Unit economics review (LTV vs CAC).
Restricted-market unit economics work differently from unrestricted-market unit economics — CAC is typically 2–4× because acquisition channels are constrained, but LTV is often higher because category lock-in and repeat purchase are stronger. The model has to be calibrated to those category specifics, and most operators run on assumptions imported from unrestricted-market playbooks.
Get in touch→- Service code
- SVC-010.3
- Parent service
- Strategy →
- Coverage
- 18 verticals
- Engagement modes
- Retainer · Project · Partnership
- Reply window
- < 24h on enquiries
- LTV modelling by category and cohort, calibrated to actual restricted-market retention curves
- CAC modelling per acquisition channel, with attribution that handles channel constraint
- Payback-period modelling and break-even thresholds for each channel
- Channel-level contribution-margin analysis
- Scenario modelling for channel-mix adjustments
- 01
Data ingestion
Pull from your CRM, payment processor, analytics, ad accounts. Build a unified dataset that lets us see the full LTV curve and the channel-level CAC.
- 02
Cohort modelling
Customers grouped by acquisition month, channel, product. Retention curves built per cohort. LTV measured against actual cohort behaviour, not generic averages.
- 03
CAC and contribution analysis
Channel-level CAC, blended CAC, channel contribution margin. Where the math actually works, where it doesn’t.
- 04
Scenario modelling
What happens if you shift spend from channel A to channel B. What happens if retention improves 10%. What happens if pricing changes. Documented scenarios for decision-making.
01 How long does the modelling take?
3–6 weeks depending on data quality. Most operators have data scattered across 4–8 systems; the consolidation work is most of the timeline.
02 Will this work if my attribution is messy?
Yes — we’ll work with what you have and document the gaps. Perfect attribution doesn’t exist in restricted markets (last-click breaks, MTA models break, walled-garden data is opaque). We build the most-likely model and explicit confidence intervals around it.
03 Is this a one-time engagement or ongoing?
One-time for the initial model. Quarterly or annual refresh as data accumulates. Most operators do an initial model in year one and a refresh annually thereafter.
Tell us about it.
Drop a quick brief or write directly: contacts@despitemarketing.com. Telegram @despitemarketing. Signal @despitemarketing.