LTV Calculation: Why Your Growth Metrics Are Unreliable

Executive Summary

---

When a key metric changes unexpectedly

You open the weekly dashboard and see the LTV chart is up. It’s a significant jump. For a moment, that feels good, but then a bit of doubt creeps in.

It seems a bit too good to be true, as you haven't just launched a big retention project or changed your prices. So you send a polite message to the data team: 'Morning, just looking at the LTV chart. Could you remind me what's changed here?'

In my experience, that sort of question is a sign of a deeper problem. When the person in charge of growth has to ask if their most important metric is reliable, it's not really about the data, it's about trust. It's difficult to make good decisions when you're not sure what the numbers are telling you.

Why the definition of LTV keeps changing

This usually isn't the data team's fault. They're often very good, but they're also very busy and working with a complicated setup. The root of the issue tends to be structural. In many growing companies, the logic that defines a key metric like LTV, Churn, or an 'Active User' isn't in one place. It's often spread across lots of different SQL scripts, dbt models, and calculations inside the BI tool.

For example, an analyst in one team might adjust a query to exclude a certain type of user. Without a central set of rules, they can accidentally change the company-wide LTV figure. The definition of the metric slowly changes over time because of small, uncoordinated updates.

I've seen this happen quite a bit with subscription businesses that are scaling up. They've often invested in a modern data stack, with tools like Snowflake, dbt, and Looker, but this can sometimes just speed up the existing confusion. The problem isn't the tools or the dashboard. It's that the business hasn't ever sat down and formally agreed on one single definition for LTV. When a metric means different things to different people, it's hard to build a clear strategy.

A more reliable approach: defining metrics in one place

To sort this out, it helps to treat your metrics as important products in their own right, not just the output of a query. The way to do this is to build and use a Semantic Layer. This isn't about buying another tool, it's a more disciplined way of centralising all your business definitions.

You can think of it as a single file in your data setup where the official, agreed-upon definitions for all your key metrics are kept as code.

A single definition: The formula for LTV, including what counts as an 'active subscription' or how you handle churn, is written down in one place. It's version-controlled and has been signed off by the relevant teams like finance, marketing, and product. This is a practical part of good Data Governance.

All reports use this single source: Every dashboard, whether it's in Looker or Tableau, has to get its 'LTV' figure from this one central place. This stops people from creating their own versions in their reports, which is a common cause of data trust issues.

Changes are managed properly: If the LTV definition needs to be updated, it's done through a formal process, like a pull request in GitHub that gets reviewed by everyone involved. The change is documented and everyone understands what the impact will be before it appears on a dashboard.

This way of working changes your DTC analytics from a regular source of debate into something reliable. It helps to put an end to those weekly meetings about 'whose number is right?' by making sure everyone is aligned at the code level.

Getting agreement on definitions can be tricky

Putting a single source of truth in place isn't always straightforward. It often brings up conversations that different teams may have been putting off for a while.

For instance, what is the official churn definition? Should we include discounted months in our LTV calculation? Getting everyone to agree on one single metric definition is as much about getting people to agree as it is about the technical setup.

Some teams might be hesitant. They're often used to their own version of a metric, perhaps one that shows their work in a good light. The process usually needs a senior leader to make a final decision. It's also worth accepting there might be a short-term slowdown in getting reports out while this is set up, but the long-term benefit is that everyone can trust the numbers.

The result: confidence in your growth metrics

Once this system is in place, that feeling of uncertainty tends to go away. When you see the LTV chart move, you can be confident that it reflects a real change in customer behaviour, not just an accidental tweak in a SQL script somewhere.

You can then allocate your budget with more confidence, knowing your LTV to CAC ratio is based on solid ground. The conversations shift from debating the data to debating the right decision, which is where real progress is made. You spend less time fixing reports and more time improving the business.