A customer can be easy to acquire and expensive to keep. If the acquisition report stops at the signed contract, that distinction disappears.
I would put cost to serve beside acquisition cost when deciding which customers, packages, and channels deserve more investment. The point is to understand the work a sale creates across the company.
Map the work after the signature
Start with a recent customer cohort. Identify implementation, onboarding, ongoing delivery, support, account management, and usage-related infrastructure costs. Separate work that happens once from work that repeats.
Assign time at a consistent internal cost. The estimate does not need to be perfect to expose a large difference between customer segments, but everyone needs to understand what is included.
Do not count the same expense in both customer acquisition cost and delivery cost. Write the boundary down with the person who owns the financial model.
A worked comparison
This is a hypothetical first-year comparison, not a client result. Both customers produce $24,000 in revenue. The categories below are simplified allocated costs and exclude acquisition and general company overhead.
| First-year item | Customer A | Customer B |
|---|---|---|
| Implementation | $2,000 | $7,000 |
| Ongoing delivery | $6,000 | $8,000 |
| Support and account work | $2,000 | $4,000 |
| Revenue less these costs | $14,000 | $5,000 |
The same contract value leaves very different room to fund acquisition and overhead. This does not automatically make Customer B a bad customer. A reusable integration, expansion opportunity, or strategic learning could matter. Those are additional hypotheses, not reasons to hide the cost.
Find the source of the difference
Review whether the additional work came from a weak handoff, an unclear promise, product limitations, customer readiness, or an unusually complex use case. Those causes require different responses.
If every customer struggles with the same onboarding step, improve the process. If one segment consistently needs custom work, reconsider packaging, pricing, scope, or fit. If sales promised an unsupported capability, fix the qualification and approval process.
Turn the finding into a decision
Choose one intervention and one follow-up measure. For example: standardize a data import, record implementation hours for the next cohort, and compare time to value and support burden.
Avoid treating lower service cost as the only goal. Cutting work that creates retention or customer value can damage the economics you are trying to improve. Read the cost alongside outcomes, then use the CAC payback framework to connect acquisition spending to the contribution the customer actually creates.
Co-founder and CEO of Stackmatix, startup advisor, and former Head of Sales at MightyHive. Connect with Matt on LinkedIn ↗ · More about Matt →