Instacart: The Digital Grocery Revolution

With digital platforms and big data insights, Instacart is alleviated the cumbersome experience of grocery shopping in urban areas.

The digital grocery industry:

Though we typically do not give it much thought, the traditional grocery shopping pathway is wrought with friction. First, a customer decides what to purchase. Next, the customer physically goes to the store, selects the items, waits in the checkout line, and brings the groceries home. For those frequenting urban stores, this can be a harrowing experience of bumping between people in aisle and waiting over 20 minutes in the checkout line (Figure 1). And finally, the customer uses the ingredients to make a meal. In today’s world, we often don’t reach this final step.

Whole Foods Union Sq NYC
Figure 1: Whole Foods Union Sq NYC. Source: Business Insider

Many firms are attempting to alleviate friction various portions of this consumer pathway1 (Figure 2). Restaurant delivery services (e.g., Seamless) are eliminating the entire pathway by delivering entire meals from restaurants through their online and app platforms. Blue Apron eases the selection and purchasing process by allowing a member to choose online recipes and shipping all the necessary ingredients. Fresh Direct allows consumers to view its warehouse inventory through an online portal, select a grocery list, and have it delivered to their doorstep.

Figure 2

Among the increasingly fragmented digital grocery market, Instacart requires the smallest change in consumer behavior. Instacart enables consumers to shop their local grocery stores (e.g., Whole Foods, Costco, small local chains) through Instacart’s online and phone app portals. Unlike Fresh Direct or Blue Apron, Instacart is able to work with its retail partners to mutually boost revenues.

Instacart has three primary revenue streams2: delivery fees from consumers, volume-related fees from retail partners, and promotion-related fees from CPG companies. Consumers order from Instacart’s web or phone app platforms by entering a zip code and selecting their preferred retailer among the options that Instacart populates. The platform then allows customers to choose grocery items among the thousands of SKUs from the retail partners’ inventory. Because it partners with local stores, Instacart is able to serve its customers in under 60 minutes.

An operationally-intensive service:

While the consumer interface is fully digital, Instacart’s operations are human capital-intensive. Once the consumer places an order, the shopper begins collecting the requested items, using his/her phone to scan each item as it is selected. If a requested item is out of stock, the Shopper will request the customer to approve a replacement item. The Shopper then proceeds through the check-out line to pay for the order. The orders are placed in a ‘staging’ area; hot and cold items are temperature controlled until delivery. An Instacart Delivery employee then delivers the items to the customer’s destination.

Gaining a competitive advantage through big data:

Instacart recognizes that much of its competitive advantage lies in the data that comprises its $100M annual revenues. Over the past several years, Instacart has built a team of 10 data scientists to analyze its customer, shopper, retail partner, and CPG partner data to boost revenues and minimize costs3.

  • Instacart provides its CPG and Retail partners with data that is unavailable in a traditional grocery model. For example, Instacart orders indicate ‘true’ demand for a given item, even when it is out of stock. Similarly, Instacart’s data indicate consumer preferences for replacement items when existing items are out of stock.
  • Data analytics is also enabling Instacart to bring down costs: e.g., mapping out store layouts to optimize the Shopper pathway, batching orders and optimizing delivery routes to minimize Delivery time, and using data analytics to better predict delivery windows and staffing levels.

While they have not publicly revealed all their trade secrets, I imagine the data science team has identified far more data opportunities than the 10-person team can address. There are, however, two key opportunities that I recommend Instacart’s team prioritize if it is are not already doing so:

  • Instacart already has plans to compete directly with Blue Apron and Plated through a partnership Food Network to offer 500,000+ recipe ingredients for purchase4. I recommend that Instacart use its customer-specific data to offer a curated recipe list by customer, making the browsing process more applicable and user-friendly. Additionally, this will increase consumer stickiness as they will be more likely to return to a platform that caters to their preferences.
  • Instacart has the unique position of viewing how a given customer splits his/her basket across retailers (e.g., what they purchase at Whole Foods vs. Costco). I recommend they sell this compelling data to its retail partners. This will likely pose conflicts of interest between its retail partners, however, so Instacart should be thoughtful in how to execute on this incredibly valuable set of insights.

Though it may feel unnerving that businesses increasingly mine personal data, this practice has already permeated our lives. If Instacart can channel the hunt for data insights to improve our grocery shopping experience, I’m all for it.

Word Count: 792

  1. Data Fox, “7 Instacart Competitors Offering Grocery, Meal, and Recipe Delivery Services”, November 15 2016
  2. Fortune, “Instacart Is… Generating Profits?”, March 24, 2016
  3. Instacart Tech Blog, “Instacart Spotlight: Data & Analytics”, May 17, 2016
  4. Digital Trends, “Can Instacart compete with Blue Apron now that it’s partnering with the Food Network?”, June 2, 2016


Burberry is Bringing Tech-xy Back


ESPN: The World Worldwide Leader in (the Digitization of) Sports (Media)?

Student comments on Instacart: The Digital Grocery Revolution

  1. Great post Priya. I have a sense that Instacart’s data can be misleading in a sense given the structure of their business and the kind of customers they serve. In fact, I know of a lot of people here on campus bundling Instacart orders to get better pricing on delivery and tip as it can be split among several people. Therefore, using Instacart data as being representative of the average consumer at a grocery store can be misleading, especially when it comes to forecasting demand and rearranging store layout using data analytics. They seem to be serving mostly student populations or people without cars, and given the premium pricing on their service people living nearby in an apartment building will often bundle orders.

    These same customers are also very price sensitive, and exhibit very little brand loyalty towards Instacart. It seems to me as if a model like Instacart is a temporary solution to a problem, in a time where services such as AmazonFresh are becoming increasingly available at lower prices while removing the need for tips to a shopper. Despite the useful data Instacart provides to retailers, I would be concerned with their ability to grow and be useful to their partners in the long term in a market where AmazonFresh and other equivalent services will gain scale and reduce pricing, only to capture more and more of the price sensitive segments of the online food market.

  2. This is a great example of how the digital transformation is touching on every aspect of our lives, including mundane tasks like buying groceries. While I conceptually love the idea of Instacart (especially because it requires a small change in consumer behavior which is key in new product diffusion), I have concerns about the operating model. Given that the operating model is extremely human capital intensive (picking out groceries, taking them through checkout, storing & delivering) I am concerned about its ability to scale in a cost effective manner- will prices increase and will service times decrease? On a separate note, I did love your idea about creating a curated recipe list based on historic data collected on the consumer. I think this nicely taps into another trend we are seeing, customization, which in this context could add a ton of value to the end consumer.

  3. I just registered at Instacart but I haven’t used it yet. For me its business model makes a lot of senses, as you mentioned, helping consumers trade off their precious time with money, providing extra income for the buyers, and offering invaluable data for retailers. However, there’re certain group of people, like me, just don’t enjoy losing the chance of walking around and picking up goods in supermarket every week. For me it is a great leisure time once a week, getting away from the real world and just concentrating on life. Also, per your comment of selling data to retail partners, I have a concern of data privacy issue. If trading personal consumption data really respects consumers’ privacy? And, retailers won’t be happy to share with their competitors their own data. So my recommendation will be that Instacart could provide data analysis service for retailers, instead of directly selling data. By providing data analysis and consulting service, retailers would understand their consumers better through Instacart’s data without hurting consumers’ privacy

  4. Really interesting article! I’m a customer of Instacart so I can personally attest to the value of the service – particularly for those living in a food desert! In my opinion, Instacart, Blue Apron, Seamless etc. are all luxury services because they add significant cost to the transaction. Generally, we see technology a a way to lower costs to the consumer. Do you foresee anyway that services such as Instacart can deliver the same value (in terms of lowering friction) without passing on extra costs to the consumer? One idea is to shift the additional costs to the CPG firms/retailers as they are benefiting from the data that Instacart is collecting.

  5. Thanks for sharing! As you pointed out, this business will require very smart operational planning, but I completely see the value it brings customers and retailers. The move to partnering with Food Network for recipes makes me think about how they can reduce variability in customer orders by providing recommendations. As they play around with the impact these suggestions have on customer orders, they can get smarter about how to predict demand for certain products and from where.

    Another thought I’m curious about is whether there’s a specific customer demographic that will want this, or do you see this as a mass market service? The most obvious target demographic would be busy working professionals who don’t have time to go to the supermarket after work, but are there people who enjoy the process of picking their own groceries? For example, I imagine that higher end restaurants try to maintain a competitive advantage by sourcing the best ingredients possible.

  6. Great article, Priya! While I agree that Instacart requires the smallest change in consumer behavior, I still have not used it. I’m sure once we get into winter Instacart will be my bff. I wonder if there are other possible partnerships, such as within healthcare, Instacart can enter into given this tremendous information on a consumer’s diet. What if the food you order on Instacart is preloaded into your Apple health app, which you can better track your nutrition. Even further, nutrition apps can recommend a shopping basket (and recipes) given your desired health or weigh loss goals. I think health/nutrition is an avenue Instacart should explore.

  7. Thanks Priya. Wholefoods invested in Instacart earlier this year. Has this created problems with Instacart’s other grocery partners? At some Whole Foods stores, Instacart has embedded shoppers with devoted checkout lanes and who then pass off orders to drivers. The investment from Wholefoods was a boost of confidence in the service which has been struggling to become profitable. They raised delivery fees and annual membership costs, and lowered pay for the shoppers. As you mentioned, its still a human capital-intensive business model, and according to the Economist, Instacart was “rumoured to lose around $10 on each order it fulfills.” I don’t see scale or a network effect fixing this model.

  8. great post and very very interesting Priya! I started using Instacart and although it was a big big struggle for me at first but I must agree that it requires the least change in customer behaviour. But there is one aspect that I feel requires a change in perception (a little like what we were discussing in the marketing pricing module as to what people perceive as fair) and that I feel it is a big challenge for the company. Their rise in delivery prices has already made the headlines ( ) and although I think what they are doing with the $150 year free delivery plan is much like what Amazon Prime does, there is still that tip part of the cost that does not sound right for the grocery shopping model. Any thoughts on how the company is or should be addressing this ?

Leave a comment