Lifetime Customer Value (LTV): The Right Metric?
As we discussed in one of our post on Avoiding the Reasons why most SaaS companies fail, understanding how much value a customer is generating for the business is critical to knowing whether you’re running a profitable business or driving it into the ground. The Customer Lifetime Value (LTV) is the most commonly discussed metric for figuring out this value. However, how is LTV calculated, is it even possible to calculate it for a company at a very early stage, and is it even important?
… the present value of the future cash flows attributed to the customer relationship…
Or, in other words, the LTV identifies how much an average customer will return in value to the business over its predicted lifetime. A “classic” LTV calculation includes a number of factors as part of the calculation:
- Periodic Revenue – The amount of revenue collected from a customer in the period. In the case of SaaS companies, this would be the subscription revenue if appropriate.
- Profit Margin – Profit as a percentage of revenue, which is often reflected as gross profit, subtracting the cost of servicing a customer.
- Period - The unit of time into which a customer relationship is divided for analysis. A year is the most commonly used period, with the LTV being calculated across multiple time periods based on the estimated lifetime of the customer.
- Discount rate – From the Wikipedia entry: “The cost of capital used to discount future revenue from a customer. Discounting is an advanced topic that is frequently ignored in customer lifetime value calculations. The current interest rate is sometimes used as a simple (but incorrect) proxy for discount rate.”
- Churn rate – In many simplified LTV calculations, the churn rate is not considered, but is necessary to get an accurate customer value. The churn rate is defined as the percentage of customers who end their relationship with a company in a given period, with the assumption that the churn rate is constant across the life of the customer relationship. One minus the churn rate is the retention rate. .
- Retention cost – The amount of money a company has to spend in a given period to retain an existing customer. This may be factored into the Profit Margin calculation or identified as a separate cost.
LTV = ARPU x Average Lifetime of a Customer – the Cost to Serve them (COGS)
Where ARPU = Average Annual Revenue Per User and Average Lifetime of a Customer = 1/Churn Rate.
The above calculation dismisses the idea of discounted future cash flows (assumes a future dollar is equivalent to a current dollar) and assumes retention cost is included in COGS as part of the calculation.
The folks at the Database Marketing Institute provide a very good and clear example of LTV calculations that is a good starting point for most efforts.
There are a number of real challenges with LTV, especially for companies just starting out:
- How do you know what the predicted lifetime of a customer will be as measured in months or years?
- How do you know what the expected cash flows (revenue) will be for that customer in that time period?
- How do you know what your churn rate will be?
- Is a future dollar really worth a current dollar?
- How accurately can you measure COGS and/or retention costs?
- Is measuring a customer’s value across their whole lifetime of more or better value than measuring over a shorter period?
- Do you include the value of referrals?
- Do you segment customer groups or calculate one overall number?
- When is the customer is no longer a customer?
- Do you include all customers, including those who simply trialed a product?
Also, are all customers really treated the same? What about the “power users” who derive the majority of benefit from the system and thus have a much greater LTV? Should these be lumped in with different users who see minimal value and would be willing to switch to a competitive offering at the drop of a hat? So many assumptions have to go into what otherwise would be a very simple calculation. This is where many startups get stuck — or start to fool themselves.
Is Determining LTV Even Possible for an Early Stage Startup?
As we discussed above, calculating LTV for a company just starting out is as much an exercise in predictive fantasy as the development of three-year financial projections. If you’ve been in business for just a few short months and have barely (if at all) gotten product out the door, how on earth can you know how long your customer will derive value for you, what their COGS are, the retention rate, and the churn rate? At best, you might have some comparables in the industry that might provide a clue. At worst, you’re starting out with metrics so invaluable that they are not worth considering at all.
What industry trends can you identify that indicate purchase size, retention statistics, or customer lifetime? What assumptions can you make now for the sake of practice? Try your calculation using different lifetimes, and refine your numbers every month, quarter, or year. Compare your assumptions to actual results. Compare your more conservative estimates (lower LTV) to your more aggressive estimates, and compare both to acquisition costs.Your number will be less reliable, but the calculation process will keep you focused and set the stage for increased accuracy every month.
In other words, if you’re just starting out, consider LTV to be one of those hypotheses that needs to bear continuous evaluation to disprove. Come up with multiple LTV theories. Use those multiple theories to create benchmarks so you can evaluate how you’re doing compared to worst-case and best-case calculations. But most importantly, don’t come up with a bogus number and stick with it because it’s a metric.
LTV vs. Annual Customer Value (ACV)
One of the things we are doing at Yobiz is using Annual Customer Value (ACV) as a proxy for the LTV. We don’t know how long a customer will stick with us because we don’t have a long-enough track record to reliably make such a prediction. However, a year is a good enough horizon that allows us to make calculations that are still useful. If a customer provides value for greater than a year, then this is purely upside in our calculations. If they don’t stick with us for a year, then we have other problems we need to address.
The ACV is simply the LTV calculation (using David Skok’s simplified one) with a lifetime set at one year. Since we’re using a one-year calculation, we don’t need to factor in a Discount Rate since the time horizon is set at a year. The real challenge thus becomes predicting a churn rate and establishing the COGS. Our models thus involve setting a few different churn rates and a few estimates of COGS (worst case, expected case, best case) and developing metrics and models around those. Making sure we achieve at least a one-year customer life with churn rates and COGS are between the limits we set up will help us get to where we need to be.
Calculating LTV for Customer Segments
As you run your business, you’ll start to notice that not all your customers are the same. Some are heavy, power users from specific market segments that not only derive and generate significant value, but also have very long lifetimes. Other customers provide marginal value, or worse, their LTV is exceeded by the costs to acquire them. It doesn’t make sense to lump these two groups together for one specific reason: you want to focus on growing your pool of best customers.
Pareto’s principle says that , for many events, roughly 80% of the effects come from 20% of the causes. As applied here, this means that roughly 20% of your customers will generate 80% of the value for the company. The challenge thus is to identify those 20% in order to maximize the 80%. If you’re just starting out, it will be very hard to find those 20% beforehand. By measuring LTV (or ACV as a proxy) for each identified customer segment, you can identify those trends early and focus your acquisition so that you not only maximize overall customer value, but minimize acquisition costs and inefficiency.
For an example of how market segment-based LTV calculations would work, check out Avinash Kaushik’s post on the subject, although note that he includes the Customer Acquisition Cost (CAC) in his Net profit calculations. This can potentially be a problem since you need to compare LTV to CAC to make sure that you’re moving in the right direction. If you do include CAC in the LTV calculations, then note that you are trying to avoid at all costs a negative number. Make sure to do LTV and ACV calculations if you’re using this approach and realize you still need to compare CAC against an LTV number that doesn’t include it in the calculation to measure these independent trends.
What are your thoughts on LTV and ACV? Is this a useful metric?