Infor's retail ambitions - on machine learning, Whole Foods, and where the traction is

It's been six months since I got my last retail update from Infor at NRF, which in retail terms is about six lifetimes.

Heck, even since my diginomica colleague Madeline Bennett reported on Infor and Whole Foods in May (Infor’s Duncan Angove – riding out the retail sector turbulence), Amazon acquired Whole Foods, leading us to wonder how Infor will fare in the middle of two converging relationships.

During a sit-down at Inforum 2017, we got some answers. Infor's Doug Tiffan, who sports the intriguing job title of Chief Merchant, also revealed which of Infor's retail pursuits have the most traction.

Infor has one of the most aggressive industry orientations in the (cloud) ERP space, with retail being a core vertical. Though Infor didn't have a dedicated retail unit until October 2015, its CloudSuite Retail push has been aided by several key acquisitions, including Predictix, which comes up in this article. Big customer wins, from Whole Foods to Home Depot to DSW - featured in the second day keynote (pic above) - have bolstered Infor's retail swagger.

Oh, and don't give Infor the song-and-dance that aside from Amazon (and maybe Walmart), the rest of the retail industry is borked. Angove believes that Amazon is the model, not the grim reaper:

Amazon is a technology and supply chain company. What’s going to save retailers is behaving in the same way.

Why retail needs demand management

With that in mind, where is Tiffan seeing the action? One biggie is Retail Demand Management, powered by their Predictix buy:

There's very public use cases out there with Home Depot, for example, with assortment planning and how they've leveraged it to become more efficient with their inventories and their assortments.

The brick-and-mortar expansion model is giving way:

The old way of retail that I grew up in was that you expanded. More stores was the key to growth, right? Using Home Depot as an example, what they realized was that you're going to run out of space if you've got that mentality. And now with this omni-channel world coming at us, it's no longer a brick and mortar growth story. It's about become more operationally efficient with your inventory.

That's where predictive retail comes in:

Demand management can recognize relationships between items and other causal factors that humans typically would never have the time to look into... You can get much more accurate forecasts. Of course, nothing's perfect, but the accuracy level's significantly better.

Tiffan sees demand management needs across forecasting, assortment planning, pricing, and allocation. Give CFOs better inventory control, and you'll peak their interest:

When you talk to a CFO and say, "Hey, you're running a $50 billion company and I can drop your average inventory by millions of dollars - and you don't lose service level," they're like, "Tell me more." Because inventory is where you tie up so much cost.

Tiffan sees the chance to build machine learning into demand management:

This product is a cloud-native solution. So when it was designed, it was never designed to be constrained by computing power... The proposition of having machine learning embedded in our planning solution gives it so much potential.

How will Amazon's Whole Foods pursuits impact Infor?

Prior to Amazon's announced acquisition of Whole Foods, Infor was already knee-deep in a co-innovation effort with Whole Foods, building a next-gen retail merchandising system. No small endeavor, this system will eventually replace twelve legacy products across Whole Foods' 400+ stores.

Tiffan says by all indications he's seen, the co-innovation proceeds:

What I have heard internally is that everything is still full-speed ahead. There's no plans to change.

This merchandising system was always intended for all retailers. One feather in the cap is Crate and Barrel signing on as the next co-development customer. As for Whole Foods, Tiffan reports that Item Management is now live:

I want to say it's somewhere between 700,000 and a million items being managed by that solution in their "365 store" format right now.

Tiffan sees more project wins coming: "We're setting them up to be able to have perpetual inventory. Ordering is easier."

Whole Foods is also putting in Infor Demand Management; some pricing and category management functions are already in place. Tiffan says shareable results are on the way; for now, foundational work is happening. As he puts it, "We're putting in the plumbing right now."

In keeping with the AI themes of Inforum 2017, Tiffan sees a big role for machine learning in this merchandising system: "Let's let machine learning think about assigning attributes." Humans might not spot the "significant relationships" between items that could impact purchasing/inventory.

Previously, Angove shared how Infor's GT Nexus network will affect Whole Foods:

Whole Foods will be bringing their 16,000 suppliers onto the network, publishing and managing their item catalogue on the network, they’ll be doing all of their ordering, they have complete inbound visibility of all of the products coming into their shops and depots, and we’ll be doing all of the financial settlement on the network.

We’re rethinking every aspect of the retail supply chain with a network mindset, from product discovery and sourcing to purchasing and financial settlement.

My take

Tiffan says that the demand management functionality in CloudSuite Retail will eventually be offered to midsize companies, in a "templated" format. Beyond demand management, another hot ticket is Infor's cloud-based point of sale (POS) solution, in use by Nordstrom:

That will probably be one of the largest POS deployments in the cloud that we've seen. It's not just point of sale. It also has store inventory management in it; it has a pricing module in it, so that one's been getting a lot of play also.

Infor is now pushing to integrate that with their "beyond omni-channel" offering: "Eventually, stores will be tied in to our Converged Commerce solution set."

Amazon's acquisition of Whole Foods creates both pressure and opportunity for Infor. Whole Foods has a decentralized approach to IT that would present a challenge for any vendor. If Infor takes the Whole Foods account beyond merchandising and into their entire CloudSuite, they will be on the hook for IT credit or blame.

The opportunity is to build a merchandising system so compelling that Amazon wants to invest and replicate it. In a pre-Inforum chat with CEO Charles Phillips, my colleague Derek du Preez quoted Phillips on the potential to bring new "pop-up retail" models:

I also think this Amazon thing accelerates one of the concepts we were thinking about. We had this concept of pop-up retail, which means that if we could make the fulfillment turnkey – where we connect the fulfillment economy to the engagement economy (the front end), we could basically do an outsourced, turnkey fulfillment system.

In this scenario, you don't try to compete with Amazon's supply chain dominance on your own. Instead, you piggy back it. You could imagine seasonal retailers activating and de-activating their distribution network after the holidays.

There are tough times looming for "legacy" retailers who will struggle to make the right decisions on store modernization versus closures - perhaps enough tough times to curb even Infor's enthusiasm.

Infor has a multi-year journey ahead building their retail co-innovations. If they play their cards right, they may find themselves with Amazon's retail wind at their back, be it pop-up stores or some other disruptive model not yet foreseen.

That's good progress since 2015. I expect to get updates from the Infor team at Shop.org in September and NRF's "Big Show" in January - stay tuned.

End note: Steve Brooks of the Enterprise Times joined me for this interview, asking plenty of useful questions. Look out for his story.