Retail Pulse Report: The Consumer Waiting Game and CFOs Love GenAI – When They Don’t Have to Pay for It
Bonus question: are we still feeling the impacts of the pandemic now that we’re five years in? Or is this the “new normal”?
Adobe Stock: The consumer, still waiting for things to get better.
Lots of people have remembered that it has been five years since everything closed down for the pandemic, and five being halfway into a decade, it has been cause to take stock on where we are, so lots of coverage came out last week on this topic with more to come this week.
So fair enough. Where are we? Not back to “normal,” whatever that means. Consumers are still as wary as ever. GenAI is something to contend with that almost no one was talking about in 2020 – and it is driving as much bad as good and future impacts are still unknown. One surprising idea: we are still feeling the ripple effects of the pandemic, and not the last little waves of it either.
Let’s dive in!
Retail Economic Indicators
Our global view of economic indicators will include stops in the US, the UK, and Germany. First up, in the US, the jobs report came in for February, with nonfarm payrolls increasing 151,000. This was less than the consensus forecast.
One aside here: the more I look at these numbers and how the press reacts to them, the more I am skeptical and careful about the tone to take. A bunch of economists and investment analysts made some guesses about how much payroll will change, and the reaction is less about the actual number reported, and more about how well the analysts guessed. The number is the number. If there is a big gap between expected vs actual, I would think it would be more interesting to understand WHY there was a difference, rather than clutch pearls over the fact that something that largely cannot be predicted didn’t turn out as predicted!
Anyway, I don’t know why economists thought payrolls would be higher, but I do know that this report barely starts to bake in DOGE and its layoffs. And this part is interesting: average hourly earnings climbed 0.3%, and the annual increase was 4%. All things considered, wage growth and increasing payrolls indicate that the job market is still doing better than anyone expected (except the nameless “analysts” or “forecasters” cited against the numbers). That could definitely take a nosedive. Scenes of angry voters yelling at their representative aside, I don’t think anyone has a great handle on how much consumer and labor market pain is being inflicted as I type this.
In the UK, some new shop price inflation numbers landed. Inflation rose 0.7% year over year in February, which was the same result as January. Where inflation in January was blamed on travel staying expensive coming out of the holidays, in February it was due to food prices, which were up 2.1% year over year. Breakfast staples (yes, eggs, though much less so than in the US), but with the threat of more to come due to supply chain disruption for coffee in particular, as well as upcoming levies expected on packaging. Hidden in the average was non-food price deflation of 2.1% year over year in February, which was up over the 1.8% deflation figure from January.
Last stop, Germany. The HDE consumption barometer for March (HDE is the German Trade Association) was unchanged from February. I can’t read German, so I am dependent on Google Translate, but they used the word “stagnant” a lot, which perhaps has a more negative connotation than they meant? Because while it’s true the overall sentiment rating was unchanged, their intent to consume was slightly higher compared to last month, and while their tendency to save remained the same, their future income expectations were higher. At this point consumers seem to be holding their breath until the new government forms.
We’re on the cusp, is how all this feels to me. It could be the cusp of recovery or the elusive “return to normal” that everyone has been asking about now that we’re officially hitting up against 5 years since the pandemic. It could also be on the cusp of economic disaster as on-again / off-again trade wars and general global bullying and strife continue to play out in unexpected ways.
Retail Tech & Research Data
Two surveys this week, both related to fraud and theft.
SEON surveyed business leaders (not retail business leaders specifically), and found that 85% are increasing fraud budgets and 88% their fraud teams. 86% say they spend over 3% of revenue on fraud prevention alone – I hope that’s on top of the regular IT budget, because that would eat way over half of a retailer’s IT budget. 65% of business leaders surveyed said they will slightly increase their budget in the next 12 months, and 20% said they will significantly increase. And 76% reported in-demand fraud preventions skill gaps are mostly related to AI and ML expertise.
At the same time, Flock Safety and Zencity conducted a survey of consumers and found that one-third said they saw retail crime as a major concern and 58% said it has made them want to shop online instead of stores. What was interesting to me is that 63% of those who said they had actually witnessed retail crime said they preferred online shopping as a result (vs the 58% overall). In consumer survey margins of error, that 63% vs. 58% difference is not as significant as it sounds. But it should be worrying to retailers that consumers see retail theft as a deterrent to shopping in stores just as retailers are trying to embrace having consumers shop there over online (as that channel has become less profitable).
There’s an angle here that retailers do not always take into account: yes, consumers don’t like it when you lock up things they want to buy. But if they think your store isn’t safe, then they’re not going to show up at all.
And also, theft and fraud get more sophisticated every day, aided by AI of all kinds (voice and video are improving by leaps and bounds almost daily). The pace at which theft adapts and evolves has never been so rapid, and I’ve heard fascinating ones that exploit everything from offline mode for a pinpad to timing gaps between when a gift card is activated and then deactivated (“Oh never mind, I don’t want that”) versus how quickly a thief can redeem the card during that brief active period.
Deterrence is cheaper than enforcement, but I’ve seen a lot of retailers allocate budget for theft based more on high theft rates occurring than on keeping it from getting high in the first place. And the side effect may be driving more customers from your stores than you realize.
AI & Retail
When I said that video and voice AI are evolving by leaps and bounds, I meant it. And Google and Amazon are going head to head to prove it, with both of them introducing all kinds of visual search tools. For the longest time, consumers called Amazon the “Google of product search” and both have been competitive about owning product search for a long time (Google has had many attempts at upping their game).
Well, now TikTok is the “Amazon of product search” and both Google and Amazon are raising their bar in what looks like head to head competition, but is more likely trying to win back product searchers from TikTok.
On the Google side (the company has already launched a whole range of visual tools, including virtual try on based on a model you fine-tune based on your characteristics), they have launched “vision match”. This sounds pretty cool, and it’s on my list to try: you describe a garment you have in mind and it uses GenAI to create some images of what that could look like (which you could then adjust to get more like what you want) – but then throws in some real products that are similar that you could buy.
They have also expanded their virtual try-on to makeup using Augmented Reality (AR), and you can try on looks inspired by celebrities and influencers. This used to be really hard. I remember consulting for a company that was working in this area in the early days – long before Sephora and L’Oréal introduced smart mirrors in stores. I feel like I should volunteer as a test case too, because the color in the package NEVER looks like the color on my skin, and I remain skeptical that these tools will be useful for me. So if Google is successful with this one, that would be a real, actual thing.
Amazon is catching up, and has launched a laundry list of new visual search features. I can’t even do them all justice, just read the article, but one that I really liked was adding text to an image search, so that you could, for example, upload a picture of a beige sofa and say “find me one like this but in blue”.
While I’m sure many of these were in progress long before consumers flocked to ChatGPT and Perplexity to help them find Christmas presents last year, I am also sure that many of these have been accelerated in order to stave off encroachment from GenAI product search at the same time that TikTok has already encroached in the visual space.
Last week I said that everyone is waiting for GenAI to “show them the money” and that it could end up becoming the free shipping of productivity. So I felt obliged this week to include this survey of CFO’s on the ROI of GenAI. Apparently, 90% of CFO’s surveyed report a “very positive” ROI from the tech, a big increase from the 26% who said so in March of 2024. Cybersecurity saw the highest adoption, with 75% saying they have increased the use of GenAI to strengthen their cybersecurity efforts. Overall, the main sources of ROI have been detecting fraudulent behavior, customer service chatbots, and content generation.
Some scary reality checks: 97% of CFO’s surveyed expressed “high” or “complete” trust in GenAI’s output for risk management and 98% said the same for strategic decision-making. Yikes. Not to worry, though, 30% worry that GenAI might produce less insightful results over time, and 28% are worried about data security. 75% agreed that GenAI content generation needs human oversight, while 69% agreed the same about product innovation. This amount of naivete is a bit breathtaking – have they exported all their concerns to their CIO/CISO?
Let’s also remember that they’re not paying what it actually costs in those ROI calculations of theirs – not even close.
All that said, I – you – we cannot dismiss that we’re still only at the very first part of the upward curve of the inflection point that is GenAI’s disruption. And disruption encompasses business models, consumer behavior, the nature of employment, and organizations. As cautious as I am about embracing these things as whole-heartedly as the CFO’s in the surveys above, I still keep an eye out for big picture thinking about the impact in the future. Found one in Neil Perkin’s position on what “the agentic organization” could look like.
I’m still absorbing it myself, and there are some good analysis points and frameworks for thinking about things. But a lot of the journey that he throws out there is basically cataloguing every workflow and process in the company, and then classifying them along a range of capable of being fully autonomous to fully human-enabled, with some mix of human dependency in the middle.
I know of only very few companies who could identify, let alone then analyze the potential of, the majority of processes they have. And retail is far worse than other verticals. Already you have the case of having very little understanding of what happens in a store – foot traffic goes in, sales come out. Who knows – consistently across all stores – what actually happens between those two things? No one. So, it could be interesting to, say, record all conversations or have body cams on all employees to understand what they do all day. It could also be one, incredibly expensive, and two, incredibly creepy. For both the store associate and the customer.
Same for district managers. They visit stores, they fill out forms and paperwork. They give advice to store managers and serve as the front line for rolling out new initiatives. How they do that, though, that’s a whole business case in itself, finding ways to tease out the special sauce of what makes this district manager more effective than that one – and it’s a team sport, so that a district manager who is effective with this set of stores may not be effective with another set of stores.
Yes, GenAI could potentially suss out the patterns - if it has a large enough data set about what’s actually happening.
Which leads me to one last comment about agentic organizations. One use case is that of digital twins – that you could model your organization and its competitive and economic environment to a degree that you could use it to simulate future actions. We’ve been talking about digital twins in supply chain for a long time now and still aren’t there. Still aren’t even at “good enough” to take any prediction as at least a directional indicator. I can only imagine all the broad swaths of assumptions that we will make or miss in the attempt at taking digital twins not just to an enterprise level but to a whole market level.
If all this really comes to pass, this is decades worth of work to get there.
Retail Winners and Losers
I’ve gone on long enough already, so I’ll get through the last two sections quickly. As I noted above, it has been five years since the pandemic. Cara Salpini of Retail Dive put together five charts to illustrate where we are vs. where we were. Some interesting conclusions:
Ecommerce may well plateau at around 35% of total retail sales. (I have a lot more I could say about this one!)
Store closures are going to be much higher in 2025 than they were in 2020 – propped up by stimulus money, the closings are a delayed pandemic impact.
Big ticket items may have finally bottomed out from pandemic stocking up (but high interest rates would have to come down to get to recovery mode)
And in 2024, retail hiring was 40% less than it was in 2019.
Two take-aways for me: while everyone, including me, is ready to move on from the pandemic, we are still feeling its effects today – in tough times for big ticket items, in the slower-burn impact of store closings, even in the job prospects in retail. There is no “going back to normal”. But also, it may not be true that where we are today is “the new normal” either.
Emarketer summed up a trends webinar from Placer.ai, and one of the topics struck me. It’s about designing stores for experiences vs convenience. That during the pandemic consumers just wanted convenience, because they wanted to be in and out and back home again. But now we’re entering an era where they want experience, even third spaces to come back to give them a place to hang out that is not home or work. The webinar did not appear to come to a conclusion, but the idea that experience and convenience are somehow naturally mutually exclusive and that designing for one means giving up on the other – I’m not sure that’s true. No shopper comes to a store to have the same exact customer journey as everyone else. So to say that a store cannot enable different objectives sought by different shoppers at the same time – I think we’re underestimating stores.
And last, poor Target. After every known ability to measure store and web traffic found that Target actually did take a blow by the one-day consumer-organized boycott, a 40-day consumer strike is now planned. Which customer base are you going to keep, Target? The one on the left or the one on the right? Because hopefully you learned from Budweiser that if you don’t take a stand, and quickly, you’ll lose both.
Store Innovations
I debated about including this store innovation, because it’s not really that innovative – it’s been done before, like when Macy’s teamed up with Facebook to bring new product discovery to select stores. In this case, Primark has teamed up with Pinterest for a new “trends-led” homeware range. There are a few things that are interesting and different, though, worth pointing out.
This picks up on 3 trends that Pinterest identified from user behavior and searches – Cosy Cottage, Minamaluxe, and Dainty Décor (I can already tell I’m not going to be in the target market for any of this). It rolls out to 100 Primark stores but also has a stand-alone store opening in Belfast. And it’s tied in with Pinterest’s new Collages tool that allows users to curate mood boards (with Primark products, of course). But the most innovative part is the masterclass on how to use Pinterest Collages, being offered in real life in four flagship locations throughout the UK. Sounds like an organic social media bonanza to me!
What Did We Learn This Week?
Whew! Lots to cover this week:
Consumers are still waiting – they want to be optimistic but not willing to commit to it.
Retailers have to think about how theft impacts consumer willingness to come to stores, not just the bottom line of reducing theft.
Fraud is getting more sophisticated and retailers have to up their game to keep up
CFO’s love the return on investment they’re getting for AI – because they’re not actually paying what it costs (and they seem way too trusting about it).
It’s worth thinking about what an “agentic organization” looks like, but don’t get too carried away – retail has lots of dark corners of processes that no one really understands, let alone tracks.
And two questions that I’ll pin for now and come back to at some point:
Will eCommerce sales top out at 35% of total retail sales?
Do stores have to sacrifice experience for convenience? Or vice versa? Is it impossible for stores to serve both needs at the same time?
Until next week!
- Nikki