Customer Lifetime Value: Hartnow
A customer analytics simulation for a B2B technology company, completed with my team for Customer Data Analysis (Honors), spring 2026.
The question
Hartnow was a young company that investors judged on both profit growth and customer value, and our team was asked to decide which customers it should pursue and how it should keep them over time.
Where we started
In the early rounds we followed the revenue. Automotive customers brought in $32.3 million of $47.5 million in total revenue, so we built most of our marketing and service spending around them.
What the data showed
I ran two regressions on the full base of 940 customers. The first explained nearly all of the variation in revenue, and only two drivers mattered: the number of contracts each customer held, worth about $24,700 in revenue per contract, and the share of those contracts that were advanced. Sector and acquisition channel had no significant effect.
| Driver of revenue per customer | Effect | Significant |
|---|---|---|
| Number of contracts | about $24,700 per contract | Yes |
| Share of advanced contracts | about $12,400 from all basic to all advanced | Yes |
| Customer sector | no reliable effect | No |
| Acquisition channel | no reliable effect | No |
The second regression found that no visible customer trait predicted who would upgrade to advanced contracts, which told us that upgrades would only come from deliberate account work. Our retention analysis also showed that defense customers churned far less than automotive ones, which made them more valuable over time even though they brought in less revenue.
What we changed
We shifted our spending toward retention, contract depth, and advanced product adoption, and when we raised retention from about 31% to 40%, our estimated customer equity rose significantly.
The same course also included cohort and decile analysis of a retailer's customer base, where I tracked how order frequency, order value, and retention changed as each group of customers aged.
The simulation showed me how easy it is to let the biggest revenue number set the strategy, and how much clearer the picture becomes once you look at what each customer is worth over the full relationship and test that against the data before committing the budget.
The full Excel analysis is available on request at tkachr@bu.edu.