From the perspective of retail economic indicators, it’s a mess out there, more so than at any other time since I’ve been covering economic indicators impacting retail. Here are some of the frustrating and contradictory headlines that were published just last week:
- “US consumer confidence rose in May” (US)
- “April trade data reflects unrelenting pressure on retailers” (AUS)
- “Retail sales recover as inflation slows” (UK)
- “UK retail footfall sees 3.6% drop in May 2024 (UK)
- “Euro zone inflation rises to 2.6% in May, but bloc still seen heading for interest rate cut”
- “India’s inflation tug-of-war: Falling headline rate versus rising food prices
And this one seems to sum it all up in one: “US economic growth last quarter is revised down from 1.6% rate to 1.3% but consumers kept spending.” In other words, depending on which eye you’re squinting with, it’s either getting better or it’s getting worse.
But that’s not the only news in retail this week – let’s dive in.
Retail Economic Indicators
As promised, the two-sided story of consumer health continued. On the bad news side, Power Retail reported that April sales in Australia grew just 1.3% year over year, well below inflation, and taking a particular hit in discretionary categories. The CEO of the Australian Retailers Association (ARA) pointed to spending levels as a possible indicator of a looming recession. Over in the UK, retail footfall fell 3.6% in May, based on Sensormatic IQ and BRC data. That’s an improvement over the 7.2% drop in April, but there is still the caveat of trying to compare year over year against an early Easter.
In the “I still don’t understand why” category, India’s inflation is falling (and it has been below the central bank’s target benchmark rate for a while now), but food prices have remained stubbornly impacted by inflation, sitting at 7% inflation rate for the period between November 2023 to April 2024. In the US, GDP was revised downward last quarter to 1.3% from 1.6%, but that didn’t really stop consumers from spending. And Eurozone inflation rose to 2.6% in May (keep in mind, from 2.4% in April), but analysts still expect an interest rate cut from the ECB in the next week or so.
And then the good news: US consumer confidence rose in May – sooner, actually, than I expected – but improved to 102.0 from 97.5 in April. And it came after 3 straight months of decline. Fewer respondents said jobs were hard to get (which may be delayed from reality, if the job market reports are to be believed), and fewer said they expect deterioration of future business conditions, job availability, and income. However, high prices are still the largest contributor to the negative side of their view of the US economy. Also good news, retail sales in the UK grew at their fastest pace since December 2022, rising 8% in May according to the CBI.
As I said, if you have an angle you want to promote here, just pick which half of the situation you want to focus on, and you’ll find a headline that will support your position.
Retail Tech & Research Data
One particular angle that retailers need to pay attention to is not just consumer weariness over high prices, but actual consumer backlash. And sometimes that means paying attention, once again, to the difference between what a consumer thinks they would do vs. what they actually do in the moment.
Case in point, the Canadian boycott of Loblaws. According to a survey, 70% of Canadians were aware of the boycott, which was planned for May. It started in a Reddit group, which blamed the grocer for its high prices. However, only 18% of those surveyed said they actively participated in the boycott, a big gap from the 58% who said they supported the boycott. 65% said they didn’t feel like it would make any actual difference in lowering prices.
One interesting difference between Canadians surveyed vs. Americans: 76% of Canadians living in Atlantic provinces said they believed the price of at-home food is worse now than it was 2 years ago, which most US consumer surveys also say. The difference is in who is to blame. For Canadians, 29% blamed grocery stores – and just shy of 50% of those asked about Loblaws said that it wasn’t fair to target Loblaw’s and not any other grocer. But Americans were more likely to blame the government – they blamed government policies as the primary reason for high prices, followed by product manufacturers and then suppliers.
The Reddit group promised to play on, by the way, continuing the boycott into June.
Confirming at least the potential for danger, an Axios Harris poll exploring retailer and brand reputation found that many reputation scores fell this year, in a survey of 16,500 US consumers conducted in March. Impacted brands included Amazon, Costco, Kroger, P&G and KraftHeinz. Trust fell the most for clothing companies, followed by quick serve restaurants and then big box stores, and consumers cited high prices as the reason why. And which is also probably part and parcel behind why retailers are starting to go big with announced price cuts (see more on that below).
In a random collection of interesting studies, a report by LexisNexis Risk Solutions found a 19% increase in global human-initiated digital attacks in 2023 vs. 2022. This is based on 92 billion transactions processed through the LexisNexis Digital Identity Network in 2023. The leading type of fraud in the human-initiated category was third-party account takeovers. The report found that automated bot attack rates did not increase in 2023, which it attributed to stronger bot detection. Turn on two-factor authentication, people! And don’t ignore the random text giving you a code when you didn’t ask for one – that means it’s definitely time to change the password.
And a study by Team Global Express found a close link between the delivery experience and customer retention in Australian shoppers. The study metrics aren’t great – 750 online shoppers – and it’s definitely there to drive retailer interest in the company, and, if that weren’t enough caveats, it’s kind of a “dog bites man” story, but one callout worth making is that consumers value different things at different parts of the sales journey, for example they place a higher value on information about clear delivery fees at checkout, vs on package arrival, they want clear information about what to do if the package is damaged. In another case of “what I say is not what I do”, 91% of consumers surveyed said they would abandon your brand if their package is damaged, which of course is not reflected by reality.
Finally, in a story to make children everywhere shudder, consumers have apparently already begun back to school shopping. According to a survey by LTK, 14% of shoppers surveyed said they started their back to school shopping in May. 19% say they will begin in June, and half will have already begun shopping by the time we hit mid-July. It’s not a stretch to think that consumers are looking for big discounts on the kinds of school supplies that they can predict will be on the supply list come fall. But there will always be that pressure to buy back to school clothing as late as possible.
AI & Retail
On the positive side of AI, WWD provided a pretty comprehensive look at where AI is getting adopted in the fashion vertical, with a deeper focus on shopping assistant bots. This is in line with what I’m seeing as well, where the first wave of GenAI adoption in retail focuses on an area where “popular” is more valuable than “precise” – a difference where GenAI can make mistakes and not even have it noticed. If a GenAI bot makes a mistake in what it recommends, then you risk not making a sale, but not having the bot in first place also means you might not have made a sale. Whereas having a GenAI bot at least increases the opportunity to influence customer decisions (vs nothing at all). The article indicates that global virtual shopping assistant market will reach $6.9 billion by 2032, from $516 million in 2022. And it also points out that this is really the first generation of such shopping bots. Multi-modal bots, the kind that can “see” and “hear” and communicate back in much more than just text or just images, aren’t even fully on the scene yet, and promise to do even more to change the game.
The challenge, though, always comes back to fit. And it may be that shopping bots are what really bring adoption to fit solutions, which have had some success but have still not become ubiquitous. Fit applications have two-sided market problems, where they have to get enough brand manufacturers to provide garment data, and get enough consumers to provide fit profiles, to really take off, but TrueFit’s recent tie-in with Shopify as an offering to sellers might provide that tipping point.
On the other end of the AI spectrum is AI for the sake of trying to cash in on the mania, and I would definitely put Eddie Bauer’s smart sunglasses into that category. The manufacturer, Innovative Eyewear, is delivering access to ChatGPT, using Bluetooth. What is the consumer need here, exactly?
And finally, two stories that relate to how GenAI actually works, and that reemphasizes that there is still a lot about GenAI that we don’t actually know. The first story covers research that explores the challenge of what happens when, if you ask the same question two different ways, you may get two different answers from the same model, even when a human would understand and accommodate the differences to get to the same answer. The difference is if you ask an open-ended question versus a more discriminative question, one that requires precision in the answer. As people in the industry increasingly talk about, GenAI is good at the open-ended questions (as long as your sources don’t include sub-reddits that joke about glue and pizza), but it’s really bad (or even more bad, I guess) at answering questions where there really is only one right answer.
So some researchers developed a consensus model, one that pits the open-ended answer against the precision answer, to make sure that both models agree on the right answer before it is provided to the user. If you throw in some game theory in how those answers are pitted against each other, the researchers found that they could improve the accuracy and internal consistency of the answers provided to the user. Back to the story above about multi-modal bots, this kind of intermediary AI may ultimately be what’s required to get to AGI (artificial general intelligence). It’s not going to be one model that can do anything a human can do. It’s going to be one interface and intermediary between multiple different types of models to drive some kind of alignment and coherence between them to get an answer that they all agree on. I guess this is like arguing with yourself in your head!
In another example of having to get creative just to be able to get to the bottom of how GenAI works, Anthropic published a paper that explained how they sub-divided their model to get a better idea of how it works. This is basically like mapping a brain while it’s being used – just like you have regions of the brand that you rely on for certain things but not others, models apparently have pieced-together capabilities that approximate the same thing.
Other than, this stuff is kind of cool in and of itself, there is another lesson to take away from these developments: don’t just buy a chatbot and consider it “job done”. The industry is changing too fast, and today’s chatbot will be obsolete before the holiday season kicks off.
Retail Winners and Losers
Let’s start with the positive news: retailers are finally starting to cut prices. This week it was Giant Eagle (though they’re doing it as an 8-week promotion across a targeted set of 1,000 items), and Co-op cutting prices on fruit and vegetables. This is after announcements last week by Walgreens, Walmart, Aldi, Target, and Amazon Fresh. When this happened in the UK, it took a few months before it showed up in inflation numbers – a distinct slowing of inflation and in some cases even deflation.
Speaking of Amazon Fresh, the company continues to tighten its focus, ending grocery delivery in 5 UK cities. Beginning June, the delivery service ends for Glasgow, Portsmouth, Sheffield, Leeds, and Newcastle. The company points out that you can still order at least some products for delivery through its partnerships with Co-op, Morrisons, and Iceland.
And while fresh delivery may have its setbacks, fashion rental is looking at a possible renewal. Nuuly, an apparel rental business under the Urban Outfitter umbrella, reported more than 50% annual sales growth for the quarter ending April 30. This follows full-year growth of 172% in 2023 and 96% in 2022. And Rent the Runway beat analyst expectations in its latest earnings call, saying the company expects a break-even year, which would be a big milestone for the company, according to the CEO.
Finally, only because I couldn’t bring myself to put it in the store innovation section, L’Occitane has launched a new The L’Occitane Greenhouse virtual experience. The experience is a 3-room virtual world focused on a women’s collection, a men’s collection, and a Middle East Oasis that somehow features local scents. The objective is to “tell a story that evokes curiosity with online visitors, to increase brand loyalty and grow in-store foot traffic.” Of note, Emporia is the tech provider – the same one powering Walmart Realms.
Store Innovations
Back in the IRL world, Meijer has joined Walmart in trying to discover the right format and assortment for a smaller grocery store. They have opened their third now, which can be anywhere from 1/2 to 1/3 smaller than a typical Meijer store. And Amazon got approval from the FAA to operate drones beyond line of sight, a key step forward in their plan to expand drone delivery to over 500 million packages delivered by 2030. The approval applies to College Station, TX only – so far.
And last but not least, Hallmark is going to open an immersive “Christmas Experience” at the Crown Center in Kansas City, MO. The intent is to bring a Hallmark movie town square at Christmas to life. It will run from November 29 through December 23. It will have a Christmas Market that features Hallmark products as well as unique artisan-crafted items, food, drink, ice skating, a Crayola community mural, interactive light walks, and a nightly tree lighting. It will also include ticketed events and activities like photo ops with Hallmark stars, workshops with Hallmark artists, and breakfast with Santa. I feel like I’ve eaten an entire plate of Christmas cookies and binged watched twelve hours of Hallmark movies just typing all of that out. Hallmark is an interesting blend of retail and media in general, but just because this is based on the infamous movies doesn’t mean that it doesn’t set the bar for consumer expectations from retailers, or even from Hallmark stores, a lot of which are franchises.
The Bottom Line
What can we learn this week? Consumers are, indeed, a mess. And whether it’s GDP, the labor market, inflation, or consumer confidence, it’s clear these things don’t move all together and the lag between some of these indicators creates its own interplay as well. And in the meantime, the pace of change in AI – not just in the capabilities available but in our understanding of how these capabilities actually work – is moving so fast, no one should be complacent with what they’ve implemented already. Foundational models are improving at a crazy pace. Is your bot built on a foundational model keeping up?
And two competing trends in store innovations are keeping things interesting – grocery retailers trying to figure out what it takes to operate a smaller format store (critical for their expansion plans as consumers are increasingly living and working in the suburbs), and multi-industry brands like Hallmark trying to figure out how to create synergy between its multiple content channels, raising the bar on what’s required to deliver “experiential” retail. The question is, is it scalable?
Read the full article here