Missions

We’ve spent some time in this newsletter talking about some of the unique advantages of ecommerce over brick-and-mortar—like my last post about “menu costs” and how it’s way easier to change things in ecomm.

But we’ve also talked about what brick-and-mortar gets right over ecommerce (like this one about how the layout of the Walgreens by my office is different from the layout of the Walgreens by my apartment). Since then, I learned a fun new term about why stores are set up in different ways: “Shopper Missions.” Our head of data, Andrew, told us about this term from back when he was consulting for retail stores. 

A “Shopper Mission” describes why the shopper came into the store today. A given store may serve customers that have a wide variety of missions, e.g., Target may have people with grocery missions; pharmacy missions; clothing missions; online pickup; etc. They can then think about how to lay out the store to try to optimally serve the most common missions. 

The brick-and-mortar stores can’t know anything about you before you walk in the door. But they can look at past shoppers and purchases and make assumptions about the possible (or likely) intent of that shopper who comes in the door, and anticipate what they may want and need. 

Contrast this to when we talk about “personalization” in ecommerce, it seems like the focus is on how we solve for the needs and specificity of this exact person. “How do we achieve 1:1 personalization and 100% predict exactly what they want?” The promise here is exciting and we all intuitively understand how powerful this could be. We know that there’s probably some way to predict this, if we can just link all our cookies and data sources together and get more powerful prediction models. It feels so close!

Despite that—and I talk to a lot of brands around this—I have seen effectively 0 brands doing personalization like this at scale. Usually when you dig in, people are using some algorithms around product recommendations, which are useful, but a far cry from this “1:1 personalization at scale” we all talk about. It’s just really, really hard to do, and you end up needing to create a TON of content and juggle rapidly-multiplying versions of your site, and most people don’t have the capacity to do it.¹

But brick-and-mortar is constrained, and there are real physical limitations that get in the way of 1:1 personalization. So instead, they’ve solved for the mission. And I think that looking at the problem this way could actually make personalization more approachable for ecommerce experiences, too. 

As mentioned above, I’ve previously relied on the example of Walgreens’s store layouts being different depending on where you’re at: 

The Walgreens in my residential neighborhood has the toiletry and pharmacy sections up front, serving a classic drug store purpose. Near my office in FiDi, they have all the grab-and-go food and drink options up front because most traffic is coming in on lunch break. If I need to pick up allergy medicine there, I’ll still walk to the back and find it. This store has a “personalization” that is based on a broad assumption/context (people in FiDi are looking for snacks) and likely works, and I find what I’m looking for nonetheless. 

Whole Foods is another example. Adam describes them as being ambidextrous: They meet multiple shopper missions with the same store layout. Whole Foods, for example, usually has an entrance by the produce and another entrance by the prepared foods. Two different entrances for two different missions (grocery shopping versus getting food for tonight).

Home Depot does the same thing. I’ve been there a lot recently, and the rental section is close to the main door, because you’re usually carrying out heavy stuff. But there’s also the “pro” section, which has a special parking and a dedicated entrance that leads you right to the plywood and raw building materials, and a garden center with its own entrance and checkout. Three different access points for three distinct missions.

Target does things differently than both for its large format stores (but it has small format stores for different locations) and runs with a very similar store layout throughout them. They don’t have as many entrances, but they keep departments in roughly the same configuration so as to reduce friction by making it easy for a customer to memorize the layout. 

The big takeaway here is that Shopper Missions are more approachable for ecommerce brands than the “1:1 personalization at scale” we all talk about:

• It’s easier to list out the main reasons someone is coming to your site than it is to list out all the various ways you might want to “personalize” an experience for a specific customer.

• Those reasons (“Shopper Missions”) are actually surmisable. You can gauge intent by the way they “enter” your site. Which ad did they click on? What email did they click through? Have they answered quiz questions previously? What section did they visit first after coming to your site?

• You have tools (that can help you service that intent.

• Testing will tell you which tools to use, which offers to use, which merchandising tactics to use. Navigation tests are popular for this reason (what is my customer trying to find?); Product information tests are valuable for this reason (do different ads require different product copy?); Testing various product recommendation models is helpful (is my customer most frequently looking to “complete the look” or something else?

The above is so doable today.

Plenty of brands are already tackling components of this and it’s a lot easier than chasing a white whale. It might be that we should, at some point, set our sites on “1:1 personalization at scale,” but if the choice is between orienting around Shopper Missions and doing nothing, it feels to me that the choice is obvious.

1  The ecommerce tactics graveyard is littered with promises of personalization that never materialized. We could spend an entire newsletter talking about all the reasons why.