Strong Businesses and Lifetime Value
Building and evaluating companies for long-term strength
Lifetime Value or LTV sits at the crux of the startup world. It’s a term that gets thrown around a lot, but is often misunderstood. This is evidenced by the desire to calculate LTV using a very simple formula (i.e., with precision) without understanding the underlying drivers that will sustain it over longer periods of time (i.e., with accuracy).
The typical LTV calculation usually looks something like this:
This method is based on the annuity formula for financial instruments. In very few circumstances does this formula yield anything remotely useful - both for investors trying to evaluate a business and more importantly for the operators themselves. Companies and their customers typically do not behave anything like engineered financial instruments. As a result, there is so much lacking in this LTV formula that I don’t even know where to start, but the key problems are 1) at the early stages (and even into the mid stages) the data simply is not good enough to reliably establish these metrics, and 2) most people have no idea where it comes from or how it works and simply apply it blindly.
Which leads to another, more informative method:
This method is actually the same… but in the previous over-simplified case you are assuming that n = infinity. (Assuming anything equals infinity in a practical business setting is obviously questionable.) Using this sum formula with the requisite understanding of customer size and behavior can yield much more accurate results, but what it takes to really understand customer behavior is up for debate. In addition, you can make a simple assumption like 5 years for n, but why choose 5 over 3 or 10 years? If the company has only been around for 2 years, it’s very difficult to predict and quantify the customer lifetime.
That is why I came up with a third option… the customer half life formula:
Here again I use the same formula, but instead of making an assumption for n, I use the half life formula to understand the probability of maintaining a customer based on historical customer behavior. The theory behind this is what once the probability of a customer churn goes above 50%, we can no longer count on retaining that customer, so we end the customer lifetime there.
This results are finite numbers for n based on measurable annual churn numbers:
5% churn = 13.5 year half life
10% churn = 6.6 year half life
20% churn = 3.1 year half life
But we run into the same problem: in the early days of a company without much data on customer behavior over time, it is very difficult to nail this churn number which now has even more impact on the LTV calculation. If you sell monthly contracts and have a large number of customers with industry-standard behavior, this can work (i.e., consumer cybersecurity), but if you sell annual or multi-year enterprise contracts (i.e., accounting software), then churn means something entirely different and it takes years to understand this dynamic. In either case, customers are retained rather than churned - this is the essence of a subscription. The impact of upsells, usage-based billing and multi-year annual contracts throw another wrench into the calculation. Whereas a company like Spotify or Netflix with a simple monthly subscription mechanism and tons of data can probably use these simplified calculations, as soon as the math gets a bit more complicated, accuracy suffers immensely.
So we’re back to square one. And this is just the top half of the famous LTV/CAC ratio that people discuss and calculate constantly. The good news is that CAC is easier to measure, but it can be harder to calculate and predict as you move through the phases of product adoption.
For later stage companies with more data, it benefits both evaluating investors and operators to try to improve precision and thereby improve accuracy of these metrics. The result: powerful diagnostics which can help assess the viability of the business longer term as well as improve operations and resource allocation in the short term. Defining the right metrics and capturing them from an early stage make this exercise much easier down the road.
But for early stage, we are still at a loss. Trying to do calculations with just a handful of customers is not useful, but understanding the concepts behind logarithmic decay and net present value is helpful in defining metrics to drive business success in the future.
So how do we apply these very mathematical concepts to nuanced and nebulous situations to understand business model opportunities at the early stage?
The answer is to think about what we can know at the early stage, then see if any of those data points map back to the fundamentals we were trying to calculate above.
I have identified three key product-market characteristics that drive lifetime value even before launching a product:
Average Contract Value (ACV)
Average contract values need to be high enough to create interesting opportunities
High ACVs mean more margin to play with when it comes to fixed and variable costs. Mistakes and experiments are more easily forgiven.
With a lot of outstanding questions, knowing that there is pricing leverage due to market demand and value de-risks the entire situation.
Pricing is also an engagement method: for most enterprise and consumer buyers, the more they’re willing to spend, the more they care, the more they’re willing to invest in making the product work for them, resulting in longer customer lifetimes and higher LTVs.
Adoption Pattern / GTM Strategies
With the proliferation of digital marketing and modern sales tools, there are very few industries where a new company can sell efficiently at scale. The market for eyeballs and attention is priced to perfection, so any sector that is not yet efficient will quickly head in that direction once the word gets out that there is a new opportunity.
Figuring out how to create advantages in customer acquisition is critical.
For larger companies, this usually means branding and partnerships.
In certain parts of the enterprise software stack, there are well-trodden sales motions, but it’s almost always a race to the bottom before long and one competitive player with a loose budget can drown everyone in the market.
The brute force method can be an effective way to get started, (i.e., hiring well-trained sales reps who can get the ball rolling right away), but it is likely going to become a challenge without access to zero-interest-rate levels of venture dollars.
The solution:
Markets with group-think buying behaviors
Certain verticals tend to hire the same consultants, go to the same conferences, buy the same tools, etc. If you’re the first software player in that space, then there’s a good chance you could be the great beneficiary of this buying behavior if you can snag an influencing customer early.
Virality
Certain product-market combinations encourage initial adopters to bring along their friends. The product and the market must be cleverly engineered to make this happen, and it is possible for both enterprise and consumer.
Network Effects
This may be the single most powerful adoption dynamic in venture capital history. This is how Facebook got where it is, but it’s also how Bill.com has developed so impressively. Finding a market where you can build a product that can unlock network effects is not easy, but when it happens it is pure magic.
Retention Dynamics
The longer a company retains its customers, the more money it makes. Even though cash now is worth more than cash later, it still makes a big difference to maintain customers for 10 years instead of 1 year.
Finding markets and products where multi-year contract and enterprise buying behavior are evident is great.
Pairing that with variable customer adoption behavior is a super power.
There are two distinct approaches to retention, which are often intermingled: economic (i.e., revenue) and cash.
Economic retention is what is in the contract: is it monthly, annual, multi-year? Is there auto-renew?
Cash is king. An annual contract that gets billed monthly is effectively a monthly contract. Getting the cash upfront makes a big difference when it comes to retention.
Cost of capital is incredibly high for early stage companies, getting free net working capital loans from your customers in the form of upfront cash payments is a huge unlock for growth.
Upside alignment with customers is huge opportunity.
No company has done this better than Shopify:
Lock in a subscription tool that enables more sales for your customers
Take a piece of the transaction value
This results in a highly predictable revenue stream with a more volatile, but customer-aligned upside revenue stream.
Executing the land-and-expand well is incredible, but it’s often much harder than it seems.
This exercise takes me back to engineering classes in college which allowed us to have an “equation sheet” for exams. They usually looked something like this, as we tried to cram as much as possible onto a notecard to help us during the exam. The trick was always that making the equation sheet was more useful than having it during the exam. Understanding the meaning behind the variables and how to apply the equations is more important than the equations themselves. Our professors knew this and preferred to test our ability to apply our learnings over our ability to memorize formulas.
This is all to say it is important to get behind the simplified formulas and understand the variables that drive them. Despite not being able to calculate them with precision, we can still understand an opportunity pretty accurately by looking at what we do know. The key variables in the LTV formula above are:
ARR = size of a customer
n = lifetime of customer
churn = retention dynamics
Pairing that with Customer Acquisition dynamics, which are also characteristics of a market, we can get a fairly accurate picture of the unit economics of an opportunity without a single operating metric in hand. The math at this stage is simple and it’s worth everyone’s time to run through it to better understand which variables matter, how they interact, which are the most sensitive, and ultimate what to be on the lookout for assessing risk and upside along the way towards product-market fit. We can also figure out which operating metrics to track so that founders and their investors will understand if what they are doing is actually working and how well.
An example: we have a certain portfolio company that many would call “consumer fintech,” which as a category has a bad reputation for retention, distribution and transaction sizes - overall bad unit economics. But this company operates a viral GTM, has SMB-sized ACVs and retention on par with government contracts. The lesson here is that oversimplifying and labeling can cause you to miss the best opportunities, but understanding the fundamentals and abstracting key business characteristics is a secret weapon.
If you’re wondering where I have been the last few months, I have started also publishing less quantitative ideas via the Verissimo Ventures Substack, which you can find here:
