Remember ‘The Hunger Games’?
You Had to Kill Everybody to Win

By David Lobaugh, President, August Partners, Inc. | May 2021

 

The Ultimate Food Fight for $760 Billion in Prize Money

In the movie, it’s a contest with only one survivor. In our industry, it’s a battle involving almost 40,000 grocery stores and multiple corporate players – including Amazon, with their ‘Hunger Games’, data-fueled, algorithm-driven, take-no-prisoners approach.

U.S. grocery sales have increased by $357.1 B since 2000, but note the rate of the 5-year incremental percentage growth that began in 2010. From a sector that was growing less than 1% per quinquennial, it has increased by 4% from that point to 2015, and then by 6% to 2020.

According to eMarketer, U.S. smartphone grocery app usage grew 41% during 2020, during the peak Covid months. So now it’s a physical + online contest with a whole new set of rules for the stores themselves and for grocery-anchored centers.

It’s a Numbers Game, But Proceed with Caution

My friend Gregg Katz – Chief Strategic Officer of TSCG – has recently published a series of short gut-punches regarding the way our industry is using or mis-using data. You can access it here.

Gregg is one of the best in our business at wrangling immense amounts of disparate data from geofence, market, GIS, and consumer sources… and making it all make sense for TSCG’s clients and brokers. He’s the first one to advise his own folks not to make up a story that isn’t there. Quoting his post:

“We see what we want to see, especially when it verifies what you are biased towards or want the decision to be.”

My interpretation: Instead of using data to shape your case, use the data to shape your decisions.

Lately, we’ve been doing quite a bit of work throughout the country for a specialty grocery chain, as well as for our grocery-anchored center clients and for a municipality that wants to entice targeted grocery operators. We’ve evaluated potential acquisitions, quantified in-center cross-traffic pollination patterns, advised on anchor replacements, etc. – employing a combination of intel sources, including location data, consumer research findings and ambient market data to “reveal the truths,” as Gregg puts it in his series.

The North Jersey Grocery Store Tango

It takes two to tango, but let’s begin by looking at three free-standing grocery stores in New Jersey’s Hanover/Parsippany market to get our bearings:

We began at the dashboard level, but went deeper into the “raw” source data to pull usable/actionable analyses from this NJ geofence study.

The store trade areas above are defined via our proprietary methodology, based on shopper visit percentage generation (50% ± for the Core TA; 25% ± for the Extended TA) so that there’s a common-denominator standard that allows like-to-like comparisons. Underlying shopper visit data are appended to the block-level microgrids to facilitate the visit share calculations.

Wegmans and ShopRite are in a dead heat for visit share, but if you look at the TA maps, you can see that those two stores each generate their shopper visit patterns differently. Wegmans depends on an extensive geographic reach; ShopRite depends more on generating close-in shopping trip frequencies.

The graph at left summarizes how each store builds its shopper visit share over a pre-Covid, 12-mo. period (We have re-based the 5-mi. data to 100%).

ShopRite is a front-loaded-traffic store that achieves 62% of its visits within a two-mile radius. Wegmans is a back-loaded-traffic store, generating just 43% of its shopper visits within the first two miles, but ramping up to 57% of its 5-mile draw in the 3-5 mi. radius (vs. 38% of ShopRite’s visits).

Same overall visit share. Different geographic draw patterns.

Takeaway: Define Trade Areas by Shoppers, Not Shapes

The preceding grocery store trade area (TA) maps reflect detailed shopper points-of-origin down to the microgrid/block levels that are actually producing venue shopping trips. Note that in reality, a TA is made up of pockets and clusters of consumers – not by  broad, contiguous shapes; or large zip code blocks; or by drive-time polygons. By weighting and aggregating shopper visit-generation tranches into the top 50% ± (red microgrids) and next 25% ± (dark blue microgrids) retailers and centers can view an accurate Core and Extended TA picture. Not just the vague amoeba shapes found in dashboards, but instead, precise, down-to-neighborhood-level draw patterns. Subsequently, shopper demographic and category-spend data and shopping frequency levels can be appended to provide an in-depth profile of shopper characteristics.

Another way of looking at the data – useful when you want to visualize how the stores are generating their 5-mi. radius shopper visits:

The graph above visually depicts how one grocery store (Wegmans) is playing the long game and the other (ShopRite) the short game in this market. If you were ShopRite, you could 1) pull the zips from the location data, 2) isolate and correlate the Zip 9’s to where you see Wegmans exerting high traffic draws, and then 3) aim your direct mail and digital marketing to those customer clusters. If you’re Wegmans, you would want to flip the script by targeting Zip 9’s closer-in to build shopping frequencies in the 1-Mi.-2-Mi. radius range in which ShopRite excels.

Takeaway: Use Zip 9 Breakouts for Effective Shopper Segment Targeting

You’ve no doubt noticed those catalog and magazine labels… the ones with your zip code, then a hyphen and then four more digits? They are called Zip + 4’s or Zip 9’s and they match up with neighborhood-level US Postal Carrier Pre-Sort Routes. They are a mainstay of those grocery store direct mailers you get every week. When you use location data to track your shoppers, you can filter and market to shopper segments such as Covid vs. pre-Covid traffic-level contractions, competitor incursions, high-frequency/Best Customer shoppers, lapsed-shoppers, Saturday shoppers, etc. You can shape content and also save production and media costs – especially in direct mail situations, but also in social and digital outreach. We all may think that our grocery stores already know this… but it might be nice for centers to share their data with their tenant partners… just a thought.

The Shopping Frequency Game-Changer

From its Burger Bar (with kids’ meals) to its 52,000 SKUs (4,000 organic), to its “farmers-market” style fresh produce, to its “Meals 2 Go” offerings (left), Wegmans leverages its 113 K square feet not only to extend its geographic draw, but to win the Saturday shopper sweepstakes: our study revealed that Wegmans captures 37% of Saturday visit share vs. 32% for ShopRite and 31% for Stop & Shop.

And yet, the two stores have the same visit share. If Wegmans is so good at market penetration, and is the dominant Saturday destination, how can that be?

Shopping frequency. TA population is important, but the key is how often you can get each consumer to shop at your store or center.

To Gregg’s point, you have to look at all the data. And in this case, it’s the Shopping Frequency Index (SFI = total sample visit count ÷ unique observations) that stands out. This 80 K+ sq. ft. ShopRite is a large format unit – with a full liquor store, food bars, a pizzeria, a kosher butcher shop and a coffee shop – that generates an exceptionally high level of shopper loyalty and repeat visits:

Takeaway: Measure & Monitor Shopping Frequencies

Dashboards have us all fixated on rising and falling visit counts. They are important. But one of the interesting pre-Covid vs. Covid period metrics we’ve been tracking is a comparison of shopper visit/gains or losses (There weren’t near as many gainers as losers) versus our Shopping Frequency Index (SFI) shopper loyalty scores. Well-tenanted grocery-anchored centers tended to hold up well on both of those vital signs. BTW, Zip 9 targeting – especially into zips that have contracted pre-Covid to Covid – can be used to stimulate trip frequencies.

In studies across the country, we see traditional grocers in the 3.5 SFI range and specialty grocers like Sprouts and Trader Joe’s at 2 to 2.5. So ShopRite’s overall 5.2 SFI and its Saturday 6.1 – in this market – represent above-spectrum metrics that produce above-norm shopping frequency levels. Consequently, visit velocity (sample quantity visits per sq. ft.) are 22.1 for ShopRite vs. 15.7 for Wegmans.

With HH incomes in the $150’s,both units draw an above-market shopper. Ambient trade area average HH income is $129,000, and the MSA average is $115,000 (Again, to Gregg’s point, we are comping shopper-draw data to market data, i.e., what we’re getting vs. what’s out there.)

Final cautions from Gregg Katz: Be careful about turning people loose with self-serve data… and this: “Data by itself is just data. The differentiator is how you understand and apply the data.”

The Hunger Games analogy might be a little over the top. But just slightly.

Email David Lobaugh