The COVID-19 pandemic has been an unexpected boom for online grocers. However, with first-time users representing 20 percent of the surge and out-of-stock challenges continuing to plague the industry, online grocers that want to hold on to the momentum post-pandemic face two challenges:
- “How can we ensure an excellent user experience?”
- “How can we deliver what customers order?”
The response to these questions first requires understanding shopper intent. Once you understand intent, you can align that with inventory and user experience. Here’s how:
Make Smart Substitutions
In-store, “what you see is what you get.” Once a shopper puts a box of breakfast cereal (or a pack of toilet paper) in their cart, it’s theirs. Online, an item on display might not really be available, might go out of stock while in a shopper’s cart, or might just vanish before delivery. That’s incredibly frustrating for shoppers, who will remember how your failure made them feel.
To reduce “out-of-stock shock,” first tweak your algorithm to display items in high supply. For items that truly are unavailable, let user intent guide your substitutions. How does a shopper plan to use those organic, free-range eggs that just ran out? Does her selection of ham and grated cheddar tell you that these eggs are for high-protein breakfasts? Then maybe turkey sausages would be a good substitution. Or does gluten-free flour and sugar-free chocolate suggest that she’s baking a healthy cake? In that case, offer a plant-based egg substitute. When you can’t deliver the original choice, grab the opportunity to show customers that you care.
Leverage the Best of In-Store and Online
The familiar pattern of physical grocery stores has its advantages — well-organized aisles, scores of product options, and strategically placed promotions. Roaming through categories picking up essentials while also discovering new products satisfies something deep in our "hunting and gathering" psychology. A large selection means every shopper is delighted, regardless of their intent.
Online, organizing products that way can be overwhelming. If Netflix made you search through all of its 12,000 programs without curation, you wouldn’t use it. But that’s exactly what we do in online grocery. What would happen if, instead of making users search through 25 pasta sauces, you curated a selection according to inventory and intent? None of us was looking for a show like "The Tiger King," and yet 20 percent of Netflix U.S. subscribers devoured it within 10 days of release. Online grocery can create this sense of curated discovery too.
Know What’s in Stock
All of this requires that you know your inventory in the first place, of course. Supply management systems designed for brick-and-mortar stores are doomed. Grocers need artificial intelligence (AI) that connects every local stockroom and that can dynamically predict the supply needed, district by district. Technology is evolving quickly to deliver such insights. Wakefern Food Corp., owner of ShopRite and PriceRite, for instance, is testing computer vision-aided inventory management from AI startup Focal Systems that “sees” what's on hand at all times and updates needs from shelf to supplier.
We're still in the “internet dial-up phase” of grocery technology, but it's already possible to promise highly personalized user experiences, and then to complete orders every time. The location of your competitors and customers is suddenly irrelevant. The game now isn’t just about a shift from bricks to clicks, it's about which retailers can best delight and deliver to all shoppers in any locality.
Henry Michaelson is co-founder, president and chief technology officer of Halla, a taste intelligence company that enables retailers to predict the personal preferences of their shoppers, all in real time.
Henry Michaelson is Co-Founder, President and CTO of Halla, a taste intelligence company that enables retailers to predict the personal preferences of their shoppers, all in real time.
While studying computer science at UC Berkeley, Henry co-founded Halla. He is responsible for constantly improving Halla's machine learning algorithm and for internal leadership, especially with respect to technology. Henry's previous projects include machine learning based classification of supernovae in the UC Berkeley Astrophysics department, a speaking role in the Warner Brothers blockbuster comedy Project X, a three year stint as lead guitarist for Joe Banks, and a patented algorithm that has distributed over $7M in awards to mobile gamers.