Reflections on Groceryshop 2023

This week was my first time attending Groceryshop in Las Vegas. The event brought together numerous #retail, #CPG and #tech companies to talk innovation. The explicit theme of the event was ‘unified commerce.’ The aspiration for a more integrated shopping experience that marries online and offline rather than simply coordinates them is a lofty one.

However, two implicit themes underpinned nearly every conversation here - even more than unified commerce:

Retail Media Networks

First, Retail Media Networks have reached a new level of maturity and a new level of breadth across the sector. Retailers at Groceryshop acted like tech companies. They sponsored large booths, brought large sales teams, and they pitched hard on the power to reach their audience. For several years, RMNs were primarily the domain of national retailers. Now, the regional grocers have networks of their own. These are a major new source of profitability that are offsetting the long term decline in gross margin that national retailers have spurred through more price competition.

At Groceryshop, many of the retailers and the technology providers discussed the work they are doing to enhance the networks. They are increasing the available inventory by creating more assets that can be sold in the network from the physical store like screens and radio. Retailers are also trying to prove that they are driving incremental sales. However, there is still not consistent measurement across retail media networks when it comes to measuring ROAS (or iROAS).

Retail Media Networks have probably yet to reach the peak of their ‘hype curve,’ but the level of hype is certainly high. Oliver Wyman expects in the coming years for the industry to start integrating them into their core customer strategy rather than a standalone offering.

Gen AI

Second, there is near universal agreement that GenerativeAI will be a massively disruptive technology but not yet agreement on which parts of the food retail ecosystem will be disrupted and many people misunderstand what they should already be doing. Some senior leaders highlight the exact right actions the whole industry should be doing – they are putting LLMs in the hands of their associates in safe and secure ways to drive early quick wins while evaluating whether additional colleagues hassles.

And yet, there are still several myths that still need busting:

• Myth: Using LLMs requires significant infrastructure investment to get value for associates. The reality is that by enabling associates to access LLMs through secure corporate endpoints requires only a few days of development and a low cost per-query.
• Myth: Using LLMs poses a security threat for company data and their intellectual capital. The secure corporate endpoints do not expose the data they access or even the questions asked to the outside world. These are no riskier than the cloud providers powering the internet today.
• Myth: Using LLMs will have a negative environmental impact even as many retailers and suppliers have made net-zero commitments. The initial training for an LLM is very computationally intensive. However, the training does not need to be repeated from company-to-company. Building new use cases that tap into an LLM is not a carbon-intensive activity.
• Myth: Generative AI will solve personalization challenges in retail. Generative AI has very rapidly unlocked new ways to engage customers with truly personalized solutions. Many of the more innovative providers at GroceryShop are already finding ways to embed GenAI into more personalized experiences. Oliver Wyman’s own experience is that GenAI solutions can be an important addition to personalized solutions – but they are not sufficient to fully personalize a consumer experience. Creating fully personalized omnichannel experiences is an entire business transformation.
• Myth: Generative AI is the best solution for solving supply chain challenges. It will be many years before the food industry uncovers the full power of Generative AI. However, even today there are many Machine Learning / Artificial Intelligence solutions that can improve supply chain performance with better forecasting, better product visibility, better optimization of routing and other operational improvements. GenAI solutions may support more efficient development of analytical solutions but is not a replacement for the hard work of data source identification and cleaning, model selection and hyper parameterization and tuning.

Over the next few years, retailers and their suppliers will certainly find many new ways to leverage Generative AI. Many have the right business approach to investigate, develop and deploy these new solutions. While Generative AI may have been the most discussed topic at Groceryshop, there are still many misconceptions out there.

My overall take

I believe the rise of the Retail Media Network and the omnipresence of Generative AI were justifiably top priorities for many here. In addition, there were many providers talking about shoppable video and other forms of social media. I loved seeing how dynamic some of these solutions are - and how retailers like are already using them to drive commerce. These are part of an overall push to create better shopping experiences – something the digital grocery experience desperately needs.

Several panelists – especially those discussing consumer trends – noted the high inflation that has affected so much of how consumers choose what food to buy today. However, while there were a few exhibitors and panels on new ways to get products to consumers under ‘unified commerce’ there were not as many technology solutions that were striving to drive down the cost of existing channels.

Food retail continues to be a dynamic sector and the level of energy behind new technology solutions at Groceryshop serves as further validation of this. I'm curious what other attendees thought or what you heard from peers who did.