Unit Economics: Lifetime Value
The practical guide for assessing the unit economics of your early stage marketplace business.
Of all the metrics that you should be measuring in your marketplace business, your ‘unit economics’ are perhaps the most important. Despite this, they are poorly understood and prone to errors in calculation. This series offers a practical guide as to what you should be measuring and how you should be measuring it. It will be most relevant to anyone involved with a marketplace businesses in the Seed to Series A range.
What are Unit Economics?
Unit economics are the basic underlying building blocks of the most simple element of your business model:
- The cost associated with acquiring a customer. This is know as your Customer Acquisition Cost (CAC) or Cost per Acquisition (CPA)
- The revenue associated with each customer over the time they spend using your product. This is your Lifetime Value (LTV)
For your business to become profitable (which seems a bit unfashionable today), your LTV must be more than your CAC. This makes sense as you do not want to be spending more to acquire a customer than the amount or revenue returned to you by them.
Generally a ratio of >3:1 is seen as strong but take this with a pinch of salt. There are many subtitles in the analysis of unit economics and they are often open to interpretation. It is therefore essential that you understand how you assess your business’ unit economics so they can withstand inevitable scrutiny.
In the first of this series, we will kick things off by addressing LTVs. Subsequent articles will tackle CACs and the nuances associated with applying these metrics to different marketplace businesses.
Why do we measure Unit Economics?
In the earlier stages of your start up, you will most likely be looking into unit economics as part of a conversation with investors. Indeed, when going into a fundraising process, it is important to have a good idea of how unit economics work.
Even pre-revenue, you should be able to make some educated assumptions as to how things will shape up. At a Series A raise, LTV and CAC should be your middle names. At any stage, an appreciation of unit economics is a big tick for any potential investor.
As your marketplace business matures, your understanding and subsequent use of unit economics will migrate from ‘display purposes only’ to becoming operationally essential. Understanding the core drivers of your business will help guide decision making. It is worth putting in the work to be able to access these insights quickly and accurately as it will let you identify the levers that you can pull in order to drive growth.
Cohort Analysis
Before we dive into analysis of unit economics, we need to ensure we have a good understanding of ‘cohorts’.
Cohort analysis is an analytical technique that takes the data from a given set and, rather than looking at all users as one unit, breaks them into related groups for analysis.
There are many ways you can cohort your users within your marketplace and your chosen cohort characteristics will depend on the intricacies of your model. At Florence, we split out users by three distinct cohorts:
1) Supply vs Demand. This one is almost too obvious to mention but you need to be able to differentiate behaviour between buyers and sellers.
2) Geography. Almost every marketplace business that involves the exchange of non-digital services will have a geographical component.
We split the UK into ten regions and allocate our supply and demand side users into each region based on their home postcode. This is the most important cohorting we do to drive operational decision making.
3) Time. In order to assess changes in our marketplace dynamics over time, we create cohorts depending on when users first transact on the platform.
Time-based cohorting has less of an operational impact to us today but is important when it comes to analysing trends over time and being able to forecast supply and demand as we launch new markets.
Calculating Lifetime Value
There are a number of caveats when calculating lifetime values and the road to accurate prediction is fraught with difficulties. Below I have provided two techniques that we have found useful.
Method 1 — Using Historical Data:
The simplest way to start thinking about LTV is to set a time period (say one year) and, using historical data, examine the revenue to your platform of a user cohort over this period.
In the worked example in the GoogleSheet: Method 1 (https://docs.google.com/spreadsheets/d/1xt0zQ1ZRgfSHfHud0M6-pPprcL1aw3DtFYnU3wLnZAU/edit?usp=sharing), we have taken a hypothetical group of users from an on-demand staffing platform and:
- taken a cohort of users who first became active in a given month
- taken the total volume of work completed by that cohort over the subsequent 12 months
- divided that work by the number of users in the cohort
- multiplied that by the average hourly rate and, finally, the platform take rate (or commission)
Although this method gives a quick and dirty insight into your users’ LTV, it is fundamentally flawed. As this analysis requires you to have a significant data set (at least a year of user data), by the time you are able to draw conclusions about the examined cohort, you are already a year behind the events that caused their behaviour within the platform.
If you are looking to use LTVs to assess the impact of process or product change in your organisation, then this tool is too blunt. In a pre-Series A company, one year is far too long to wait to draw these conclusions.
Method 2 — using users lifespans:
LTV can be calculated more accurately (and without having to wait for years of data) using the formula below:
LTV = Lifespan x Average Earnings per Period x Platform Take Rate
This method relies on being able to calculate a user’s ‘Lifespan’ — the amount of time a user, on average, stays with your platform before they stop transacting. In order to calculate Lifespan, you need to get a handle on ‘Churn’ i.e. the number of users you lose from a given cohort in each subsequent time period that passes.
Example: You have a business that sells lemonade to your neighbours. You set up your lemonade stand in June and sell 100 different people a glass of lemonade over the course of the month. In July, 80 of those 100 people come back to your stand. This gives you a month on month churn rate of 20%. Your retention equals 1 — Churn; in this case, 80%.
N.B. It is important to note that for this calculation, it does not matter how many lemonades each of those customers bought in June or indeed if you attracted any other first time buyers in July. Churn is constituted only from the month on month change in active customers in each cohort. The time period you use to measure churn needs to make sense to your business but monthly or quarterly periods are used most frequently.
The example seen in Method 2 (https://docs.google.com/spreadsheets/d/1xt0zQ1ZRgfSHfHud0M6-pPprcL1aw3DtFYnU3wLnZAU/edit?usp=sharing) analyses user LTV based on monthly Churn of individual cohorts. You can see in this example the importance of being able to cohort your users based on when they first transacted on your platform.
This analysis looks into the ratio between how many users were Retained/Resurrected vs Churned in each cohort and takes a weighted average of those monthly churn rates. From this number you can calculate the lifespan of a user by dividing 1 by the churn rate.
Lifespan = 1 / Churn Rate
Although more complicated to set up, this method for calculating LTVs allows you to draw significantly faster conclusions about your user base. Having a firm handle on churn (and its closely related cousins of retention and lifespan) is completely essential for any marketplace business. It is a discussion that merits an entire post on its own.
Measuring LTV over time
Do not fire and forget with your unit economics — the more you put in, the more you will get out of it. When you can afford it (around your Series A raise), it is worth investing in a data analytics tool that gives you easy access to this information. Whatever your underlying business model, as your product improves, your user base grows and network effects take hold, you should see your LTVs improving. An example of this in practice can be found in Evolution of LTV (https://docs.google.com/spreadsheets/d/1xt0zQ1ZRgfSHfHud0M6-pPprcL1aw3DtFYnU3wLnZAU/edit#gid=1784914791)
Conclusions:
- Unit economics are not just numbers to tell investors — they can help assess the health of your marketplace and drive critical business decisions
- Understand the most relevant ways to cohort your users but don’t get bogged down in over analysis too early
- LTV = Lifespan x Average Earnings per Period x Take Rate
- Lifespan = 1/Churn
- Start thinking about LTVs soon. Start off by doing it manually with a spreadsheet. When you can, invest in an analytics tool that let’s you pull out information easily.
If you’d like to read more about marketplace unit economics, be sure to check out the practical guide to customer acquisition costs.