Was last month’s marketing worth it? A cohort dashboard that answers early.
A fintech client acquires investors who keep investing for years, so a month’s ad spend cannot be judged by that month’s revenue. We built a cohort dashboard on their own backend data with a custom lifetime value, so the team can see within weeks whether a month’s spend will pay back. The version below is a rebuild on invented numbers.
One table answers one question per month: did that cohort pay for its marketing? It follows every month’s registrants over time, projects their lifetime value from real behavior, and sets it against what it cost to acquire them and how many projects were open when they arrived.
- The one number to watch
- pLTV/CAC
- Lifetime value projected per cohort
- 24 mo
- When a first read gets confirmed
- Month 2
- Built on the client’s own backend data
- 100%
Cohort LTV: what each month’s marketing spend returns
| Cohort | M+0 investors | Total spend | Open projects | pLTV/CAC | ROI | Payback | Verdict |
|---|---|---|---|---|---|---|---|
| Sep ’26* | 390 | €53.4k | 12 | 2.13× | 113% | 6 mo | Efficient |
| Aug ’26* | 343 | €49.6k | 14 | 1.99× | 99% | 7 mo | Efficient |
| Jul ’26 | 149 | €43.4k | 3 | 0.85× | -15% | 14 mo | Does not pay back |
| Jun ’26 | 246 | €45.3k | 8 | 1.50× | 50% | 8 mo | Pays back |
| May ’26 | 324 | €48.2k | 12 | 2.00× | 100% | 7 mo | Efficient |
| Apr ’26 | 245 | €60.5k | 5 | 1.02× | 2% | 12 mo | Pays back |
| Mar ’26 | 347 | €49.2k | 13 | 2.14× | 114% | 6 mo | Efficient |
| Feb ’26 | 289 | €41.3k | 10 | 1.97× | 97% | 7 mo | Efficient |
| Jan ’26 | 309 | €42.3k | 12 | 2.07× | 107% | 6 mo | Efficient |
| Dec ’25 | 119 | €31.6k | 4 | 0.96× | -4% | 12 mo | Does not pay back |
| Nov ’25 | 285 | €40.1k | 11 | 2.04× | 104% | 6 mo | Efficient |
| Oct ’25 | 234 | €36.2k | 9 | 1.77× | 77% | 7 mo | Efficient |
* Two months of data or fewer, so the row is mostly projection. Below 1.00× a cohort does not pay for itself.
Cohort matrix: how each month’s registrants behave over time
| Cohort | Size | M+0 | M+1 | M+2 | M+3 | M+4 | M+5 | M+6 | M+7 | M+8 | M+9 | M+10 | M+11 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Sep ’26 | 3,240 | 12.0% | |||||||||||
| Aug ’26 | 2,980 | 11.5% | 7.1% | ||||||||||
| Jul ’26 | 2,300 | 6.5% | 4.0% | 3.5% | |||||||||
| Jun ’26 | 2,540 | 9.7% | 6.2% | 5.4% | 4.8% | ||||||||
| May ’26 | 2,760 | 11.7% | 7.6% | 6.8% | 5.8% | 5.6% | |||||||
| Apr ’26 | 3,120 | 7.9% | 5.3% | 4.1% | 3.7% | 3.4% | 3.5% | ||||||
| Mar ’26 | 2,890 | 12.0% | 7.1% | 6.5% | 6.2% | 5.7% | 5.4% | 4.8% | |||||
| Feb ’26 | 2,450 | 11.8% | 7.9% | 6.5% | 5.7% | 5.4% | 4.9% | 4.5% | 4.5% | ||||
| Jan ’26 | 2,610 | 11.8% | 7.3% | 6.4% | 6.3% | 5.2% | 4.8% | 4.5% | 4.3% | 4.6% | |||
| Dec ’25 | 1,720 | 6.9% | 4.3% | 3.8% | 3.3% | 3.4% | 3.1% | 2.6% | 2.7% | 2.5% | 2.5% | ||
| Nov ’25 | 2,380 | 12.0% | 7.8% | 6.5% | 5.8% | 5.2% | 5.3% | 4.7% | 4.6% | 4.4% | 4.5% | 4.1% | |
| Oct ’25 | 2,140 | 10.9% | 7.3% | 6.0% | 5.6% | 4.9% | 4.6% | 4.5% | 4.4% | 4.1% | 4.0% | 3.9% | 4.0% |
The question
The marketing team had a monthly target for new investors and a pacing report to track it. That told them whether they were hitting the number. It did not tell them whether the number was worth what it cost.
On an investment platform the first investment is small and the value comes later, as people keep investing month after month. Two months can bring in the same number of new investors at the same cost and end up worth very different amounts. Ad platforms cannot see any of that. They see a registration and a first deposit.
What we built
A cohort is everyone who registered in a given month. That group is fixed forever, and the dashboard follows it month by month: how many of them invest, how much on average, and how fast it fades.
The Simple view keeps it to the numbers that decide whether a month’s spend was worth it. The Detailed view groups every metric by the question it answers: volume, cohort quality, acquisition cost, market supply and return. Underneath sits the cohort matrix, which shows the raw retention and average investment behind every projection.
One detail mattered more than it looks. The dashboard separates investors who registered and invested in the same month from everyone whose first investment landed that month. They are the same people counted from a different date, and mixing the two is the fastest way to get two teams arguing over whose number is right.
A custom lifetime value
Nobody can wait two years to learn whether a cohort paid off, so the dashboard predicts it. For each cohort it takes what is already visible, what share still invests each month and how much, and projects it forward using the shape older cohorts followed. The platform’s own revenue model turns projected investment volume into projected revenue per investor. That is pLTV.
Divide the cohort’s pLTV by what that month’s marketing cost and you get pLTV/CAC: how much lifetime revenue one unit of marketing spend returns. Below 1.00× the cohort does not pay for itself.
The number moves as a cohort matures, because every month replaces a bit of estimate with reality. The dashboard shows how many months of real data sit behind each row, and flags the young ones as mostly projection. The rule the team follows: take a first read immediately, confirm it a month later.
Why supply is on a marketing dashboard
New investors can only invest if there is something to invest in. In months when few projects were open, people registered, found nothing that suited them, and left or invested much later. The ads had done their job and the cohort still looked weak.
So the dashboard carries the number of open projects next to every cohort. In the demo data, look at the months flagged in red: the cohort with the highest spend of the year barely pays back, because it was acquired into a thin shelf. That is a budget decision, not a creative problem. When supply is low, the right move is to spend less, and the table makes that visible.
How the team uses it
Pacing is read against the monthly target. Cohorts are read for whether the spend pays back. The two are never compared against each other.
In practice that means a monthly budget conversation with evidence behind it. If a cohort lands well above 1.00× with healthy supply ahead, there is room to push. If supply is about to drop, the budget comes down before the money is wasted, not after.
What to take away
- 1
Judge the month by the cohort it acquired, not by the month’s revenue.
In a business where customers keep investing for years, this month’s revenue mostly comes from people acquired long ago. The only fair test of this month’s marketing is what this month’s cohort goes on to do.
- 2
Predict lifetime value early, then let reality overwrite the prediction.
Nobody can wait two years to learn whether a month paid off. Project each cohort from the shape older cohorts followed, and replace estimate with real data every month. Take a first read straight away and confirm it a month later.
- 3
Put the business constraint next to the marketing numbers.
Here it was supply: how many projects were open to invest in. A cohort acquired into an empty shelf converts badly no matter how good the ads were. Without that column the team would have blamed the campaigns.