The dust has yet to settle from Walmart’s recent announcement that the retailer will soon make its products available for Instant Checkout via ChatGPT.  Since then, OpenAI has announced a flurry of additional releases – from a new ChatGPT-embedded web browser to a partnership with Paypal, to an entire third party app ecosystem within ChatGPT.  We’re witnessing the beginnings of a tectonic shift in consumer behavior and the ecommerce landscape.

‍

The scale of AI shopping

Much has been made of OpenAI’s 800 million global weekly active users (77 million in the U.S.).  But how many of these shoppers are turning to answer engines like ChatGPT to ask shopping related questions?

Over the past year, Stackline has started tracking “shopping question volume” – any question with retail purchase intent – across the largest AI answer engines (ChatGPT, Perplexity, Amazon’s Rufus, and more).  In the U.S., shoppers are now asking ChatGPT over 84 million shopping-related questions on a weekly basis.  And the volume shows a marked acceleration since the beginning of summer.  

‍

Retail search is dead.  Long live retail search.

While the growth in ChatGPT’s shopping question volume points to continued disruption in path-to-purchase, it’s important to contextualize that volume against retail search (think: the keywords that shoppers type into a retail search bar).  On Amazon in the U.S., Stackline data shows that shoppers conduct an average of just over 1 billion searches each week. That puts Amazon’s search volume in the U.S. at 12x the shopping question volume that ChatGPT sees on a weekly basis. But looked at from another angle, ChatGPT has grown from less than 1% of Amazon’s weekly search volume to over 8% - all in less than a year’s time.  We’re poised to see ChatGPT capture more and more shopping journey share, especially as the platform rolls out Instant Checkout with Shopify, Walmart, and other brand partners.

‍

Consumer use cases for AI shopping

Stackline recently conducted a large-scale survey (n=2,500) of U.S. consumers to understand shoppers’ mindsets and use cases around AI-assisted shopping journeys. 41% of respondents indicated they had recently used AI to help with shopping (up 250% from our last benchmark just a month prior).  This rapid increase highlights the impact of OpenAI’s recent press and partnership announcements, as the platform gains more mindshare and promises more utility.  

When we asked consumers how they are using AI to help with shopping, use cases hit nearly every stage of the funnel.  The top 5 uses cases, however, were largely focused on mid-to-lower funnel use cases around consideration and conversion:

Of shoppers already using AI shopping assistants, frequency of usage is key.  Nearly three-quarters report that they’re now using AI to help with shopping on a monthly basis.  Nearly one-quarter (23%) do so at least once a week.  When we asked these shoppers how they see their AI-assisted shopping occasions changing over the next 6-months, nearly half (43%) said they’ll increase their usage.  Just 11% said they’ll decrease usage – pointing to a high level of utility that shoppers are finding with these AI assisted experiences.

‍

The evolution of the digital shelf

All of this has massive implications for how consumers enter categories, discover brands, and compare product offerings.  For the past decade, retail search has provided one of the best windows into the mind of the consumer – as each shopper approaches the digital “aisle” through keyword-based search.  Measurement has largely followed suit, with brands using share of retail search as a leading indicator of both ecommerce performance and market share.  

As Stackline retail search data shows, that mode of shopping isn’t going away (at least anytime soon).  But it’s no longer a given that retail search will continue to drive the big year-over-year volume increases it has seen in years past. Take the electronics department as an example – which is arguably among the most digitally penetrated and mature in ecommerce. For the first time in history, annual search volume for electronics products across Amazon, Walmart, Target, and Best Buy fell over 5% year-over-year.  It’s not that shoppers aren’t still looking for headphones or televisions or tablets – they’re just starting to research and discover those products elsewhere, such as ChatGPT.

‍“AI visibility” is the new share of shelf

Now more than ever, there’s a need for a new set of data and KPIs to understand these shifts in consumer behavior, and how brands show up:

• Question Graph: What are the questions that shoppers are asking AI agents like Amazon Rufus, ChatGPT, Perplexity, or others?
• Question Volume: What is the volume behind these questions? Which questions are must-win today, and which are accelerating?
• Answer Impressions: How many impressions is your product or brand receiving when featured in the answer or recommendation from ChatGPT or Amazon Rufus?
• Answer Impression Share: What is a brand’s share of all recommendations and answer impressions in your category? How does that vary by question type or category? Where are the gaps?
• Citations: What are the sources these AI agents are using to generate their recommendations?

Now’s the time to get ahead of the curve.  Let’s set up time to discuss how your brand can leverage these shifts in consumer behavior – and how Stackline’s rapidly evolving suite of AI measurement and activation capabilities can help provide you with competitive advantage.