Last week, the retail economic indicators were cautiously optimistic, with decent growth predicted for US retail sales in 2024 and bright spots in UK consumer confidence in March. The industry learned through surveys that store employees want enabling technology and cite it as a reason to be loyal to a company, and that stores are better than online for both preventing returns (this isn’t a surprise but nice to see some data behind it) and for discovery, especially in fashion. A possibly AI-generated cookbook (and author) gives us a glimpse of just how junky a GenAI future could be. A few retailers had some technology missteps that hampered their ability to sell for a day or two. Some grocery customers are organizing to boycott unless their targeted retailer drops prices by 15%, and self-checkout gets more adjustments in high-theft stores. Let’s dive in to the latest in retail and technology.
Retail Economic Indicators
NRF published two big numbers in the last week. The first was a take on Easter spending. Spending is expected to reach the second highest level in NRF’s consumer intent survey history, surpassed only by last year. This number goes to show how challenging it can be to predict big spending events. Last year came in higher than this year’s prediction in part because last year had nine more shopping days leading up to Easter than this year. But regardless of time available, while the same percent of Americans plan to participate in Easter spending this year vs last year, the average spending per person is expected to come in lower – an average of $177.06 per person vs. $192.01 spend last year.
NRF also released their prediction for all of 2024 US retail – they are forecasting an increase of 2.5-3.5%, which would fall around or slightly above the Fed’s target inflation range. In 2023, retail sales grew 3.6%. Overall, it’s more of the same in terms of a middle-of-the road prediction that probably hides the potential for actual results all over the place, and definitely keeps it open for consumers to continue their unexpected resilience. Or not.
In the UK, inflation in February fell to 3.4%, which was welcome news. This was down from January’s 4% rate, and the lowest rate in 2.5 years. Food and restaurant drove the decline, with housing and fuel still relatively high. And while UK consumer confidence remained unchanged in March, flat at -21 according to GfK data, there were some positives in some of the survey numbers. Consumer expectations for the general economic situation over the next year rose one point to -23, which is 17 points better than March last year, and the personal finance forecast rose two points to +2, which is 23 points better than March last year and the first positive score since December 2021. Not just in the UK and not just related to consumer confidence, there appears to be some “blips” between February and March data – US March inflation data seemed to be a “blip”, but we’ll need to see another month to know if that’s really true or not.
Retail Tech & Research Data
This week saw a raft of research related to the impact of stores. First up, a survey by eMarketer, which found that physical stores have the most influence on discovery in fashion than any other channel. The numbers weren’t maybe as definitive as store operators would want them to be: “nearly” 40% of consumers say they learn about fashion by visiting a physical store. Brand and Retailer websites came in 2nd and 3rd on the list, also with responses in the 30%’s. However, store preference was higher for Gen Z than for other generations.
A new Scandit survey of 2,000 store associates across 9 retail verticals found that store employees want more technology available to them. 40% said their employer doesn’t invest in their technology needs and 20% say they feel like it’s “actively” not a priority. 41% say easy-to-use technology is a top-3 driver of their loyalty, behind work/life balance and competitive pay. It’s the “actively” not a priority that stings, but I can say I’ve seen it too – just a complete lack of belief that technology in store employees’ hands drives any kind of value, or serves only as a distraction.
And a new Consumer Returns Survey from the International Consortium of Shopping Centers (ICSC) found that shoppers return brick & mortar-purchased items at a far lower rate than online transactions. They found that online returns average 15.2%, which is three times the return rate for in-store purchases, at 5%. That was across industries. For apparel in particular, online returns hit 22% on average, while stores hit 6.2%. And 87% of consumers surveyed who report overbuying apparel online said they do so in order to try it on at home. For the basket of Aptos customers that I track, I can see very similar averages.
AI & Retail
I’m a big fan of The Future Normal, a newsletter run by Henry Coutinho-Mason, and last week he had a monster issue that covered both sides of the coin around the impact of AI. It could make the world worse, but it could also make it better. But there were two examples that he used that I thought were worth calling out. One, the example of IKEA retraining its 8500 customer service operators to be remote design consultants after replacing their old job with a chatbot. Through this upskilling, IKEA generated EU1.3 billion in sales in 2022. You can’t cut your way to growth. But you can invest to get there.
The second example was Kernel, a new effort by Chipotle founder Steve Ellis. Kernel is a new “robot-powered, partially automated restaurant concept.” Each outlet requires only three workers, but Ellis pointed out that it was also an opportunity to take a “McJob” and turn it into $27/hour with a paid vacation. An interesting way to look at how to invest those automation savings.
On the downside of AI, I could see this one coming a mile away, but it’s nice to have some data to back it up. It’s true that retailers have a lot of data. But AI’s hunger for data far, far outstrips retailers’ ability to supply it, not least because no matter how hard they try a lot of it is locked into siloes or not in any way prepared to be training data for an AI model. Salesforce and the Retail AI Council surveyed 1400 retailers and found that nearly half of respondents say they’re struggling to make their data accessible. Only 42% are connecting their various data siloes. Only 17% say they have a complete, single view of their customers. And 49% say they are still in the preliminary stages of building or even considering the creation of a complete customer profile.
It gets even worse – half of respondents say they lack the ability to fully comply with data security standards and data privacy regulations, but the adoption marches on regardless – 36% say they think their employees are already using GenAI today and that the number will grow to 45% by the end of 2025. The focus area? Personalization, with 93% saying they are using GenAI for some sort of personalization, like with email copy or product recommendations.
And if you want a taste of at least one possible future for AI, it’s already here, with this great story about a gift that turned into a dizzying journey into nonsensical cookbook and its possibly AI-generated author. Lawmakers want to do more to put a stop to this kind of fakery, and keep trying to find ways to make the marketplace – in this case Amazon – responsible. But where I keep getting hung up is around the money.
Marketplaces say “hey, look, buyer beware. I can’t be responsible for the millions of sellers and listings.” But they sure have found a way to monetize those millions of sellers and listings. And that’s where I feel like they should be vulnerable. To put it in real life terms, it may be true that a flea market, for example, can’t be held responsible for illegal activity that happens on its property during its operating hours. But shouldn’t they be held at least partly responsible if they’re letting in (and making money off of) sellers that are selling counterfeit goods? I’m not a lawyer, but that’s what I spend my time thinking about when I hear these kinds of stories – and I think we’ll hear more and more of these in the future.
Retail Winners and Losers
It was one thing to hear about Southwest’s extraordinary meltdown over the holidays when their scheduling software basically went belly up after years of underinvestment. It’s another thing to hear about it from retailers who are actually actively investing in tech. First McDonalds with its app going down across multiple continents, then Sainsbury having to cancel nearly a whole day’s worth of orders placed for home delivery. It just goes to show that creating a seamless experience is still not so easy, and requires a lot of coordination across a lot of different technologies and touch points.
Meantime, Loblaws shoppers have threatened to boycott in response to what they cite as high prices. It all started last November on – where else? – Reddit, with a subgroup called “Loblaws is out of control”. The group started out by venting, and then by organizing. More than 25,000 members post photos of items they claim are outrageously priced, and they’re targeting the whole set of Loblaws banners including Shoppers Drug Mart and T&T Supermarket for a boycott in May if the company doesn’t cut prices across the board by 15%, among other demands.
On one level, I feel like this isn’t about Loblaws in particular, just that these shoppers got together first. And of all the industries, it has been grocers stuck the most between shoppers feeling the pinch and CPG brands who have definitely raised prices beyond inflation – as Carrefour went so far as to ban selling Pepsi in January, citing basically “outrageous prices”. Will we see more consumer efforts like this? My bet is yes. Social media helps people who would normally never connect find each other, and it also naturally (or through “The Algorithm”) puts the more “engaging” content (read: rage inducing) front and center. So yes, we’ll see more of this.
In the great Self-Checkout (SCO) debate, Five Below and Dollar General both are changing up SCO to either reduce its availability, lower the limit on the number of items you can take through SCO, and/or provide more staff assistance in SCO lanes. Target made changes too, among many others on the list trying to shore up loss prevention in their highest-theft stores.
And finally, several more examples of retailers who are no longer just in retail. Walmart is going to sell its AI-based logistics solution (called “Route Optimization”) to third party businesses of all sizes, all through its Walmart Commerce Technologies subsidiary. Marks & Spencer is reportedly nearing a deal with HSBC, the provider of M&S’s banking services, to provide a “superapp” that will combine banking, payment, and the retailer’s loyalty program all in one.
Shein is going to sell its small batch manufacturing model to anyone who wants to use the capacity. They’re billing it as “supply chain as a service” but I feel like that’s a bit misleading – it’s the quick, small batch capacity that is valuable here, not necessarily how it gets from China to anywhere else. And lest you forget about Amazon, the ultimate in “not really a retailer”, it announced that its Zoox subsidiary – the company’s autonomous vehicle subsidiary – is expanding the scope of its pilots in Foster City, CA and Las Vegas, NV. The pilots are for robotaxis. At the moment.
The Bottom Line
It’s gratifying to see something that you believe is true get supported by actual research, even if you have to be careful to make sure you understand what the research really means (and be careful of my own confirmation bias). But the benefits and ROI for eCommerce are easy to see, and in this muddled omnichannel world, I think it’s harder to see the value contribution of stores – and the squeaky wheel is the one that gets the oil, more often than not.
It’s not gratifying to see retailers already starting to run into a data wall when it comes to their aspirations for GenAI, but it’s not surprising either. I think this will be a theme through all of 2024 – “that proof of concept went great but that’s as far as we can go until we do a lot of work on making our data more organized and accessible.” But – as the strange cookbook experience shows – we are still in the early days of what GenAI can do, both all the good and all the bad. And, as McDonalds and Sainsbury show, even the more basic tech journeys are not easy. However, consumers don’t care. They have their expectations, and we have to meet them.