Much product discovery happens before the digital shopper even opens the shopping app -- via conventional media, social media, research and word of mouth. AI can drive message personalization in media and assortment personalization in the digital store, but it still very early days. A key limitation IMO is the linear nature of shopping on a screen. (Doesn't matter much if its a mobile or a desktop screen.) You can only evaluate one item at a time -- not scan a rack full of suits or dresses, or a wall of TV screens, or an aisle of arts & crafts items, or a rack of garden tools. An AI driven online merchandising system will ideally present the most likely items first, based on things the system knows about the individual and/or others resembling them. But is that really "discovery"? Seems more like prescriptive selling to me. Useful when the shopper has an objective in mind, but not really a browsing experience. Digital shopping has displaced and will continue to displace a large number of purchasing missions, but it's not ideal for finding new things. When you start searching, many alternatives are instantly filtered out.

Wise move by Walmart.com to rein in its ecosystem of third-party sellers. Quality and reliability should take precedence over sheer scale and breakneck pace -- in fact, those values may be mutually exclusive to a degree. Amazon and Alibaba may each follow a different path. So be it. Walmart should project and protect its brand reputation in its own way. Lore's comments last month about acquisitions suggest Walmart.com is assembling a house of brands, not an electronic swap meet. That's a distinct strategic difference.

For my money, urban shopping districts succeed best when they are hyper-local. That means a preponderance of independent merchants who live in the community. Replicating the regional mall store assortment on Main Street seems like a poor answer. These commercial neighborhoods thrive on foot traffic; they are not drive-to destinations. The type of merchants matter, Even in the face of gentrification, merchant associations can be instrumental in preserving design standards and engaging neighborhood residents. However, when Orchard Supply or CVS wants to take over a double storefront at above-market rents, that presents a tough battle. Here in Tucson, AZ, we are enjoying a downtown revival that has been powered by two main factors -- an active community development effort and the installation of a modern streetcar line that has stimulated new business investment along its path. In the past 5 years, we have seen new residential, hotel and restaurant projects that have instilled pride and optimism.

I take a radical (and admittedly futuristic) position on this issue. Each customer should own and control their own personal data profile, in a personal, encrypted vault. Access to this information should be transactional in nature -- if you want a glimpse of mine, offer something of value. Saving meta-data about those interactions is OK, but making a copy of any individual profile is forbidden. True data security is by now proven to be impossible. So why keep denying it? Retailers should simply not be in the business of gathering, accruing, and protecting the data of millions of individuals. Their data troves are an irresistible temptation to criminals -- like so many Fort Knoxes but housing a currency that is far easier to steal, transport and spend. Sure, customer databases are hard-won assets, so it's not easy to walk them back. But the cost of maintaining and securing them (including escalating technology and damage to brand value and customer relationships due to breeches) may soon exceed the benefits. Worse yet, each consumer-facing company assembles its own version, which means that hundreds of copies of our personal data are out there waiting for thieves. The personal data lock box would have the effect of distributing information, requiring it far more effort for bad actors to gather them up. Discontinuing retail customer databases would also eliminate a major market for hackers. The tech to enable personal data lockboxes is on the horizon -- possibly making use of the Blockchain. Like I said, this is a radical perspective. Where are the visionaries who are ready to consider, debate, and create it?

I give Walmart's Pickup Tower points for innovativeness, but I don't love it as a mechanism for optimizing the store pickup experience. It's essentially a giant vending machine and billboard. Useful for non-food orders that are not too large or heavy. Not well-suited for multi-item grocery orders that require temperature control. I worry that the Towers pre-empt human interaction that might lead to incremental sales. They need to be loaded by human order-pickers in the stores, too, so it's not much of an automation play. On the positive side, Pickup Towers may help alleviate a bottleneck at the customer service counter for uncomplicated orders. A location close to a main entrance might facilitate quick service. Placing them in the rear of the store seems manipulative and mean. Don't get me wrong. In-store pickup is an essential service element for Walmart. It can and must strategically allocate some floor space for this. It feels to me that the Pickup Towers are a temporary measure in play while the company works out a more comprehensive solution for store order fulfillment.

Amazon versus Walmart is a proxy for the larger war: digital-first versus store-first retailing. It's amazing to me that in the present era, a large real-estate footprint with long-term leases can turn out to be a liability for some retailers, where not long ago it was evidence of market dominance. Grocery will be the pivotal theatre of battle, due to purchase frequency patterns and total value generated. In the near term, Walmart (and other supermarket operators) retains the power of incumbency. Amazon may succeed in reinventing the category, but it will need to reinvent stores in the process. It appears to be trying pretty hard to do just that.

It strikes me as ironic to cover Walmart as an "underdog story" in the digital retail era, but there it is. Amazon has set the tone, both with its breakneck innovation and its unprecedented financial muscle. Walmart seems to be attempting to reinvent itself in anticipation of a future that it could not have anticipated before the present decade. Walmart's digital brand portfolio could soon give birth to scattering of specialty stores located in select downtowns and malls. It has an edge over Amazon in physical retail know-how, although the gap will narrow fast. Amazon is digital-first; Walmart is store-first. Their trajectories have them on a collision course that is re-shaping retail.

For what is essentially a tech-enabled convenience store experiment, Amazon Go is garnering a huge amount of attention. Most of it is deserved. The presence of so many helpful humans has at least two crucial purposes: They train and reassure shoppers who are still learning how to use the store. They perform stocking and merchandising tasks that cannot (yet) be automated. The absence of checkout lanes and registers may be the most visible departure from the retail norm, but check-in-and-walk-out transactions are just the beginning. The tech in the ceilings and behind the walls is about in-store sensing. Amazon wants to know as much about its shopper behaviors and store conditions as it knows about its online environment. It's this digital-first orientation that store-first retailers should take to heart if they want to be among the survivors.

If Walmart's broad strategic goal is to simply win a larger share of America's wallet, then it makes sense to go after the heaviest spending customers and categories. Health care is a natural add-on that has synergies with the company's existing pharmacy, food, and wellness categories. But it should beware the complexity. Health insurance and healthcare delivery are heavily regulated and subject to ever-changing political winds. HIPAA privacy rules require a firewall around healthcare data that could (and probably should) block some tempting cross-marketing concepts. In light of the recent CVS-Aetna and Albertsons-Rite Aid deals, Humana's pharmacy benefits business may be the gem Walmart covets most. As Walmart customers age, they may be expected to spend a larger share of their budgets on healthcare and prescription meds. Taken in aggregate, the recent wave of acquisitions and deals in this sector are troubling to me. They send health care further down a for-profit spiral that will be increasingly difficult to unwind.

Despite the headline, I don't think store-as-media is the most important takeaway here. For these three retailers, at least, physical presence is largely about customer acquisition, brand experience or meeting short-term needs. There was a consensus that long-term lease commitments and multiplying prototypes across the landscape are incompatible with their "agile" approach. Pop-up shops, roll-up shops, and a revolving array of short-term leases are at odds with past norms which assume a vast lease portfolio is a primary indicator of retail success. All stores are communications environments. It's always been that way. The key question is: What messages are being delivered?

Clearly online retailing has evolved past its race-to-the-bottom origins to embrace new digitally-enabled experiences. The luxury brands mentioned here have discovered that their shoppers value other aspects of the experience - especially personalization that involves actual persons. Who woulda thunk? I don't really think this is specific to Millennials, however. Plenty of folks in other cohorts can afford to purchase premium experiences. The aha! here is that some brands are working out how to appeal to those customers recognizing that the experience should align with luxury expectations, not just low price and free delivery.

"It's the store, stupid!" Most are now digitally enabled, true, but their status as value-added intermediaries between makers and consumers is as essential as ever. This is true whether shoppers access them on the asphalt highway or the digital highway. Many of us in this forum already think similarly about this. It's remarkable that NRF and Forrester have put the question to shoppers in order to confirm evident reality. Retail pros don't need to be convinced, but it seems to me that the Wall Street financial community has been generally slower to understand.

Reliance upon a summary statistic to guide retail returns policy is not adequate. The dynamic is likely to vary greatly by retail channel or product category, as Paula smartly observes. Retailers who pay active attention should be able to identify specific items that are commonly associated with returns abuse. Why not label those items as "subject to limited returns policy" with associated fine print, while leaving the broader, more liberal policy in place for the rest of the assortment? Tracking frequent returns behavior in the context of a frequent shopper program should enable very bad actors to rise to the top. Where a liberal return policy engenders stronger loyalty, it may be in the interest of the retailer to minimize obstacles.

Setting aside the excellent commentary here about Nordstrom's technology strategy, I'd prefer to pursue the question, "What will each of these innovations do for Nordstrom shoppers?" BevyUp, the tech-enhanced tool for creating "style boards," seems like a natural upgrade to the traditional little black books kept by Nordstrom sales associates to help them communicate with relevance with customers. The tech adds value because it provides perfect memory of interactions along with enticing visuals. Great for maintaining contact with shoppers who make personal style a priority. MessageYes, the "conversational commerce platform," seems oriented toward mobile-first shoppers who don't want to interact with people very much. Embedding an intelligent agent within the app to help shoppers zero in on preferred items more easily is a good objective for all digital retail platforms, however. Nobody really enjoys scrolling randomly through hundreds of garments on a tiny screen. Controlling the I.P. for these solutions may let Nordstrom box out competitors on these service experiences for a short while.

Hmmm... is it the Van in the band, or the band in the van that really matters most?