While I love New York, January is really not the ideal time to visit the Big Apple. However, it is an ideal time for retailers to gather themselves up, dust themselves off from the tail end of the holiday season, and figure out what they need to do next.
To be fair, most retailers have 2-3 years’ worth of “what to do next” piled up, especially when it comes to technology. But something about coming fresh off the holiday season, whether successful or not, seems to really sharpen the focus. This year, holiday results seemed uneven. While overall the holiday season was better than expected, that success was not evenly distributed – there were some definite winners and losers. But whether retailers won or lost 2018, they came to NRF equally determined to tackle their technology to-do lists.
Exhibitors brought robots and drones (flying robots?), artificial intelligence, augmented reality, personalization, and more. But after the dazzle was done and the booths packed up, what stuck on retailers’ to-do lists? Based on what I saw and the discussions I had, here is my take on the top four tech topics we’ll still be talking about by the end of the year.
While it may seem so 2010, cloud is still a big topic in retail. Tech-driven companies like Netflix and Amazon have long embraced cloud technologies and architectures. They’ve spent at least the last decade deploying cloud-native software that was designed to take advantage of the promised benefits: scalability, flexibility, speed to market, the ability to innovate.
There are a lot of reasons why cloud is “better” for everyone – for retailers and their solution provider partners both. For solution providers, it means keeping everyone up to date on the same version, easier deployments, faster cycles of updates. For retailers, it means never getting version-locked or so far behind as to be off maintenance. It should also mean faster access to innovation and more flexibility in keeping up with consumers’ shifting behaviors.
You would think, based on the way vendors talk about cloud, that it’s a foregone conclusion that the future of software delivery is cloud. However, there are still a lot of questions about cloud that are stopping some retailers from doing more than hosting (outsourcing the computing power and operating stack, while still running and paying for a license application). The biggest stopper seems to come from CFO’s, who see more and more solutions coming in as operating expense subscriptions, which are fine when they make up a few parts of the technology portfolio, but start to look pretty intimidating as more and more of the portfolio moves over to that subscription model.
However, part of the challenge comes from a belief that software doesn’t need to change or really be upgraded that frequently. There are lots of retailers operating on-premise software that is ten years older or more, and have convinced themselves that if it worked ten years ago, it should work just fine today. If that’s the going-in belief, making arguments about speed to market and the flexibility to innovate are going to fall on deaf ears.
Retailers seem to now fall into three distinct buckets when it comes to cloud: those who have fully embraced it and are looking for business benefits akin to those achieved by Amazon or Netflix, those who flat-out refuse to believe that cloud is anything more than a scam concocted by software providers, and those who understand the benefits of hosting, and feel the pressure from the pace of change in retail today, but have not yet fully embraced cloud.
So, thought leaders in the industry may have moved on from cloud, as if it’s a done deal, but I would say it’s not so done as it seems. And vendors and retailers both need to make sure they understand – and place a real value on – the ability to keep up with the speed of the consumer. Which means we’ll still be talking about cloud and what it really means and what the value really is for the next year. But hopefully not too much more beyond that, as being able to match the speed of the consumer may easily become a key differentiator in 2019.
Ironically, a lot of the retailers who are cloud-skeptics are also big fans of Artificial Intelligence or Machine Learning technologies – which, when you think about it, can really only be effectively delivered via the cloud. AI platforms, the algorithms, the data to train those algorithms, plus the access to the computing power necessary to spin up AI results are all best delivered through cloud, otherwise you have to spend an enormous amount on infrastructure in order to support an AI capability, not to mention the (expensive, hard to find) people who maintain AI algorithms.
Some of the love for AI comes from a fascination with “shiny objects” – if AI is the next big thing, then it must be the thing that solves all a retailer’s problems. However, a lot of the talk about AI is just that. First, you have to separate different kinds of AI from each other. There is a lot of activity going on within natural language processing with chatbots and AI’s that can write product description copy, for example. And there’s a lot going on with image processing – recognizing the difference between a dress and a skirt, or assigning attributes to images that can be used in recommendations or other customer-facing interactions.
Forecasting, however, is another matter entirely. While chatbots and image processors grow basket size and improve conversion rates, they just don’t operate on the same scale as buying the right products in the right quantities and putting them in the right places and the right prices in the first place. Every stockout, unplanned promotion or markdown is a forecast error, and if those errors can be fixed, they contribute to both the top line and the bottom line – at the same time.
Right now, all of the AI-driven forecasting activity is happening in grocery replenishment. That’s fine – it’s delivering value, which even toddler-aged AI should do. But in the forecasting game, grocery replenishment, where sales data is high-frequency and inventory is generally plentiful, is probably the easiest problem to solve. New product introductions, where an item has never been sold before, or intermittent demand items, or short lifecycle items where you have to get it right without having much insight into initial demand – these are all much more difficult problems to solve. And outside of grocery, it’s these problems that really need solving.
Personalization has been a hot topic in retail for a few years now. At this point, personalization online is ante to the eCommerce game – it works, it delivers value, and consumers notice more (in a bad way) when it’s not there than they now notice that it’s there.
But even maxed out, online personalization only reaches about 15% of all retail (at least in the United States; global mileage varies). If retailers really want to unlock value, they need to figure out how to do that in the store.
The quick and easy temptation is to throw personalization onto the mobile app and create some way of giving it a store context, so that the same online personalization can be taken to consumers in the store. There are two problems to this approach. One, it assumes that consumers turn to their mobile devices for in-store help. But usually, if consumers pull out their phones in stores it’s because they had a problem the store couldn’t solve. Personalization at that point is too late – the consumer is either looking elsewhere (Amazon maybe?) or moving on to the next thing.
Two, it leaves an enormous opportunity on the table, because that approach ignores the most valuable part of a store: the store employee. Yes, it’s tricky to deliver personalization through a store associate. You do it wrong and you end up in creepy-land very quickly. But the only way employees in stores are going to continue to be successful is if they can play an effective role in initiating, assisting, or completing an omnichannel journey. And that means having enough information about a customer to be useful in that role.
I’ll be honest, I don’t like talking about payment technologies. It’s a complicated space, full of many different layers of players, some of which actually serve multiple roles within the ecosystem. But the reality is, if you thought “EMV and done” when it comes to payments, think again. Even in countries where retailers have long moved on from EMV to contactless, omnichannel payments, mobile payments, and cross-border payments still confound.
Upgrades to providers can mean requiring upgrades to peripherals. Fraud is still as big a problem as ever. And the only thing retailers know for certain is that all they want is to get out of holding credit card numbers or any other part of the payment transaction. That means holding tokens instead, for the lowest cost possible and with the least amount of complexity, and that’s just not that easy to do or find. Which means that as much as everyone would prefer to move on from having to deal with payments, 2019 will still be a year where that’s a big topic.
The Bottom Line
Rather than pushing the envelope on robots or some kind of magical connected in-store consumer experience, some of the things retailers will be talking about in 2019 will be things the industry might have moved on from too quickly. Payments is a tangled knot that is expensive to untangle and full of plenty of unintended consequences. Cloud still have too many detractors to just call it a done deal. But retailers are also focused on innovation too – and AI and in-store personalization will play big roles there. However, in the end, and as always, it will be consumers who will dictate just how quickly retailers will be able to move on any technology topic.
Date: January 31, 2019