Data and Machine Learning: The New Grocery Battle Ground

Kroger looks to leverage big data and machine learning to win its share of consumer stomachs, and wallets

“Data is the new battle ground” according to Stuart Aitken, chief executive of Kroger’s analytics unit, 84.51°. [1] Since Amazon’s arrival on the retail scene, retailers have witnessed the power of big data as Amazon has leveraged its consumer data to capture share. Now with Amazon’s stunning 2017 acquisition of Whole Foods, America’s grocers are now staring down the same barrel retailers have been. This shift to omnichannel retailing has created many challenges for retailers as they need to profitably provide a compelling offering to consumers across each channel. But this widening of the shopping funnel created many opportunities for retailers – and for grocers now too.

To date, the grocery channel has seen lower online penetration due to consumer hesitance to buy food sight unseen. [2] Consumer purchasing behavior is changing however, thanks in large part to the growth of same-day delivery options like Instacart and the growing Whole Foods / Amazon platform. Food shoppers are increasingly splitting their dollars between supermarkets, discounters, club stores, specialty shops and online food retailers, as compared to the historical “one stop shop” approach. [1] In the one stop shop world, Kroger, the largest US grocer by revenue, only worried about competing with Walmart, Target and other supermarket chains with brick and mortar footprint. But this shift in consumer behavior has sparked an online arms race not only with their traditional brick and mortar competition, but the online behemoth that is Amazon. A major change with this blurring of channel lines has created a huge opportunity for Kroger and other grocers as they are able to amass vast amounts of proprietary consumer purchasing and behavior data that can be used in previously unimaginable ways.

In 2015, Kroger acquired dunhumbyUSA, a subsidiary of the UK-based analytics company and since renamed as 84.51°, to harness the power of big data and machine learning. Kroger has turned to machine learning to develop predictive models that can in turn be used to inform strategic decisions across all aspects of the business – everything from assortment selection, store layout, pricing decisions, promotions, and supply chain and logistics decisions. Kroger 84.51° deploys the “supervised learning” method of machine learning, which “trains a model on sets of labeled data and then uses the resulting models to make predictions or classifications on data where the outcome is uncertain.” [3]

Kroger has sought to embed machine learning across the organization to automate processes and products. In late 2017, Kroger launched the “Restock Kroger” initiative. A 3 year, $9B transformation program to improve its ecommerce, digital, and omnichannel business and redefine the customer experience. [4] The effects of these efforts can be seen in assortment and store layout changes in over 1000 Kroger stores across the US. On perhaps the more innovative front, we are seeing increased personalized marketing efforts.

“Today Kroger uses 850 algorithms to personalize the coupons it mails to 12 million households. The company can use purchase data to determine whether someone has gone on a diet, had children or retired, and to market different products to those customers accordingly. Some Kroger coupons have a redemption rate of 65%, compared with a national average of about 5%, executives say.” [1]

Kroger has not disclosed the financial impact from its personalization efforts, but according to BCG research, personalization efforts can drive incremental revenue growth in the range of 6-10%. Kroger delivers over 3 billion personalized recommendations annually, and over 1.1B personal digital coupons were downloaded in FY18 Q2 alone. [6]

Investors have rewarded Kroger for these efforts over the last year, as the share price has risen roughly 50% from $21 a share in October 2017 to almost $32 a share in November 2018, despite disappointing second quarter results. [7][8] These are the right moves for Kroger to deliver a compelling omnichannel offering. For Kroger to maintain its leading position in the grocery world, it will need to accelerate its efforts in the short- and medium term. In the short-term, Kroger’s use of data for machine learning should:

  • continue to focus on developing localized assortments across the US
  • accelerate its space optimization efforts by aligning shelf space to sales to “Amazon-proof” its stores and drive productivity
  • deliver consistent omni-channel personalized offerings

Over the medium-term, Kroger should further synchronize data across its DCs, stores, and vendors to improve operations. Accurate store inventory data combined with predictive analytics to forecast orders, fully Automated DCs with store ready picks, and a smart vendor collaboration portal can drive significant operational efficiencies. These efforts can minimize stock outs, and reduce spoilage to free up critical cash flow.

It remains to be seen if Kroger’s investments in machine learning and Restock Kroger will be enough to maintain grocery leadership in the US, or if these investments are just table stakes to stay competitive with market giants like Walmart and Amazon in the short-term?

 

[800 words]

 

[1] “Grocers Imagine the Store of the Future,” Wall Street Journal (October 2017)

[2] “Online Grocery Retail Forecast, 2017 to 2022 (Global),” Forrester (March 2018)

[3] “84.51° Builds a Machine Learning Machine for Kroger,” Forbes (April 2018)

[4] “How U.S. Retail Giant Kroger Is Using AI And Robots To Prepare For The 4th Industrial Revolution,” Forbes (July 2018)

[5] “Profiting from Personalization” BCG (May 2017)

[6] “2018 Investor Day Conference,” Kroger (October 2018)

[7] S&P Capital IQ (November 2018)

[8] “Don’t Panic Over Kroger’s Stumble,” Wall Street Journal (September 2018)

[9] “Building Watson: Not So Elementary, My Dear!,” Harvard Business School Publishing (October 2016)

[10] “Aspiring Minds” Harvard Business School Publishing (May 2016)

Previous:

The Power of Patient Data: How Flatiron Is Combining Big Data and Machine Learning to Revolutionize Cancer Care

Next:

Can AI save brick-and-mortar retail?

3 thoughts on “Data and Machine Learning: The New Grocery Battle Ground

  1. I really enjoyed learning about Kroger’s machine learnings initiatives. In my mind, what they are doing is impressive and if executed properly going forward, can help them maintain their leading position within the grocers’ landscape. It is important for Kroger to find ways to upsell to consumers – they can use the coupons to attract consumers to stores based on prior purchases (i.e., repeat buys), and simultaneously introduce new products to consumers as they walk the aisle (TBD how this may be implemented, but would need to involve personalized recommendations based on prior purchases).

  2. The machine learning initiatives they have implemented to date are very interesting, and seem quite innovative for a brick and mortar grocery store. From what I understand, Kroger is fairly mission-driven and people-focused, especially with their front line employees, to a greater extent than some other grocery retailers. Given that, I do wonder what the effects of expanding machine learning into things such as auto-pick in distribution centers would be in terms of morale (though I do understand the need for cutting costs if they want to exist in the long term).

  3. Great insights here, the entrance of Amazon into grocery is sure to shake things up in an already very competitive industry with low margins. To compete with Amazon, Kroger and other traditional grocers will truly be forced to raise their games, which will ultimately (hopefully) improve the shopping experience for consumers. Kroger’s investment and partnership with British online grocer Ocado will be another item to watch, as they work together to build out a delivery distribution network in the US.

Leave a comment