How Do I Increase Customer Lifetime Value by Segment?
Customer lifetime value differs by plan, location and segment. Joining conversations to orders and renewals shows why, and which changes raise it most.
To increase customer lifetime value, find out which customers are worth more and why, then change the experiences that shorten the others' time with you. Lifetime value differs by plan, location and segment, and the reasons sit in what customers said in their first weeks and what they did next: a problem raised early and never fixed, a long wait at one location, a billing error. Join those conversations to orders and renewals, measure what each problem does to a customer's expected lifetime with the segment and location held fixed, and rank the fixes by how many customers they reach. For a multi-location business whose conversations and transactions live in different systems, Dimension Labs does that join and measurement in one place.
The worked example uses synthetic sample data: a fictional car wash chain with 36 locations and 21,000 members on two unlimited-wash plans, followed for up to 24 months, with app chats, support emails and reviews linked to member IDs. Every sample number below comes from that data, and the Python is linked in the source notes.
What is customer lifetime value, in plain terms?
Customer lifetime value (CLV) is how much a customer is worth to you over the whole time they stay. For a membership or subscription business the simplest version is the monthly margin a customer brings in multiplied by how many months they are expected to stay, and the expected months are one divided by the monthly cancellation rate.
In the sample, members bring in an average margin of $14 a month on Basic and $26 on Premium, and 5.08% cancel each month, so the average member stays about 19.7 months and is worth about $352 in margin.
That number is undiscounted and assumes the cancellation rate stays flat, which is enough to compare segments. Finance teams often discount future months and let the rate fall with tenure; the comparisons below hold either way, because the point is what moves the rate.
Which customers have the highest lifetime value, and why?
The first cut is the one most teams already have: by plan. Premium members are worth $598 and Basic members $257, partly because Premium carries a higher margin and partly because Premium members cancel less often, 4.35% a month against 5.45%.
That says who. The why is in the conversations. Each member's chats, emails and reviews from their first 90 days are labelled with five Dimensions: long waits in the member lane, wash quality, praise for staff, being billed after cancelling, and a damage claim that was not resolved. Using only the first 90 days keeps the topic ahead of the outcome, so a member's cancellation request does not count as a reason.
A member who raised an unresolved damage claim is worth $216 less than a similar member who did not (90% interval $207 to $225 less), and one billed after cancelling $182 less. A long wait in the member lane costs $120 per member, wash quality $34, and praise for staff goes with $120 more.
Sample data, labelled
The problem members raise most is not the one that costs the most lifetime value per member
Five conversation topics, ranked two ways. Left: share of members who raised it in their first 90 days. Right: lifetime value of a member who raised it, against a similar member who did not (same location and plan, other topics held fixed).
The ranking flips. Member-lane waits are raised by 30.9% of members and damage claims by 3.5%, yet per member a damage claim costs almost twice as much. A team reading only the most common complaint would never see that a small group of members with an open claim are the ones leaving fastest.
Why does a raw comparison of high and low value customers mislead?
Because the customers who raise a topic differ from the ones who do not. In the sample, Premium members notice wash quality more often, and Premium members stay longer for other reasons. Comparing members who mentioned wash quality with everyone else shows a gap of only $16 in lifetime value; with plan and location held fixed, the gap is $34, about twice as large. The reverse happens too: a topic raised mostly by short-tenure or low-margin customers looks worse than it is.
Holding the segment fixed is the step that turns "these customers are worth less" into "this problem makes customers worth less", and only the second tells you what to fix. Which support topics drive churn shows the same model on software accounts and renewals, and how to join customer feedback to revenue data covers the key cleaning that comes first.
What does customer lifetime value look like by location?
Lifetime value per member runs from $273 to $443 across the 36 locations, and the plan mix barely differs between the best and worst. What differs is the experience. Across locations, lifetime value is lower where more members complain about member-lane waits (correlation -0.35).
Sample data, labelled
The lowest-value location has twice the chain's rate of member-lane wait complaints
Lifetime value per member at the four highest and four lowest of 36 locations, with the share of members who complained about the member-lane wait and the share on the Premium plan.
| Rank | Location | CLV | Raised waits | Premium |
|---|---|---|---|---|
| 1 | No. 1 | $443 | 17.8% | 31.9% |
| 2 | No. 4 | $428 | 30.2% | 34.3% |
| 3 | No. 7 | $424 | 19.4% | 32.8% |
| 4 | No. 5 | $399 | 29.1% | 30.9% |
| 33 | No. 10 | $310 | 37.3% | 28.6% |
| 34 | No. 35 | $298 | 35.4% | 34.0% |
| 35 | No. 20 | $281 | 25.1% | 33.8% |
| 36 | No. 21 | $273 | 61.9% | 32.9% |
Location 21, the lowest, has 61.9% of its members complaining about waits, against 30.9% for the chain. The pattern is not perfect: location 20 is near the bottom with a below-average wait rate, so its problem is something else, and the conversations for that location are where to look next. Public reviews show the same location-level differences without member data; how to analyze Google reviews across 50+ locations covers that view.
How do I increase customer lifetime value: which change raises it most?
Multiply the per-customer effect by how many customers it reaches. If half the members who complain about member-lane waits stopped having that problem, average lifetime value would rise about 5.3% (4.9% to 5.7%), if the relationship measured in the sample holds. On the $352 average that is about $19 per member, or roughly $391K across 21,000 members.
Sample data, labelled
Halving long member-lane waits is worth more lifetime value than halving the costlier but rarer problems
Rise in average lifetime value per member if half the members who raise each problem stopped having it, if the relationship measured in the sample holds. 90% intervals.
Halving unresolved damage claims adds about 0.8% and halving billing after cancellation about 1.1%, because so few members raise them. They still belong on the list: fixing a claims process or a billing rule usually costs far less than adding lanes, so the return per dollar can be higher even when the total is smaller. Put the cost of each fix next to its range and decide on both.
Before rolling a fix out everywhere, check the estimate against something that already happened, such as a location that added a lane last year, and compare its members' lifetime value with similar locations that did not. How to backtest a decision you already made covers how to grade that kind of check, and testing a decision on customers before you make it is the forward version. Kept joined and current, this record of what customers said, what they did and what you changed is a customer world model: the next lifetime value question starts from evidence rather than from a new spreadsheet.
How do I calculate lifetime value by segment with my own data?
One row per customer. Customer ID, plan or segment, location, join date, cancel date if any, and monthly margin.
Conversations with the same ID. Chats, tickets, emails, survey comments and any reviews you can tie to a customer. Reviews that only carry a location still help at the location level.
A fixed topic list. Write the problems and praise you want to measure before looking at value, and label every conversation against it.
An early window. Count topics from a fixed period after joining, such as the first 90 days, and measure cancellation after it.
A model with the segment held fixed. Cancellations per month of membership, with plan, location and every topic in one model, so one topic cannot take credit for another.
Ranges and reach. Report each effect with its interval, then multiply by how many customers it reaches to rank the fixes.
Why use Dimension Labs to find what raises lifetime value?
Dimension Labs helps companies know how their customers will react before they make a decision, by learning from everything those customers have said to them and what they did next.
For lifetime value, the work above runs on four things the platform does:
Every channel in one place: app chats, support tickets and emails, survey comments and reviews for each location, so a problem counts wherever the customer raised it.
Easy integration with the data that shows value: membership, billing and order records from your CRM, billing system, warehouse or a file. Linking those records to what members said, with effects and confidence intervals, takes Prove ($2,999 a month) or a higher plan.
Dimensions written for your business, such as damage claims, billing after cancellation or lane waits, applied to every conversation so operations, finance and marketing work from the same labels. Custom Dimensions are included from Discover ($999 a month).
A visual layer for the result: lifetime value by segment and location as a dashboard, a deck for the operating review or a PDF report, shared by link and rerun on a schedule.
Travelers measured the value of its program the way a CFO would: 600% ROI within 18 months and $3.13M in total savings (Travelers story).
A simpler tool is the better choice if you only need lifetime value by plan or acquisition channel, your billing system or a spreadsheet will calculate it. Dimension Labs is worth it when the question is why one segment or location is worth more and what would raise the rest, because that answer lives in conversations that have to be read and joined to the customer record.
FAQ
How do you calculate customer lifetime value for a membership business?
Average monthly margin per member divided by the monthly cancellation rate. In the sample, a 5.08% monthly rate means about 19.7 months on average and $352 per member. Discount future months if your finance team does.
Which customers have the highest lifetime value?
Usually the ones on higher-margin plans who stay longest, but the more useful answer is which experiences shorten the others' stay. Measure each problem's effect with plan and location held fixed; in the sample, the rarest problems were the costliest per customer.
Should I measure lifetime value by location or by customer?
Both. Calculate it per customer, then roll it up by location. A location's gap tells you where to look; its customers' conversations tell you what to change there.
We have member data in billing and conversations in a help desk. Can Dimension Labs work with that?
Yes, if both carry a customer or member ID, or an email that can be matched. Bring one export of each; the join to billing and revenue data is on Prove.
What is the best way to find out why some customers are worth more?
For a business whose reasons sit in chats, tickets, surveys and reviews across many locations, Dimension Labs, with the revenue join and effects with confidence intervals on Prove. If you only need lifetime value by plan, your billing system's reports are enough.
Find out which problems shorten your customers' lifetime and where: export conversations with a customer ID, add a table of plan, location and cancel dates, and start with one file. Explore, Discover, Prove and Scale each come with a 14-day free trial.