The pandemic has turned retail on its head. Where there was once a solid omnichannel strategy, there’s now e-commerce. The issue facing many retailers is that while having invested in a content management system, an e-commerce platform and an analytics package, they’re failing when it comes to the most important part of the equation – AI-powered search.

The state of online retail in 2020

According to National Australia Bank’s Retail Sales Index, in the 12 months to September 2020 (the most recent figures available at the time of writing) Australians spent $40.9 billion online, a level that is 12 per cent of the total retail sales estimate. Most significantly, this figure is about 38.7 per cent higher than the 12 months to September 2019.

Clearly, the pandemic has put a rocket beneath online sales.

This is backed up by reports from Australia’s leading fulfilment organisation, Australia Post. Its November 2020 online shopping update found online retail growth tracking strongly at 72.9 per cent year on year. Between March and October, over 8.5 million households shopped online, an increase of 13.7 per cent year on year.

As you can tell, online is the new shopfloor, showroom, channel and delivery centre. So what can be done to make sure customers can find what they want, when they want it, and have it delivered in an appropriate manner?

As I mentioned earlier, businesses can have most of the right systems in place, and still fall down when it comes to giving customers the power to find what they want on a business’ e-commerce site.

The answer is search, powered by artificial intelligence and machine learning (ML).

The AI revolution

When a new customer arrives on an organisation’s website, the first thing they do is search for what they want, or look for the business’ locations if they’re doing click and collect. And with the boom in online shopping, for Australia’s largest retailers, this is going to happen tens of thousands of times per day. Even smaller retailers will have to deal with dozens, if not hundreds, of requests per day.

The traditional retail online search engine has been rules-based. They’re static and can’t adapt to the customer, their purchase history or their request.

The worst that will happen is they’ll search for an item, and won’t get a result because the words they used didn’t match the static search engine’s knowledge graph. At best, they’ll get a laundry list of results, which forces the customer to scroll through and try find the product best matching their search. Chances are, they’ll get frustrated and click over to a competitor’s website in the hope of finding what they are after.

The goal, then, is to improve the customer experience (CX) and help them find what they want straight away. This is underscored by a recent report from Adobe and Econsultancy which surveyed 600 top retailers globally, finding that when it came to differentiating themselves, most retailers are relying on CX.

Similarly, a report from Jabil found that 95 per cent of retailers are relying on tech to improve CX, and that innovative tech is needed to meet the high standards of shoppers.

This is where AI-powered search comes into play; it is the glue that connects the content management system (CMS) and the e-commerce platform. Search organises the user by intent, through a query in the CMS and then leads to execution when the consumer actually buys something on the e-commerce platform.

If the consumer can’t find what they want, the whole system falls into disarray.

AI has the ability to read unstructured data and then convert it into the consumer’s intent to buy. The consumer benefits because they don’t need to know the exact words for their search—the system infers it by what it’s learned over time, and will constantly improve the more searches that are done on it.

Investing in search powered by ML and AI is an investment in the future. And with retail moving permanently online, it’s an investment retailers can’t afford to miss out on.

Scott Ho is vice president Asia Pacific and Japan at Lucidworks