Imperfect Produce is a Bruised Business

There are few industries which scare venture capitalists as much as grocery. I’m not surprised. Almost after a decade of food retail experiments, few investors are willing to invest in the business due to a reasons, including high procurement and processing costs, perishable inventory, variability in supply, a haphazard regulatory climate, unbearable tight margins, and the uncontrollable temperment of Mother Nature. No wonder investors stay clear of their fruits and vegetables.

It also doesn’t help that many investors still remember the dot-com hubris of Webvan, a spectacular failure which burned through more than $1.2 billion in two years. Though the climate has changed a bit and a growing number of investors have invested in a new cropping of food+tech startups in the e-commerce, on-demand delivery space.

One of those companies is Imperfect Produce, which sells the slightly blemished, surplus harvest of farmers to over 7,5000 retail and consumer buyers. The question is whether the business model is innovative or, itself, imperfect. The purpose of this essay is to prove that Imperfect Produce and many other food+tech startups are interesting, but not innovative, and should not receive venture capital investment.

The layout of the essay is as follows:

  1. Why customers don’t care about how fast you can get (im)perfect produce to their door
  2. How Imperfect Produce is focused on the least important parts of the food value chain
  3. What e-commerce grocery startups can do to win customers and retain margins

It seems counterintuitive to claim that people don’t care about how fast you can get produce to their door when a slew of companies are competing on this very metric. Amazon Fresh promises 2-day delivery, Instacart deliveries are within an hour, and Postmates Fresh groceries in 30 minutes or less. This milieu makes Imperfect Produce 7-day delivery time, what they call “assigned delivery,” seem odd especially because it is not in the meal-delivery business, like Seamless and Caviar.

It is important to note that Imperfect Produce is meeting the needs of the urban, educated, 40-year-old mother of two that plans her family meals a week in advance. It is not going after the college-student with a late-night craving for Thai food or the last-minute shopper who forgot a few ingredients on his shopping list and would rather not go back to the grocery store. Regardless, from the consumers perspective, short delivery times are an expectation, not a differentiator. Customers do not base order decisions on delivery times because the grocery industry continues to compete away the cost as an expensive, profit-curbing necessity in a new competitive climate. The shift of costs from retailer to customer that has occured over the past few years cannot be overstated (even if it has become business as usual).

Let’s break down the food value chain in order to better understand the “necessity-to-expectation” shift. The value chain includes: 1. ingredients, like carrots; 2. production, or turning carrots into carrot juice; 3. delivery, which is the vehicle and/or person that gets the carrot juice to your door or the grocery store; and, 4. brand/distribution, which is the platform and increasingly the marketplace which consumers are engaging. Below is a chart that lays out each component of the food value chain.

Food Value Chain

Ingredients

Production

Delivery

Brand/Distribution

We could clarify the components of the value chain by providing startup examples that best represent innovation in each component. For ingredients think Memphis Meats or Ripple; for production think Abundant Robotics and Momentum Machines; for delivery think Marble and Flirtey; for brand/distribution think CommonSense Robotics and MealPal. But not all components of the value chain are created equal. While some components are easily defensible, have inherent network effects, and allow startups to quickly scale, other components require a lot of work, for little to no returns in ensuring high margins and extreme scale.

Most Valuable Components of the Food Value Chain

Most valuable is defined as the component(s) where incumbents struggle to innovate due to how hard it is to build and scale. The most valuable components of the value chain are highlighted.

Ingredients

Production

Delivery

Brand/Distribution

Production and brand/distribution are the most value components of the food value chain. This has always been the case in the food business but it may not be immediately obvious so let’s unpack each component, starting with ingredients. And here we’re talking about raw materials: carrots, apples, milk, chickens, etc. The question we should be asking ourselves as we think about this category is: “what would it mean to innovate in the creation and sourcing of food?” For some, like Uma Valeti and Nicholas Genovese at Memphis Meats, this means developing meat directly from animal cells without the need to raise and slaughter livestock; for others, like Neil Renninger and Adam Lowry of Ripple Foods, this means using pea-protein as a an alternative form of diary. Imperfect Produce is primarily innovating in this component of the value chain as well. Ron Clark, Benjamin Chesler, Ben Simon this means salvaging cosmetically-challenged produce that would otherwise be thrown away.

The problem with innovating in the ingredients component of the food value chain is that is requires heavy investment in changing consumer behavior: people need to stomach meat made in petri dishes, be fine drinking pea-milk, and to overlook aesthetic flaws in their produce. And I actually, think the latter is the most difficult. Even when rifling through a bin of lower quality produce, consumers consumers still spend time selecting the most aesthetically-pleasing of the bunch. One could say that we have been nurtured to prefer perfection, but I think it's partly nature too: we feast with our eyes first. Besides, even if these companies were able to change societal norms around food consumption, it does not lead to a defensible moat. A slew of copycat companies (Minnesota Meats, Rigatoni Foods, and Inferior Produce) would emerge free-riding on the hard work of early entrants in the space.

From plow shares to combines, innovation in food production has increased the abundance of food, made it cheaper relative to income, and freed people from tiring work to pursue other trades and talents. The classic historical example is the agricultural revolution. In the 1800s, 80 percent of the U.S. labor force worked on farms. Today it’s 2 percent. The farm is not the only place where mechanization is leading to greater abundance, lower prices, and increased freedom. Zume Pizza is using robots to cook pizzas, Momentum Machines: to make burgers; and, Abundant Robotics: to harvest apples. Notice that these innovations are all using technology to automate human labor which increases efficiency and reduces operating costs. The typical pizza restaurant might spend 25 percent on workers, whereas Zume pizza, in part by robots, only spends 9 percent. Companies innovating in production can spend more money on increasing quality of their product at lower prices for the customer.

We started the conversation talking about delivery. Remember that this component of the value chain is an expectation from the customer’s perspective and an expensive necessity for companies. It probably goes without saying that any business placing band-aids on hemorrhaging costs is not innovating, but struggling to survive which probably means they have not built a superior product, do not own the end-user relationship, and are still trying to find product-market fit in an increasingly competitive market. It is near impossible to build and scale any food+tech startup innovating on delivery. The unit economics just doesn’t add up. Selling groceries online means taking on additional costs — in delivery vehicles, drivers, and other ancillary items — that are higher than the fees customers are willing to pay for delivery. The average cost of mixed produce being by Imperfect Produce is $15. That means the delivery charge that a customer is willing to pay will have to be less than 50 percent the cost of goods sold for the convenience of delivery, especially for low price point items. Since grocery products have an average price of around $3 at a 30 percent gross margin, a business is left with only $0.90 to pay for all handling, selling, and delivery. Profitability at these margins is near unattainable, and made worse by seasonal variations in supply and demand.

The last component of the food value chain is brand/distribution. It is one of the most valuable and unique in that it is also the “access layer.” As I have argued in other posts, the “access layer” is the component(s) of the value chain which most directly correlates to ownership of the end-user relationship. Innovation at the “access layer” determines the course of an industry. It is where you will find the type of consolidation that most directly impacts consumer experience. It is how a startup can acquire traction amongst early-adopters and thereafter quickly gain market share as it vertically integrates. It is when incumbents begin to waste precious time trying to determine if their systems and platforms can be adapted to compete -- unaware, that for many of them, it is too late.

The “Access Layer” of the  Food Value Chain

The Access Layer is defined as the component(s) of the value chain which most directly correlates to ownership of the end-user relationship. The “access layer” is highlighted.

Ingredients

Production

Delivery

Brand/Distribution

This last component of the value chain is a bit difficult to comprehend because it is part aesthetic; for example, the wanderlust of walking through a farmer’s market; and, part process, like bartering down the price of arguelles because you only have $2 on hand. In other words, when you think about brand/distribution, forget the produce, and think platform. It is the platform that creates a process that engages customers in an aesthetic that drives customer retention. Remember the examples of CommonSense Robotics and Mealpal? Both of these companies have similar characteristics: 1. liabilities are being turned into assets; 2. there is widely distributed supply; and, 3. technology is leveraged to streamline or automate human labor. CommonSense Robotics is building micro-fulfilment centers (1) that can support high-capacity on-demand operations because their inventory is distributed throughout the county (2) and managed by a complex system of robots and artificial intelligence (3). Mealpal is building an application (3) that connects restaurants (2) that are too small to market their foods with customers that are too busy to wait in line but don’t want to eat the same thing for lunch (and now dinner) everyday (1). Notice the benefits of each of these characteristics. Liabilities turned into assets is capital efficient: existing infrastructure obviates the need for physical capital investment. A widely distributed supply is operationally efficient: the fulfillment of a product or service requires the enterprise to only use a subset of resources, no longer taxing the entire system. Leveraging technology to replace humans is labor efficient: increased automation reduces cost, a dividend that can be passed on to creating a better product or experience for the customer.

Imperfect Produce has none of the potential advantages of an e-commerce grocery startup because it is focused on the wrong parts of the food value chain. As a result, it will eat all the associated costs. All hope is not lost. The company could still pivot their business strategy to focus on production and brand/distribution. For example, imagine if you could scan every slightly-blemished fruit and vegetable that entered into your warehouse and then sell the produce at grocery stores? Imperfect Produce could leverage this scanning technology to help farmers (at a fee) to maintain ownership of their produce until it was scanned at checkout and to provide fast and accurate information about what is selling and where. Retailers could also be charged for the service of being connected to farmers in which they don’t have to pre-purchase a contracted amount of produce.

Consumers wouldn’t have to make grocery decisions a week in advance and would still retain the ability to make purchase decisions based on their price sensitivity. As things currently stand, I predict that Imperfect Produce will not change consumer behavior for blemished goods; if anything, it will aid the steep price competition already underway for “perfect” produce. For these reasons, Imperfect Produce is a bruised business.


Thanks to Nikhil Basu Trivedi and Chelsea Miller for reading early drafts of this essay.