Reverse-Engineer the Bottleneck: CR, CPA, and a Misleading 20% Headline
Objective: Given only raw ad spend, traffic, and revenue numbers with no conversion rate stated, back into CR and CPA, then use relative-vs-absolute lift framing to diagnose whether spend, conversion rate, or AOV is the real problem.
Casper Sleep's paid social channel doubled ad spend last quarter, but revenue only grew 20%. Marketing is calling it a win. You have the raw numbers, not a labeled report.
Compute before you diagnose. Don't accept the '20% CR lift' headline until you've checked it against the absolute numbers.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, formula-driven, sufficient for a diagnostic memo
Free, standard source for the raw traffic and conversion data used in the calculation
No access? Use the numbers as given in the scenario if GA4 access isn't available
The process
2 steps
Step 01 of 02
CPA = Ad Spend / Conversions. As CR rises with traffic and spend held flat, more of the same visitors convert, so CPA drops.
Last quarter: $18,000 ad spend, 500,000 visitors, 3,000 conversions. What is the conversion rate and CPA?
Procedure
- CR = (3,000 / 500,000) x 100
- CPA = $18,000 / 3,000
- Record both figures
CR = (3,000 / 500,000) x 100 = 0.6% CPA = $18,000 / 3,000 = $6.00
Healthy
CR and CPA are computed from the same period's raw numbers before any comparison is made.
Unhealthy
Accepting a stakeholder's stated percentage without recomputing it from the underlying visitor and conversion counts.
What this means
A 0.6% CR and $6 CPA are the real baseline this quarter's '20% growth' claim needs to be checked against.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| A report states a growth percentage with no CR or CPA shown | Recompute both from raw visitor/conversion/spend numbers before trusting the headline | 5 min |
Step 02 of 02
A headline percentage lift can sound extraordinary while the absolute percentage-point change is small. Always anchor percentage gains to absolute numbers and real dollars.
Prior quarter's CR was 0.5%; this quarter's CR (computed above) is 0.6%. Marketing calls this a '20% CR lift.' At $150 AOV and 500,000 visitors, what is the absolute lift in percentage points and in dollars, and was doubling ad spend worth it?
Procedure
- Relative lift = (0.6% - 0.5%) / 0.5% = 20%
- Absolute lift = 0.6% - 0.5% = 0.1 percentage points
- Dollar impact = 500,000 x 0.001 x $150 = $75,000 in added revenue
- Compare $75,000 in added revenue against the $9,000 increase in ad spend (from doubling $9,000 to $18,000)
Relative lift: 20% (sounds large) Absolute lift: 0.1 percentage points (0.5% -> 0.6%) Dollar impact: 500,000 x 0.001 x $150 = $75,000 Extra spend: +$9,000 Verdict: revenue grew $75,000 against $9,000 in added spend, a real win, but CPA still rose from prior levels because spend outpaced the CR gain, not a pure CR success story.
Healthy
The team reports both the relative and absolute figures together with the dollar impact and the spend delta.
Unhealthy
The team repeats '20% lift' in every slide without ever stating the 0.1-percentage-point absolute change or the CPA trend.
What this means
The real story is a modest CR improvement riding on top of doubled spend, not a CRO breakthrough. The bottleneck is spend efficiency, not the page itself.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| A results deck leads with a relative percentage only | Require the absolute percentage-point and dollar figures on the same slide | 5 min |
Final deliverable
A short diagnostic memo stating the computed CR, CPA, and RPV, the relative-vs-absolute lift comparison, and a verdict on whether the real Q3 problem is spend, conversion rate, or AOV.
See a reference example
Lenskart paid search diagnostic (excerpt) CR: 1.4% (up from 1.2%) | CPA: $11.90 (up from $9.80) Relative lift: 16.7% | Absolute lift: 0.2 percentage points Dollar impact: 300,000 visitors x 0.002 x $2,200 AOV = $132,000 Verdict: CR improved, but CPA rose faster than revenue, meaning the real bottleneck this quarter is rising traffic cost, not conversion rate.
Success criteria
You're done when you can:
- CR and CPA are correctly computed from raw spend/traffic/conversion inputs, not assumed
- Relative and absolute lift are both stated, with a dollar figure attached
- The verdict correctly identifies spend efficiency, not conversion rate, as the actual driver