Funny you should ask this question. Because of a knee injury, I had to use Instacart for a couple of months. At first it seemed addictive, but it got really tiresome … Publix in particular was always missing several items I requested (Whole Foods actually did a far better job). It got to where I had to say no substitutes but I still kept getting them. I’ve given up on Instacart, and my knee is better … but this is a real problem and I don’t see it going away any time soon.

Product just turns too quickly for inventory records to stay accurate.

I think Amazon has the chance to bring a level of execution to online grocery shopping that doesn’t appear to be in place yet. I don’t want to judge an entire industry from my experience with Safeway last week, but it may be typical. While on vacation, I ordered groceries to stock up our rental house for a week. The order did not show up in the scheduled delivery window (in fact, Safeway was running three hours behind and I canceled the order) and would have been one-third short-shipped. What I thought would be a convenience turned out to be a customer service nightmare, after spending nearly an hour on hold to fix the problem.

Again, it’s a small sample size but the combination of stock-outs and late delivery does not exactly inspire confidence in the process. I believe Amazon has the capacity to make this work, and they won’t roll it out aggressively until they are ready.

As Paula’s experience underscores, the status quo is definitely not optimal. Customer-friendly substitution policies can help in the short-term, but retailers will need to continue to invest in better inventory management systems as demand for online grocery builds.

There is also the challenge of serving two masters with a single physical asset. Online grocers like FreshDirect and Amazon fulfill orders from fulfillment centers optimized for non-store operations, enabling better demand forecasting, more efficient order-picking and, ultimately, fresher perishables.

When online volume accounts for more than 10 percent of a store’s volume, it starts to have serious impacts on the in-store shopper’s customer experience and the overall store’s economics. Staying in-stock at the physical shelf is harder, labor management is harder and shoppers don’t like competing in the aisles with eager uniformed order-pickers.

As noted in the article, this potentially real differentiating factor for brick-and-mortar retailers over pure-play online grocery retailers does represent an Achilles’ heel. I am a strong proponent of BOPIS (Buy Online, Pickup in Store) primarily because research shows that the total market basket of BOPIS is larger than in-store only or online-only grocery shopping. Customers enter the store with empty baskets (collecting their groceries at the end of the visit). Empty baskets translate into the perception of more money to spend, particularly in those non-center store categories (think perimeter) that command high margins.

However, if the grocery retailer cannot execute the concept properly, it should probably not offer the service until it is flawless. This is not an area where the retailer can “roll and fix.” They need to get it right from the start.

E-grocery is reliable for dry and frozen foods, but produce is too ephemeral to be easily trackable via inventory management systems. After all, produce is a living thing and it’s tough to bag and tag a lifespan. Perhaps we need to change our expectations of online produce — making room to account for variables such as bruising and turning — instead of expecting it to function like dry goods.

Time from order to fulfillment and accuracy are directly proportional to inventory availability. Add to that the fact that a human fulfills these orders and you can expect errors and dissatisfied customers. It’s amazing to me how many issues we’ve discussed here over the years have lead back to inventory monitoring, availability, etc., and the number of problems and challenges that would go away if retailers and brands just knew where their stuff was at any point in time! It’s time for all products to become “connected.”

Why should we even expect in-store fulfillment to be successful? Fulfilling an online order out of a warehouse versus in a store are two different operations. Operationally they don’t match up. The focus of the jobs are different. The KPIs are different. The labor skills are different. In-store fulfillment only complicates inventory control in an area where stores aren’t terrific controlling it as it is.

The fulfillment problem is just the ignored tip of the iceberg. The entire brick-and-mortar system is built on PALLET logistics, delivering pallets of merchandise assembled in warehouses to stores where paid stockers put the stuff on shelves. Amazon’s entire logistics is built on ITEM delivery to individual shoppers, not to stores.

The brick-and-mortar retailer’s entire model is heavily dependent on UNPAID STOCKPICKERS, aka shoppers, to do that final “picking” and taking it to checkout for purchase. This is why I refer to brick-and-mortar retailers as merchant warehousemen. They do not SELL anything to anyone. They simply keep the WAREHOUSE, aka store, stocked, and other than stocking and COUNTING the money and sales, they are fairly ignorant of the actual mental/physical PROCESS by which shoppers sell to themselves.

It’s too complicated! 😉

Some superb observations on this thread – from both the retail operator and the “informed consumer” perspective. I’ve been arguing vehemently for some time that store-level inventory accuracy is the central discipline for success in supermarkets. Only a handful of chains have the skills in place to maintain a true store-level perpetual inventory in real time with computer-generated ordering. But the existence of that handful – among them Sobey’s, Wegmans, Price Chopper NY – is evidence that the challenge can be addressed with practical measures. Let’s face facts: In-store order fulfillment introduces an added layer of chaos. In a store with poor or marginal inventory accuracy, this can lead to even worse on-shelf availability. A digital ordering platform that’s not “aware” of actual item availability at the store where the order will be fulfilled is a virtual guarantee of missing or substituted items. Here’s where I beg to differ with Paula a little bit. Grocery inventories, which are re-order dominated, shouldn’t suffer from widening inaccuracy as turns increase. This is a major difference from fashion… Read more »

The concept of in-store fulfillment of online retailing is going to become, if it is not already, “table stakes.” If you choose not to be proficient in this area you may be at a competitive disadvantage. Accept it or not, a grocer — or any retailer — is competing with Amazon, who sets the bar for fulfillment.

I wonder if the best way to do in-store fulfillment is to maintain separate stock count for online in the store for high velocity items, kinda like airlines increasing flight time estimates to increase ontime departure. The nature of shelf inventory in stores means you can’t be 100% accurate all the time in the system. With online orders being critical for completeness, raiding store shelves ahead of in-store customers doesn’t help with store experiences either. Analytics should be able to ID the fastest turning items that are most often missing from e-commerce orders, and retailers need to hold stock accordingly.