Usually day one of a show is more about impressions than penetrating breakthroughs.
But NRF 2018 day one in frosty NYC brought with it some big surprises. Hands-on demos showed the sexy future of hands-free shopping, and the practical ability to drill into data anomalies. Five quick hits and surprises:
1. From retail apocalypse to online shopping traction
It's no wonder the retail apocalypse talk has died down. As I discussed with Jeff Warren, Oracle Retail Vice President of Solution Strategy, Black Friday sales showed that online shopping is getting traction:
The scale is tilting. Retailers generated $7.9 billion in online sales on Black Friday and Thanksgiving - up roughly 18 percent from a year ago - which is clearly displaying an ever-more digital shopping preference. So how does this impact retailers?
2. Dark side of mobile traffic bumps - conversion issues
Mobile consumption, however, is not without problems. Peter Sheldon, VP Strategy at Magento Commerce, believes mobile has enough traction that "mobile only" is the development strategy of the future. But there's a big conversion hurdle:
We track a lot of the stats over the holidays, and consistently, what we've seen with our merchants is 60+ percent,of total web traffic came from mobile, but only about 30 percent of orders came from mobile. We've still got this challenge that the conversion rate of mobile is not where it needs to be.
The reasons? That's a longer conversation about the challenge of building highly effective mobile apps/experiences. Oracle's Warren added to the retail optimism:
Retail grew about 4 percent this year in North America.
3. Personalization at scale remains elusive
For those like me with a cynical "How much of that is Amazon"? rebuttal, Warren says Amazon is about $100 billion against a $4 trillion number. But seizing that growth means conquering omni-channel data barriers.
During a lunch with Oracle's Mark Koehler and IDC Retail Insight's Andrea Sangalli, we got to talking about my bugaboo: the immense problem brands have with true personalization at scale. Instead we get generic email blasts driven by cynical click metrics that favor volume blasts.
Sangalli said that disconnect jives with IDC's research. They classify retailers into four levels of sophistication in terms of segmentation and hyper-personalization at scale. Though there are some global variations, in general, the number of retailers IDC notes at the highest level of personalization are only about five percent - or, in some cases, even less.
4. Omni-channel customers are more valuable - but harder to serve
Oracle's Warren challenged how I think about omni-channel. At the same time he underscored why data disconnects are such an obstacle:
To me, the numbers that are most interesting - I think I referenced them in the webcast - of the 174 million shoppers over the five day holiday period post-Thanksgiving, the omni-channel shopper spent, on average, $82 more than an online-only shopper in that five day period. And they spent $49 more than an in-store only shopper. So, as retailers think about an engagement strategy and who are the most highly valued customers, this isn't just a single brand, it isn't just a single segment. This is across the aggregate of 174 million consumers.
I typically think about omni-channel as serving all customers effectively across touch points. But Warren takes it further, with data that "omni-channel" customers actually spend more. But that finding comes with a caveat: the data breakdowns that ruin omni-channel shopping better get solved. Warren:
One of the biggest asks that we get from the retail community is: as we help them figure out how to unlock more value from their data, and deliver better value to the consumer is, they ask us: "How do we simplify how we deliver value along the way?" They've made the business of delivering to the end consumer too complex.
Warren believes cloud platforms can reduce these dreaded data silos. He told me about customers who found a breakthrough as they pushed towards a "single source of truth," even as they reckoned with new external data sources:
The opportunity - especially if you think about structured and unstructured data - as you start to pull all of the data out of the silos and bring it together in the cloud, you get the single view.
5. If you don't have a good value proposition, data science can't help you
A tour of NRF would not be complete without some AI chinwagging. I figured I'd get that with dunnhumby CEO Jose Gomes and David Ciancio, Global Customer Strategist. Given that dunnhumby bills itself as "Customer data science for retailers and brands," I was expecting a heavy slather of data-science-cures-retailers-woes. But I didn't get that. Instead, Gomes surprised me by arguing that the real problem in retailing is lack of distinct value proposition. And, if you don't get that right, forget about data science:
If you don't have the value proposition that's catered to what your customers want and need, you're going to struggle... Once you've understood what the value prop is, that's when the data science becomes really powerful.
I'm simplifying a bit. Ciancio, who has a background with retailers like Kroger, reframed our discussion by saying that "customer science was vital for us to understand what our value proposition ought to be." But it's a good reminder: tools and tech won't save a brand with a fuzzy market focus, or lack of consumer loyalty.
Demo award #1 - practical category
The most practical/useful demos came from a show floor tour with Wipro, via their Data Discovery platform. One was a terrific Customer Churn example of drilling into a sales anomaly amongst a category of high-value customers via dashboards, until the problem demographic was identified. I plan to share more of that demo later, but here's one screen shot. This one shows how the Customer Churn anomaly is identified prior to drilling into the issue:
Wipro also had a useful example of speech analytics. I'm including that here, because we're hearing a load of hype around voice-enabled retail at this year's NRF. But this Wipro voice example is a practical angle:
What you see here are keywords that jumped out from analysis for hundreds of customer calls. These are ported into transcripts and run into Wipro's NLP engine for further analysis. This is a solid example of how previously underutilized info from customers (voice recording) can be aggregated for trend analysis (e.g. useful product keywords like "dry wall").
It was striking to hear the Wipro retail team talk about their retail traction, and how they achieved it. They talked about design thinking approaches, customer partnerships and advisory - themes I also heard from assorted customers I interviewed today. Customers are choosing retail partners with the tools and expertise to work with them over the long term - not just staff bodies or outsource operations. You're not playing in the retail consulting game if you can't deliver that.
Demo award number #2 - forward-thinking category
My second demo award goes to Yi Tunnel, who have developed some remarkable self-checkout technology you can see in action during this impromptu, narrated video I shot:
My experiences with self-checkout are so crummy I tend to use humans. Yi Tunnel is a different matter; if they can tell the difference between nearly-identical apples, I'll give them a go. They want to take on Amazon and Alibaba, an ambitious goal. Whether or not they get there, they do have 10,000 + products SKU'd, and they say their hardware is 5 percent of the cost of Amazon Go.
That's it for the mid-show wrap. I finished this piece amongst loads of customer use case interviews, so I'll have plenty more on where the successes are happening - and what the gotchas are.