Meal delivery: here to stay, but only with a healthy side of good unit economics

It's an interesting time to be in the food delivery industry. On the one hand, VC's are flooding food delivery startups such as Postmates, Doordash, GrubHub, and UberEats. A reported $3.5B has been invested in this industry, nearly triple what was spent in all of 2017, at very high valuations (WSJ). On the other hand, meal-kit delivery services have been struggling. Blue Apron, which at one time was set to IPO at a $3B valuation, slashed its IPO price by 40% and has seen its stock fall a further ~90%, to 1.12 at market close on October 25th. While Blue Apron has been the poster child for headwinds in the meal kit delivery market, others have struggled as well, including Chef'd, which ceased operations in July, and was then sold to a food consultancy firm which permanently shut down its e-commerce operations (WSJ).

These challenges have led many to question why the meal-kit delivery industry is so different, and so much more challenging, than food delivery, with some going so far as to call the former industry a "fad". My own personal view is that the trend towards buying meals is here to stay, but that many companies fulfilling this consumer need have not been mindful of what I would contend is the most important determinant of durable, long term business success: unit economics. A business can grow rapidly, and may even be profitable due to operational efficiencies, but if it has very weak customer retention, it will not be successful for very long. Conversely, another business may be growing more slowly, and may even be unprofitable, but if its customer retention and variable profitability are strong, and its customer acquisition cost is low, it could be very successful in the long run.

Motivating customer-based corporate valuation

Traditional financial valuation frameworks will not be able to differentiate companies such as the two described above without digging beneath the surface to uncover the underlying customer-level profitability dynamics at the firms. For investors to be able to uncover these truths, they need the right customer data and the right predictive models for customer behavior. It is precisely for these reasons that I have devoted most of my academic and non-academic energy (as an assistant professor of marketing at Emory University, and as a co-founder of Theta Equity Partners, respectively) on this topic, inventing what Wharton professor Peter Fader and I call "customer-based corporate valuation" (CBCV).

CBCV entails explicitly driving a traditional valuation model -- DCF, multiples, you name it -- off of the underlying behaviors of the target company's customers. At its most basic level, it exploits a very simple accounting identity: total sales in any period must come from customers who were either acquired during that period, or from customers acquired earlier who have remained with the firm. Those customers must place orders, each of which has an associated spend amount, some of which will flow through into variable profit for the firm. Because this decompositional accounting identity is true, it mathematically must be the case that customer acquisition volume, its associated customer acquisition cost, customer retention, order rate, basket size, and contribution inform traditional fundamental valuation models. A schematic visualizing this process is shown in the figure below:

This framework is valuable not only because it results in more accurate valuation forecasts, but also because it offers stakeholders of a company diagnostic measures for how that valuation forecast came about ("my valuation fell because it was a lot more expensive to acquire customers this year, even though the value of those customers remained the same after I acquired them"), a richer framework for peer benchmarking ("my retention curve is weaker than my main competitor's"), and a common language that all major stakeholders of a firm, from the CMO to the CFO and CEO, to the shareholders, can agree on and appreciate. This is a "Trojan horse" for the CMO, who traditionally is the "owner" of a firm's customer relationships. The CFO and the CEO gain more accurate insights into their stock price. And shareholders do too, as long as firms disclose a relatively meager set of customer disclosures on a semi-regular basis (SSRN). Bringing these stakeholders more in sync with one another has value in its own right.

A Tale of Two Worlds: Blue Apron versus eMeals

Let's bring this to life with a couple of real-world examples.

First, there is Blue Apron.

I had put together a detailed analysis of Blue Apron's unit economics here on LinkedIn almost a year and a half ago (LinkedIn), right after they had filed their form S-1 stating their intention to become a publicly traded company. The main conclusion I had reached was that even though they did not disclose any customer churn data, I could infer what it was most likely to be using a combination of models and a clever estimation procedure. My main conclusions were two-fold:

1. Blue Apron's retention curve was challenging, with over 70% of customers churning out after 6 months.
2. Blue Apron's CAC had been dramatically rising, to the point that they breaking even or losing money on their most recently acquired customers.

This is Blue Apron's retention curve:

This is Blue Apron's CAC relative to the value of those customers after acquisition:

This work ended up receiving widespread media attention, including most major financial news outlets (see my website for references). For those interested, I have also published an article in HBR and a case study available through HBS featuring this analysis. We all know the conclusion to this story.

Second, there is eMeals.

Some time after I had posted this article, I had received an email from Forrest Collier, the CEO of eMeals, who shared eMeals' customer metrics with me. The results were striking.

Before getting into the analysis, a little context on eMeals themselves. They are a privately held digital meal planning subscription. Subscribers pay $5 to $10 per month (depending upon how long they prepay for), and in return, get access to a wide variety of meals. After selecting the meals that you would like, you can have the ingredients going into those meals delivered straight to your door, or depending on the area that you live in, picked and made available in an Amazon Locker-type system at the grocery store itself. They offer these services through partnerships they have with Walmart, Kroger, Instacart, Shipt, and Amazon Fresh.

They don't have any warehouses and they don't do any of the cutting and chopping of ingredients themselves, eliminating that process completely from their cost structure. They are simply an intermediary, facilitating the budgeting and purchase process for meals with the grocery store directly. Because customers are buying ingredients directly from the grocery store, they get grocery store prices for those items, making the cost much lower for consumers (about half that of a traditional meal kit). In sum, customers get access a much wider variety of meal kits at half the cost, with much (but not all) of the convenience of the more traditional incumbents. In return, eMeals is a more scalable company, light on assets and labor, with an extremely high contribution margin (I estimate it to be 90%).

Holding aside any preconceived notions about what sort of business model should or should not work, let's let the data tell us the goodness of eMeals' unit economics, and how it compares to Blue Apron's. That is what I will do next.

Customer acquisition volume and cost

Many paying customers are coming in the door. Last quarter, gross customer acquisitions were up 63% over the comparable period the previous year:

It is not very expensive for eMeals to acquire these customers. Their historical CAC has fluctuated between $50 and $60 per customer, versus $100 and $120 (or even higher) at Blue Apron:

Customer retention

After customers have been acquired, they stay around for a very long period of time. Below is a chart of eMeals' customer retention curve, cohort by cohort:

We can see extremely high levels of customer retention before month 12, as many customers prepay for the year (~98% of customers are with the firm a month later, versus an estimated 45% at Blue Apron). However, even after their retention drops due to annual membership cancellations, their month 12 retention is nevertheless 43%, which is far higher than the 23% I had inferred for Blue Apron. Part of this is likely due to the low monthly price that they charge, allowing eMeals to "fly under the radar screen" relative to other subscription offerings that are more expensive on a per month basis.

Their customer retention has been very stable over time. Customers acquired more recently are showing nearly the same retention patterns as customers acquired further in the past:

Customer profitability

While eMeals' monthly fee is lower than Blue Apron's, it comes at a dramatically higher variable profit margin (90% at eMeals versus the high 20%'s at Blue Apron). This, coupled with the long customer lifetimes we see above, lead me to infer that eMeals customers are worth about $146 per customer after they have acquired. When we subtract off the $58 that eMeals is spending on average to bring those customers in the door, that leaves them with a healthy $88 profit per customer, or a 152% return on their customer acquisition spend.

In contrast, I estimate that a typical Blue Apron customer is worth about $133 after acquisition. When we subtract off the $132 that Blue Apron had spend over the six months prior to when my analysis was done, that left them with a meager $1 of profit per customer.

Summarizing all of these figures into one table, this is how Blue Apron and eMeals stack up relative to one another along key measures of unit economic health:

Future Growth Initiatives

Of course, neither company is standing still, raising the question of where we go from here. Blue Apron has made two major announcements.

1. The first is a pivot towards selling off of subscription in grocery stores. While this sounds very appealing, the devil is in the details. They are selling the kits at a 30% discount to their prices on subscription (BusinessInsider), creating pricing and retention pressure for their core business. Selling into grocery stores is also far more challenging logistically (especially for meal kits, which must be maintained at very specific temperatures and have short shelf lives), requires more working capital, and often requires the producer to bear more of the risk if the product does not sell.
2. The second is an initiative to sell meal kits on-demand through services such as GrubHub. While I do not typically gravitate towards consensus views, I would agree with most of the commentary on this move (e.g., CNBC) - if I am ordering through GrubHub, I want my food on the quicker side. I do not want to pay a relatively steep price for a rush order on a meal that I must then spend yet more time preparing.

eMeals' future growth opportunity is very different -- changing the default ingredients that are purchased when people "buy a meal." Take canned tomatoes, for example. When people are free to choose the brand of canned tomatoes they would like with their favorite pasta dish, they chose Hunt's 30% of the time. When Hunt's was set as the default option, customers chose them 64% of the time, a 113% increase:

It would be very easy for eMeals and Hunt's to set up some sort of an arrangement to share in this growth. And as long as the default brand is high quality, consumers are completely unaffected (and may even come out ahead). This change would add no additional logistical cost or complexity to eMeal's business -- it would be no more than a few lines of computer code -- implying any revenue they generate through this arrangement would be pure profit, and may actually take their contribution margin above 100%. This is a CBCV-friendly growth opportunity.

Wrapping up

In summary, it is an almost unambiguous fact that eMeals (1) acquires its customers at a much cheaper cost than Blue Apron, (2) holds on to those customers for a much longer period of time after they have been acquired, (3) brings in less revenue while they are alive but (4) retains a far greater proportion of each incremental revenue dollar as profit. The underlying fundamentals from a customer perspective are far superior, before we even consider the fact that eMeals new customer growth is north of 60% and their future growth opportunities seem more promising.

These results were very heartening for me. There is a real demand that consumers have for meals in lieu of ingredients, and as we can see above, there are firms like eMeals that are able to satisfy those demands while also operating financially sustainable business models. Let's not throw the baby out with the bath water in concluding that the whole pre-prepared meal delivery space is a fad. Instead, let's turn to methods like the one I'm presenting here to make sure we have a clear understanding of the underlying health of these firms. Let's ask companies to disclose the metrics we need to know to perform the analysis in the first place. Finally, to the extent that they could benefit from it, let's fund the companies with the sound business models so that they can grow and satisfy customer needs at a larger. This is a winning combination for everyone -- entrepreneurs, investors, and customers.

This post is based on a talk I gave at Groceryshop.

Full disclosure: I have received no payment, direct or indirect, from any company mentioned in this article.