Author: Tim Ouimet

05 Dec 2018
Agent-based Shopping

The Rise of Agent-based Shopping

 

Retail is at a tipping point:

This was a main theme at GroceryShop’s inaugural event last month in Las Vegas. The event was attended by over 2,200 retail and CPG executives and was billed as the industry’s leading event for innovation.

Opinions about the tipping point were punctuated by Nielsen’s prediction: in 5-7 years, as many as 70% of U.S. consumers will regularly purchase consumer packaged goods online (“Digitally Engaged Food Shopper”). By 2022, they speculated consumers could spend $100 billion per year for online groceries (equal to $850/year/household).

For shoppers, digitization lowers barriers, making it easy to source product, compare buying options, find offers, transact, and take possession. For retailers, this means further downward pressure on price as competition evolves across more market segments than ever before: store, product, time, customer, channel, pickup and delivery options, cross-sell alliances, marketplaces and shopping apps.

 

Direct to consumer

Another big topic at the show was brands going Direct To Consumer (DTC) via Instacart, Amazon, and other platforms. For shoppers, it’s only getting easier as computers and AI get better at sourcing, comparing, and finding buying options that best meet a shopper’s current need. In this digital future, relevant offers will find a shopper based on her context (location, preferences, urgency, etc.). Margins will follow a retailer’s ability to make its assortment, offers, and delivery hyper-relevant to a shopper’s context.

 

Artificial Intelligence

Innovative retailers are leveraging AI to defend margins by segmenting markets better and by personalizing services and offers. In her talk, Google’s Laura Antonolli demonstrated Google’s AI-driven conversational assistant, in a machine-to-human interaction. Traditionally, a sales assistant’s role is to connect, understand, personalize and serve. In Laura’s demonstration, Google’s AI assistant searched for a hair salon, called to coordinate calendars for an appointment, and selected a haircut from an array of services.

 

Agent-based shopping

We are witnessing the rise of a new paradigm in retail – agent-based shopping. Laura said shopping agents and voice-activated search are a new battleground, stating that 22% of Google’s mobile search is voice. Forrester Research Analyst George Lawrie reported that “digital is no longer just a marketing channel, it’s now a sales channel.”

In one retail use-case, George shared that Alexa users spent, on average, £8 (eight British pounds) more per basket than at Morrisons. As of today, use cases for digital agents and household consumables are limited. That said, the promise ahead is for agents to create hyper-relevant offers and close sales based on an individual shopper’s context, values, and preferences.

 

Demand-side attributes

In large part, limitations come from the need to understand how shoppers compare substitutable product offerings and how those comparisons change under different pricing and offer scenarios. This applies to both within and beyond a retailer’s four walls. As an industry, we’re really good at supply-side product attribute data (i.e. weight, size & ordering info.). But we have yet to understand product attributes on the demand-side. Whether online or in-store, the old adage “price drives sales like no other factor” still holds true. From that perspective alone, comparability of pricing and offers within a category and across the marketplace are significant dimensions of demand-side attributes.

Comparability lies at the heart of understanding shopper values and preferences. As such, this represents a central value driver in any automated shopper feedback system. Without comparability information, offer personalization is largely blind to a shopper’s context, i.e., blind offers are not relevant.

 

Data takes a new turn

This brings me to a second take-away from the event. We are witnessing a fast acceleration in the variety of demand-side data elements offered by vendors in the space. At GroceryShop, these companies included: Engage3 (my company), Nielsen, Gladson, 1010data, EnterWorks, Label Insights and many more. Each of these companies offer unique data elements relating to different and new demand-side attributes.

As an example, Engage3 provides store-specific comparable pricing data. Nielsen provides measurement data aligned with many causal data elements. Label Insights provides detailed on-package attribute, ingredient and claim data. Data sharing and data feed integration between data vendors is also accelerating. These sharing and integration relationships open new paths to support the full promise of agent-based retailing.

Given the importance of product comparability, expect comparability to emerge as a primary focal point for integration. At Engage3, these sharing and integration relationships are beginning to yield new benefits, including:

  • More trustable competitive pricing and product comparability data
  • More precise methods for measuring price image
  • A broader and more frequent census of the competitive market
  • Deeper insights into strategic and tactical pricing dynamics
  • Improved automation and control over strategic pricing

All of this helps retailers automate and increase ROI from their pricing, offer generation, segmentation and personalization efforts.

In his talk, Earth Fare’s CEO, Frank Scorpiniti, spoke about AI as an “invisible advantage” that “removes repetition.” He reported that price promo and promo cadence all together yielded ROI of 300 basis points (3% of sales).” Adding “automation requires data quality” and that Engage3’s data has “near 100% accuracy.”

Getting data trustable is one step. Preparing integrated data for automation & advanced analytics is a step beyond. Supporting a client often means integrating a variety of data feeds from across a client organization with that of multiple data vendors. And building those feeds into a model is crucial so decisions can be automated in a controlled way.

Adding to the momentum, research dollars are beginning to flow as universities define research priorities in this space. One university we spoke with plans to create a center for excellence in food through the integration of personalized health, nutrition, and sustainability. Their effort would align both industry executives and academics from the business, engineering, supply chain, medicine & nutrition schools.

Agent-based shopping is set to emerge as a new battleground. Retailers that are positioned to make use of these new data feeds will climb the evolutionary path faster. Expect the industry to evolve rapidly in support of agent-based shopping.  It’s an amazing time to be in retail!

 

 

25 Oct 2018
KVI and Price Image

Known Value Items – Drivers of Price Image

A Shopper’s Store-switching Decision

A KVI is a known value item. It’s an item that disproportionally drives the price value perception. So, in a grocery store it would include eggs and an automotive store might include motor oil, and a convenience store it might include cigarettes.

The reason that KVIs are important is because they drive a shopper’s store switching decisions. If the retailer’s prices are out of alignment with the prices that shoppers remembered, then the shopper can reevaluate their decision to shop with that retailer.

A question you can ask a shopper is, what items do you stock up on? And at what price points do you stock up? And you’ll begin to understand what a KVI is with the answers you get to that question.

A Tiny Number of Items

Another element that’s really important with the known value items is that it’s a very tiny number of items that drive a retailer’s perception in the marketplace. Typically, about a third of the price perception comes from only two-and-a-half percent of the products. It’s a very concentrated number of items, and this holds true across grocery, drug, mass,convenience, pet, auto—virtually all retail sectors. So, getting it right is critical.

Dynamic KVIs

Traditionally, retailers will evaluate their KVIs once a year. Over time it’s gotten to a more periodic basis where they’re doing it more often, but the market’s changing faster today than it’s ever changed before. Things are getting localized, things are getting personalized, and with that the shopper’s price perceptions are being set more dynamically.

All of these things mean that calculating KVIs based at the enterprise level is the wrong way to do it. The analysis needs to come down to the store level, down to the shopper level, down to the daily level, and have items coming in and out of the KVI list at those lower levels.

Increased Complexity

The challenge is that all this results in a lot more complexity that needs to be managed. The comp shop programs that were easy for one person to manage before now explodes the amount of competitive data that’s needed and the amount of management time that’s required.

A Platform to Manage Margins and Price Image

The retail marketplace is only going to get more competitive, and retailers need a platform to support themselves in this new environment. At Engage3, we’re on a journey to build that platform to enable the retailer – the early adopters – to outpace their competition so they can outperform them in terms of Margin and Price Image.