Retail Pulse Report: Early Holiday Results, Consumer Strain, and Agentic AI
A quick peak at the news in the run up to NRF. And I repeat, if your fridge starts talking to you in Comic Sans, run.
Adobe Stock’s idea of what an “Agentic AI Refrigerator” would look like.
I am a bit punchy this week, as I’m writing this article on a plane on its way to NRF, after sitting through two flight delays at the airport. So if this article is a bit “extra” on the personality side, I promise I will dial that back to 70% next time.
In the meantime, enough things happened over the first two weeks of January that it was worth pulling it all together and taking a look, though obviously the most news is going to happen this week at NRF. However, not everyone in the retail industry makes it to NRF, so if you’re around to read this, let’s dive in!
Retail Economic Indicators
The reports are starting to roll in from Holiday 2024, and it looks pretty good, though we still have the “official” numbers to come for the US this week. In general, eCommerce was up shy of 9%, foot traffic was up in malls and stores, and overall consumer spending looks like it might beat inflation.
But because we can’t have nice things, those reports also came with some warnings. Consumers spent a lot on holidays because they used debt to finance a lot of it. Consumer confidence fell off considerably in December in both the UK and the US. Overall, the trend of 2024 continued all the way through to the end: consumers spent, but they show increasing strain in being able to sustain that spend over time.
What does that mean for 2025? I’ll give you a better take next week after NRF and the US December numbers.
Retail Tech & Research Data
Last week I talked about consumer trends that will drive 2025, and three of them combine into the drivers behind this next story. CouponFollow, a discounts platform, surveyed 1,000 Gen Z consumers and found that 3 out of 5 follow “underconsumption core” on TikTok, a trend that advocates for cutting spending as much as possible and only focusing on things that you really, really need.
The three consumer drivers that lead into this trend: Lonely (and feeling like following TikTok trends is a way to connect with others), Climate Demands (wanting to “do something” to address climate change, in the absence of cohesive government and business action), and Gaming the System (feeling like the system is rigged and all bets are off over using the tools of the system for personal benefit).
On average, survey respondents reported saving $250 per month, with about 25% saving more than $500 per month after following an underconsumption lifestyle. The leading areas for reducing spend were online shopping and fashion shopping for women, and cutting back on subscriptions and online shopping for men.
I think one definite retail need for 2025 is going to be figuring out how to make sure the answer to the “do I really need this” question is a resounding yes. The bad news is that consumers are definitely showing even more signs of strain than they did 6 months ago. The good news is that creating emotional resonance or a perception that the purchase contributes to a holistic sense of wellness can lead consumers to answer “yes” – retailers should not rely on price alone as a response. They need to bring together all potential selling points to build a total value proposition.
AI & Retail
The big story in AI in the last week was Nvidia’s Jensen Huang flanked by our coming robot overlords was that OpenAI’s CEO Sam Altman confessed that the company is losing money on the ChatGPT Pro plan, the one that costs $200/month. The reason why? “People are using it more than expected.”
We’re in the process of adopting Copilot at Aptos, and those colleagues who have been early adopters have been evangelizers – “I use it every day” and “Those long email chains? Just ask Copilot to summarize” have been phrases flying around the company internally. I agree – not having to parse through 30 email back-and-forths before someone escalated it to me just to get an understanding of what we’re arguing over is a blessing and I would use it every time. In terms of productivity, sure, yes, it easily saves me $200/month in time (Copilot doesn’t cost that much but let’s use this as the anchor price for now because there are pretty solid rumors that Microsoft is losing money too, at least on the Github Copilot).
I want to love Copilot. But I constantly worry that it will suddenly cost a lot more, and even if overall the company is feeling the productivity benefit, the initiative will run aground on pure cash outlay. There are signs that it may be possible to reduce the cost of developing models, but we’re also a ways away from figuring out where the unintended consequences are hiding – because you know there will be some of those somewhere.
OpenAI expects to lose $5 billion on revenue of $3.7 billion in 2024, by the way.
Oh yes, and in the same week that Meta announced it would be ending content fact-checking in favor of community notes, their “AI-generated profiles from hell” were also found by a reporter (they were first introduced a couple of years ago), which just served to highlight the worst parts of our potential AI future. I once read a scifi short story that had two people chatting online, matched by a dating app, but they were both extremely wary that the other person was a bot and they kept testing each other out to make sure, thus of course eventually proving they were both actually bots. Cory Doctorow’s principle of enshittification takes on a whole new dimension…
Retail Winners and Losers
Look, spend any time around me and you’ll figure out that I am a total science fiction nerd (see scifi short story reference above). I try to sneak in easter eggs where I can, but this next one requires an outright recommendation because I could not read this story without thinking about this book.
Samsung and Instacart are partnering to let consumers manage inventory and reorder products on their refrigerator screens. These are the Samsung smart refrigerators that have the cameras inside the fridge to be able to identify what you have and what you’re missing.
All we need is the refrigerator agent (the next generation of GenAI) that will combine with your meal-planning agent to both plan your meals for the next week and order via Instacart what you need (against what you already have), all on your behalf (with full access to your bank account, by the way). I’m not against this future at all, actually. But when the refrigerator starts talking to you in Comic Sans, I want you to remember that I told you so.
Back to the idea that retailers and brands have to have more than just low cost to win even cost-conscious consumers these days, Chipotle is expanding its partnership with Strava and taking it into multiple countries. I haven’t personally used Strava, so if I get this explanation wrong, apologies to the fans out there. Strava has local challenge segments, and members compete to complete the segment in the fastest time. If you complete the segment the most times in a period, you gain Local Legend status.
Chipotle is creating its own custom segments (I assume more than just the distance between Chipotle A and Chipotle B), and the Local Legend for each Chipotle segment in each city will win free Lifestyle Bowls for a year. Each city with a Chipotle segment will also compete to see who can collectively earn the highest mileage segment, and the winning city gets a BOGO entrée for the entire city. Chipotle selected the top 25 cities that eat Chipotle to compete.
This combines a sense of community, promoting wellness, and a nice little dose of city to city competition – a great way to get people thinking about the value of how Chipotle contributes to their life goals versus the price of the Lifestyle Bowl.
What Did We Learn This Week?
I got a little worried as we got later into the holidays that my post-Black Friday prediction that we were going to do better in holiday sales than predicted was going to miss the mark. There were some deep discounts on offer before Super Saturday (the Saturday before Christmas Eve). However, as the numbers are coming out, I think I did pretty well – the numbers look like they’re coming in at the top end of the predictions.
It can still come at a cost. Consumers are showing the strain. But from the pre-NRF activity we’re seeing, it seems retailers are done feeling the strain at last. They can’t wait for consumers to get their mojo back. They are on notice that consumers want to spend more time in stores, and they’re looking for ways to improve that in-store experience – and looking to ditch in-store tech investments that are getting very old after waiting through the pandemic.
The new year did not bring a new me when it comes to GenAI. I am cautiously optimistic but still mostly skeptical. My greatest fear is the cost. In stores, retailers get to the point where they say “It doesn’t matter what kind of ROI I can get, I just can’t afford the cash outlay” and GenAI could run into the same problem – it could deliver all kinds of productivity benefits, but if it drains the cash coffers dry, companies will balk. I don’t think we’ve figured out whether that will be an outcome yet.
And finally, when it comes to innovation, the things that will increasingly catch my eye are the things that will play to the consumer drivers I expect to see active in 2025 – unless something is starkly not one of those drivers, which will also be notable. This week, Chipotle provided a great example of how to appeal to many consumer drivers, and do so in a way that keeps the brand favorably in mind with consumers and out of the race to the bottom on price. Look for more of those as I find them!
And if your smart refrigerator starts talking to you in Comic Sans, run.