The ultimate guide to VC’s LTV/CAC analysis

LTV/CAC for (two-sided) Marketplace — Pt 1

Gianluca Valentini
Sep 18, 2017 · 7 min read
I see the liiiiightttt !

“What’s your LTV-to-CAC ratio?”

Does it sound familiar? It’s probably one of the key questions any VC expects you to answer, when diving into your metrics. For those who are not familiar with the terms, LTV stays for Lifetime Value (of the customer, sometimes referred to as CLT) and CAC for Customer Acquisition Costs. The ratio indicates what’s your forward-looking profitability on the customers being acquired today. In line of principle (or common sense), the ratio should return a number >1. If it doesn’t, well..

The striking thing about the LTV/CAC analysis is that it sounds very simple but it’s a major decisional aspect (or discussion point) for venture capital investors. Just the mere understanding of it is a key factor in deciding whether to invest in a founding/management team or not.

Given its importance, I was surprised by how little knowledge is presented in practical examples out there. I came to the conclusion that those who know and use it in their businesses probably prefer to keep it private, while those who write articles on the topic (often) have a very theoretical understanding and/or have a second agenda (i.e. acquiring customers to consult).

Frustrated with this dichotomy, I have decided to step in and share what I have learned so far. The more I dove into that, the longer my article became, so I have decided to split it into a mini-series, which will be published on a bi-weekly basis. In the first part, I’ll cover the main concepts and highlight the difference between a traditional e-commerce and a marketplace model. In the second, I will look at LTV (namely, customer value and lifespan) when applied to a marketplace. In the last part, I will discuss CAC and how these are split in CAC for the demand and CAC for the supply side.

An E-Commerce is run and operated by a solo supplier. The range of products or the verticals served doesn’t matter: the operator of the platform sources and sells its own inventory and bear the risk associated with it. A Marketplace in common startup lingo refers to a platform that facilitates the matching between demand and supply, but it doesn’t take inventory risk. In other terms, the business is centered around facilitating supply&demand, rather then being the supply. The main differences with an E-Commerce is that a) the platform operator doesn’t take any inventory risk (although there might be some e-commerces that use a just-in approach), and b) the supply is provided by a multitude of parties, which are free to engage the demand in the way it suits demand (save the platform directives).

From a business model perspective, while an e-commerce operator retains the mark-up on the price of the goods&services sold, a marketplace one retains a % on the volume of transactions facilitated on its platform (referred to as GMV = Gross Merchandise Value or Gross Marketplace Volume). The % of the GMV retained by the marketplace is referred to as the Take Rate.

Quick example: Uber is a marketplace. On a ride fare of €20 (GMV), 80% (€16) goes to the driver aka the supplier, and 20% (€4) to Uber. 20% represents the take rate. The sum of all the retained % is the company’s net revenues. Beautiful!

Now, what you’ll usually find by simply googling ltv/cac analysis is mainly applicable to an e-commerce model. There are plenty of articles explaining the elements and different formulas to calculate it, while the purpose of this mini-series is applying the theory to a more challenging marketplace. But a quick recap won’t hurt. So here it is.

The Lifetime value is composed of:

a) AVG revenues per customer during a timeframe (net of variable costs)

(X) multiplied by (X)

b) Customer Lifespan (that is, how long the customer remains a customer/returns)

In some cases a) can be broken down into micro-intervals. If your e-commerce sells household appliances, a) can be broken down into:

a) = (AVG basket/purchase value) * (# of times a customers makes a purchase e.g. in 12 month)

My suggestion is to focus on a set timeframe (e.g. 12 months), take the number of UNIQUE customers and total value of the baskets/purchases during the same period. The total divided by the number of unique customers will give you the AVG revenues per customer. The variable costs can be calculated as a % of the revenues, based on the type of goods&services being provided.

Then there’s the lifespan. Oh, the lifespan. Lifespan my lifespan. This is such a tricky measure that often investors simply skip it and use assumptions (e.g. 1 and 3 years). HOWEVER — don’t stop reading you lazy — it’s very, very important than any founding team at least attempts to calculate it and shows that it has a good understanding of it.

The lifespan can be calculated either historically or forward-looking. The former is implicitly better, but it’s not available for young startup companies, which have to rely on the latter.

In both cases, the key aspect to track here is the moment when a customer churns. In the historical analysis, a company can calculate the AVG lifespan of all customers who, at certain date, had churned. In the forward-looking approach, the lifespan is calculated as 1/churn rate. If the churn rate is calculated for e.g. quarters, make sure to divide the resulting number by e.g. 4 in order to get the lifespan in years.

In the coming articles, we will get deeper into this topic but the principle is: if after 1 year 25% of your customers churn, it will take another 3 years before all of them are completely gone. This is why it’s forward-looking: it’s based on the assumption that the churn won’t accelerate nor slow down.

The Customer Acquisition Cost is the sum of all acquisition-related costs divided by the number of acquired customers.

Acquisitions costs are:

a) Cost Per Click (CPC): any advertising being on search engines, social media platforms, retargeting.. you name it.

b) Customer support personnel: any manual labour oriented at helping the customer in the last steps of the funnel. In some companies these might be account managers. Usually the cost of managers/head of customer support departments are considered overhead. More on that in the coming articles.

c) Discounts and/or promotion for referrals: these suck up profitability and should be accounted for as a further CAC

d) Any other cost that is strictly connected at the conversion of a customer: it depends on the business case.

Acquisitions costs might be (in full or partially):

a) Offline marketing and other sales-oriented promotions: some might argue these costs are more for branding, thus with long-term implications and therefore not connected to conversion. My suggestion is to allocate at least part of them to the CAC.

b) Marketing personnel: same considerations as before. If some people are specifically dedicated to CPC, then it could be argued that their wage should be part of the equation.

As you can see, we are already starting to enter the grey area of the LTV/CAC, and more is coming.

For an e-commerce the equation is quite straightforward: the CAC is “simply” the cost of acquiring a customer. But for a marketplace, the operator shall acquire the final customer AND the supplier, thus running into CAC’s for both sides of the table. Things get particularly interesting because in most cases demand and supply follow totally different patterns and require a fully-dedicated set of actions and costs and the question becomes how to combine them in a way that the LTV/Total CAC is consistent. This can be quite tricky as demand and supply run into Acquisition Costs at different point in times, sometimes at intervals, but they both “convert” when they meet.

It’s hardly the case that a supplier serves only one customer. Usually a supplier (once acquired) contributes to several “conversion moments” throughout its lifespan. Likewise, a customer might have purchase only once (and churn) or purchase multiple, irregular times over its lifespan. Rarely the purchasing behaviour follows a regular pattern (that’s the beauty of subscription models). Combining these two in a way that the analysis makes sense requires the use of averages conversions on both ends. This means:

  • AVG sales per supplier over lifespan
  • AVG purchases per customer over lifespan

Let’s assume that the lifespan is similar for demand and supply, e.g. 2 years. In this period a customer purchases on AVG 4 times. In the same period, a supplier performs an AVG of 10 sales.

It’s easy to see how, on AVG, a supplier serves 10/4 = 2.5 customers during its lifespan. This means that to combine the LTV with the CAC we need to account for:

1 x Supply CAC (per supplier) + 2.5 x Demand CAC (per customer)

As you can see, the differences in conversion frequencies and lifespans can make our simple LTV/CAC ratio very, very complicated. In the coming articles I will get get my hands dirty and show how this can be put into practice. There are many more little details that should be considered but it’s hard without a case with numbers. I plan to first look into LTV and, specifically, the lifespan, and then dive into the Demand and Supply CAC and how to combine them to obtain a Total CAC of the assumed lifespan.

If you found this article useful, don’t be selfish and share it with other entrepreneurs out there. ;-)

Gringotts Ventures

An initiative of @LuckyValentini — Entrepreneur and finance…

Gringotts Ventures

An initiative of @LuckyValentini — Entrepreneur and finance geek. I write about startup life, trends in tech, venture capital or any other interesting topics I happen to stumble across. Views are my own, prophecies are wizardry.

Gianluca Valentini

Written by

VP Marketplace Ops @ | Venture Advisor @ EQT Ventures | Former founder (exit) | Angel investor @

Gringotts Ventures

An initiative of @LuckyValentini — Entrepreneur and finance geek. I write about startup life, trends in tech, venture capital or any other interesting topics I happen to stumble across. Views are my own, prophecies are wizardry.