In a recent InternetRetailing webinar, AI to survive and thrive in ecommerce with Coveo and Accenture we heard from Simon Langevin, director of product management at Coveo and Martin Bayton, business development manager search and content analytics at Accenture on the role that AI can play in transforming the ecommerce customer experience as we move beyond coronavirus lockdown. Chloe Rigby, editor of InternetRetailing.net, chaired and set the context.
- The webinar started with a brief overview of the current context – from how Covid-19 has changed the way we shop online to how retailers are using machine-learning/AI and search.
Martin Bayton of Accenture on ecommerce search
- Ecommerce search: the problem – “Customers can’t buy what they can’t find.”
- Evaluation of top 70 e-commerce sites for search maturity: websites scored against a checklist of top 10 search features.
- “For those who do engage with a search box, about 50% are more likely to convert.”
- Findings include 39% of sites evaluated did not have searchable categories – and of those that did, only 19% implemented it well.
- Ecommerce search scorecard/search optimisation quadrant: can be used to assess customers’ ecommerce search functionalities.
- Can be used to benchmark site against competitors and see how can differentiate site.
Machine-learning in search
- Machine learning to optimise results: how query chains can help. “A powerful way of connecting user search language with catalogues descriptions, tags and metadata.”
- Machine learning for personalisation: analysing user behaviour to help purchase recommendations. “This is particularly important for mobile devices where display area is limited. We can make sure the best products for the search the user is doing gravitate towards the top of those lists.”
- Deep analysis of all user journeys enables far better understanding of how well you engage customers.
Simon Langevin of Coveo on Coveo for Commerce
- Three development principles used to bring AI into search: unified, relevant and valuable
- Unified data: how can you get the data together – from the catalogue to the user journey; importance of the outcome as well as the input.
- Relevant data: how machine-learning can underpin quick instant detection of the most relevant search. finding what you’re looking for as quickly as possible and what’s relevant in your current situation.
- Valuable is being able to control it: does it bring value to your existing stack/solution
- With demonstration
The webinar ended with a Q&A session. To see the webinar in full, including the slides, speakers and Q&A session, visit the Coveo webinar page. To explore InternetRetailing’s previous webinars, click here.