Everything is coming up artificial intelligence. From predicting the season’s biggest trends to personalizing an individuals’ e-commerce experience, AI is being deployed by retailers, brands and their back-end cohorts to keep the industry in step with consumers and their ongoing list of demands and expectations.
Here, Eric Brassard, chief executive officer and founder of Propulse Analytics, a platform that tracks consumer tastes to accurately prompt product recommendations, weighs in on the future of the technology, pitfalls that retailers and brands need to avoid when selecting a specific service, and how it can assist in cracking the consumer conundrum.
WWD: What about the current consumer climate has called for AI-enhanced e-commerce experiences?
Eric Brassard: The reason AI is taking off is because it’s able to accomplish known business objectives more efficiently or simply accomplish what wasn’t possible before its introduction. Today, retailers are forced to fight it off through site experience instead of through product or promotion alone. That’s not to say that product and promotion can be ignored but rather that only using these elements as a mix isn’t sufficient to differentiate yourself. With those two elements alone, e-tailers won’t convert and secure repeat buyers as much anymore.
The ability to rapidly show a shopper the product she is seeking is fast becoming a consumer desire — and soon will be an expectation. Simply put, it’s not sufficient to have a store with the right assortment for your target market. One must also make the process of finding such products easy and pleasurable.
A large retailer that offers a ton of products that are hard to find will lose the consumer to a smaller store that makes it easy to find items. Combine this with the natural desire to spend when you’re surrounded with just the right complementary products and you have a fantastic formula for conversion to buyer, repeat purchases and increase spend all at once.
I see two reasons for this AI push: The increased consumer desire to quickly find products coupled with the ability and ease to sell quickly, sell more at a time, sell more frequently.
WWD: What is key for retailers to take into account when considering various AI-platforms?
E.B.: I would first just like to state that using AI isn’t an objective in and of itself. AI is a brilliant way to accomplish things we couldn’t achieve before (in the case of Propulse, we observe shopper tastes for instance). But AI has to serve a clear business objective to be considered a valid technology. The risk here is for retailers to compromise what they need to achieve by adapting their strategy to the AI-based technology that’s available. Strategy should be served by AI. For instance, a retailer must not accept bad machine learning recommendations, especially when the brand positioning is within the luxury category.
In terms of considerations, I would say that they’re the same as that [of] any other technology [company]….Will the technology easily adapt to my systems? Do I need special expertise to operate it? Will it require a lot of my IT personnel? How long and complicated is it to implement?
In some cases, it’s worth improving one’s own systems to enable the use of an AI technology. For instance, the company doesn’t have a valid data repository with fairly clean data in it. This is useful for many current and future corporate needs to be addressed. However, you should think twice if the requirements to implement are such that you will need to invest in external technologies that require a special internal project in order to implement it. It has to be central to your strategy to do so, otherwise you should wait for new solutions to arrive on the market.
WWD: Consumer attention spans are dwindling, how can AI assist retailers and brands to address this issue?
E.B.: This is very much where AI will have a significant impact. We live in a world of instant gratification. Shoppers need to find what they want quickly or they’re off elsewhere. The importance of the ease of the shopping experience is growing rapidly. AI offers us the ability to identify shopper needs and adapt to them as they change at an ever increased speed.
This is where I believe the most important impact will be felt. Match shoppers with the right product quickly and you become the destination of choice. You’ll become the first place they’ll look to see if they can find what they want. In fact, just recommending products, which is often enough to engage shoppers, will soon not be enough. It will be about showing a majority of highly pertinent products to a shopper’s immediate mood that will become the differentiator in shopping experience. I believe this will become the norm to compete [with] over the next few short years.
WWD: What are current challenges in the AI sector?
E.B.: We are just scratching the surface of leveraging all of the significant improvements new AI is allowing e-tailers. I see two early challenges: First off, access to data. If you want to develop effective AI, you need to train it on large volumes. That’s not always accessible to start-ups and highly skilled data scientists.
I believe there will become a culture of partnership between e-tailers and development outfits that will allow for win-win conclusions. But it’s not a simple issue to solve. E-commerce organizations are not usually geared to collaborate with tech companies. Tech companies must establish the right context to make this easy for their e-commerce partners.
The second challenge is the current access to skill. Producing good, reliable results for commercial applications requires strong AI specialists that also understand enterprise software. Today finding AI-capable data scientists is a challenge. Producing commercial-grade software with top-level AI requires a combination of skills that’s even more difficult to find.
I really believe this is an early industry problem that be solved as more professionals are getting into the field. But as for the next five years or so, this will be a challenge.
More on Artificial Intelligence from WWD: