Last week diginomica highlighted two major retailers and their generative AI ambitions. Walmart CEO Doug McMillon said:
Our approach to new tools like generative AI is to focus on making shopping easier and more convenient for our customers and members and helping our associates enjoy more satisfying and productive work.’
Meanwhile Etsy CEO Josh Silverman argued:
For Etsy, more than most, [AI has] the opportunity to unlock really incredible gains, given that there's 115 million things for sale on Etsy and none of them map to a catalog. So the ability for AI to really help to organize the world for us, I think, is a huge opportunity.
Throw in John Lewis Partnership’s AI-centric partnership with Google and three’s the charm with the retail sector and the generative AI hype cycle. But how realistic are these claims at this stage, six months after ChatGPT threw the proverbial cat among the AI pigeons?
By happenstance, a couple of global studies with a retail focus have just been published that provide a realistic assessment of the true situation around adoption and use cases, rather than just the aspirational intentions of various retailers. That’s not to say that those aspirations won’t translate into real-world exemplars, but it’s pretty much the case that the retail sector, like every other one, is still very much in exploratory mode right now.
Salesforce shopping
For the fifth edition of its Connected Shoppers Report, Salesforce surveyed 2,400 shoppers and 1,125 retail industry decision makers across Australia, Belgium, Brazil, Canada, Denmark, Finland, France, Germany, India, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Spain, Sweden, the United Kingdom, and the United States.
Among its top line findings is the bold claim that 92% of retailers say they are investing in AI more than ever to Excellerate shopping experiences. Impressive though that number is - or perhaps, because that number is! - there’s a huge caveat in the report:
Although it’s not yet clear if retailers are actually implementing generative AI in their workflows, or still experimenting, 59% of retailers say they are already using it to help store associates make product recommendations to shoppers. And when it comes to online experiences, 55% are exploring or using it to create a conversational digital assistant to help online shoppers find products.
So, see above re: curious experimenting rather than practical deployment in the main.
That said, retailer respondents do see potential for generative AI. According to the data, 59% are using it to generate product recommendations for store associates to use in store, with a further 39% evaluating this.
Some 55% are using the tech to create a conversational digital shopping assistant to help shoppers find the right product or service, something 41% more are considering.
Meanwhile 52% are tapping into AI to create virtual models for product detail pages - with 44% also looking into this - while 51% are creating personalized product bundles, with 46% evaluating this option.
Where generative AI may have a more immediate impact is in automating retail marketing. The report states:
Retailers see possibilities for generative AI when it comes to jumpstarting content creation for channels such as display ads, social media, and email. They also believe generative AI will help them automate the development of personalized marketing emails and promotional offers, saving time and money.
To that end, 58% of retailer respondents says they’re using the tech to produce creative assets for ads, emails, social media and websites. The idea of generative AI being able to help pursue the marketing Holy Grail of personalization is one that seems to have caught the imagination, with 56% saying they use it to write personalized marketing emails, 54% to create personalized promotional efforts, while 53% are auto-generating personalized product descriptions.
How personal this personalization is isn’t clear, but the p-word is also at at the center of generative AI’s potential impact on service in the retail sector. The report suggests:
Rising case volumes are putting pressure on retailers to efficiently scale service. To do this, service organizations are increasingly turning to AI: 88% of service decision makers say their use of AI has increased.* Generative AI’s capabilities hold promise in service. Retailers see possibilities for generative AI in terms of drafting personalized responses to shopper inquiries, leveling up chatbot interactions, and automating the creation of knowledge articles
With that in mind, 57% of retailers say they are generating personalized responses for agents to quickly email or message customers, while 54% use AI to power chatbots for service. On a more ‘back end’ basis, 53% use it to automate the creation of knowledge base articles, while 52% use it as a tool to create summaries of customer service cases.
AI, ML, CV
Meanwhile a survey from Honeywell - The AI Revolution in Retail - finds that nearly six in 10 retailers plan to adopt AI, Machine Learning (ML) and Computer Vision (CV) technologies over the next year, while nearly half (48%) of respondents cite AI, ML and CV as the top technologies to have a significant impact on the retail industry over the next three to five years.
Among other key findings:
- 38% of those surveyed are using these technologies for select use cases or regions
- 35% are using them on a larger scale
- 24% are in a pilot phase or in discussions
- Only three percent said they were not using these technologies at all
Most respondents see the technologies as tools to augment and maximize their workforce, rather than to replace employees, with only seven percent saying that their primary purpose for these solutions would be to reduce human labor.
The survey results also highlight three primary barriers to widespread adoption:
- Budget restrictions (39%)
- Difficulty in demonstrating business value (29%)
- Lack of internal expertise to maintain the technology (21%)
The report states:
With cost implications at the forefront of what’s holding retailers back, a greater understanding of viable early adopter use cases is needed, together with how these use cases align with budget priorities. For example, one AI application that is highly likely to deliver business value for years to come is AI-powered live chat services to provide rapid responses to repetitive customer queries. This means customer service teams can scale efficiently during peak periods without increasing employee headcounts – a benefit that’s widely sought after amid labor shortages and growing costs.
Other ways for retailers to get their foot in the door with AI include targeted marketing and offers using data from customer purchase history. AI algorithms can develop strategic offers based on each customer’s unique behavior on e-commerce platforms, putting the retailer at the forefront of customers’ minds. Because retailers often lack the internal skills to maintain such tools (21%), strategic external expertise will be crucial to optimizing usage, unlocking long-term cost-efficiencies and improvements in AI, ML and CV.”
My take
A lot of aspiration on show here. What will be interesting is to revisit this in a year’s time when there will - theoretically - be a lot more data on the practical outcomes and benefits (or otherwise) of this early experimentation in action.
What about the other side of the coin? How do shoppers regard generative AI’s usefulness?According to the Salesforce study, only 17% of shopper respondents say they’ve used generative AI to get inspiration for product purchases. Of that number, over half (52%) used it to research electronics and appliances. Other use cases include getting outfit recommendations (44%), creating meal plans (44%) and getting beauty tips (39%).
That sounds pretty much like tapping into more powerful search capabilities on the face of it. The more exciting potential of generative AI on the buy side doesn’t seem to be top of mind at present. Again, let’s revisit in a year.
Both reports are downloadable from the links above. Both are free of charge, but do require registration for access.
