hoping what PA is doing for us is going to help us become more of a destination website because people are going to come to our site because when they are searching, the results are more relevant to them.”

Don’t miss out on the wealth of insight and content provided by CMO A/NZ and sign up to our weekly CMO Digest newsletters and information services here.  

You can also follow CMO on Twitter: @CMOAustralia, take part in the CMO conversation on LinkedIn: CMO ANZ, follow our regular updates via CMO Australia's Linkedin company page