Meal-kit company Blue Apron was the dream of two entrepreneurs in 2012. Within five years, it had a more than $2 billion valuation. It's come down hard since then, with a stock price over $140 in June 2017 dropping to the current sub-$9.

Plenty of things have gone wrong, like losing half its customers in the face of competition.

Turnover of customers isn't unusual and there are things you can do, but only if you have control of the financial aspects. And that apparently wasn't the case.

Time for two tweets from Adam Keesling I ran across that raised the important point.

Let's start with addressing three abbreviations for clarity. TAM is total available market, or how many people you would potentially reach. CAC is customer acquisition cost. LTV (found in a later tweet in the thread) is lifetime value. The interplay of the last numbers is the whole point of the discussion.

Also, the addressable market was never "everyone who ate food." That's a major slip on an important topic. The market you can address includes all people who could realistically become paying customers. Otherwise, it's like saying the addressable market for Ferrari is anyone who  buy a car, no matter their budget.

I said quibble because I think Keesling was likely exaggerating for effect. But where he gets down to business is on the customer acquisition cost. For those unfamiliar with the concept, any business should quantify the cost to acquire a new customer. It will likely vary by marketing campaign because of different campaign costs and prospect conversion rates. But you still want the overall averages.

Lifetime customer value is critical because it measures how much customers are financially worth to you in margin dollars, based on how much they buy, the types of products or services they choose (different gross margins), and how long they do business with you. This is an individual measure but, like customer acquisition costs, can be averaged across your customer base.

Customer acquisition costs and lifetime value come together in an equation: CAC has to be less than LTV, or else you're automatically losing money on a customer. Keesling thinks that's what happened with Blue Apron, as the S-1.

A few more quibbles. He pointed to a company-released CAC of $94. He thought it should be more like $460: 387,000 customers added in a year with $179 million in marketing. Divide the marketing spend by the number of customers. But that's not  right. You need to know the specific acquisition marketing costs, not general marketing costs. The latter includes customer retention, brand advertising, marketing to potential investors, and so on.

Still, he's on an important track. In the Blue Apron annual report for 2018, there was about $667.6 million in revenue with a cost of goods (including both product and fulfillment costs) of $433.5 million. As a result, gross margin was about 35%. For every dollar in revenue, 35% is left over after the cost of the making the product and delivering the customer.

Average quarterly revenue per customer varied from $233 to $252 in 2018, or $81.55 to $88.22 gross margin dollars. Not available was the average lifetime of a customer or how long the average customer stays.

In 2017, there were 1.04 million customers in the first calendar quarter and 746,000 by the fourth. In 2018, the first quarter started with 786,000 and the fourth, 557,000.

The company is losing customers and new ones aren't making up the loss. That's a problem when one of the risk factors for the business, according to the company, is to "retain existing customers and attract new customers." They're going in the wrong direction.

There's no way to know from these numbers the average TV or if the CAC (they call it CPC) of $94 is still valid. However, no matter what the exact number or how much more efficient Blue Apron makes its customer acquisition, it won't be enough because customers leave faster than they can come in. The CAC clearly isn't enough, and if it went high enough, the company might never make up the money it spent on bringing bodies through the door. Which was ultimately Keesling's point, quibbles aside.

You have to bring in customers. If you can't do that efficiently, you won't last long enough to fix things.