About the Analysis
We visually coded 52 broadcast emails (30 top-RPR + 22 bottom-RPR) across 19 brands in the portfolio, then compared the prevalence of each creative element between the top and bottom quintiles. Sample drawn from 4,000+ campaigns across 40+ accounts, trailing 12 months.
The leverboard
Sorted by absolute delta. Positive bars (blue) show up more often in winning emails. Negative bars (charcoal) show up more often in losing emails. The bigger the bar, the stronger the signal.
Creative Lever Deltas (percentage points)
Source: Sweat Pants Agency portfolio · 52 broadcast emails coded across 19 brands · trailing 12 months
The CTA finding
Strong action-verb CTAs that name the action and tie to the offer (“Shop the Sale”, “Get Yours”, “Claim My Discount”) appear in 60% of top-RPR emails and 13.6% of bottom-RPR emails. Medium CTAs (“Learn More”, “Discover”) appear in 72.7% of bottom-RPR emails and 36.7% of top-RPR ones. The CTA copy is doing more work than the design.
| Strength | Top % | Bottom % | Delta |
|---|---|---|---|
Strong "Shop the Sale", "Get Yours", "I Want to Win", "Claim My Discount" | 60.0% | 13.6% | +46.4pp |
Medium "Learn More", "Find Out More", "Discover", "Explore" | 36.7% | 72.7% | -36.1pp |
Weak "Read the Blog", "View Menu", "Click Here" | 3.3% | 13.6% | -10.3pp |
Source: Sweat Pants Agency portfolio · 52 emails coded across 19 brands · trailing 12 months
The composite finding (the one that matters most)
Clear offer + strong CTA together. 57% of top-RPR emails have both. 14% of bottom-RPR emails have both. +43pp delta.
This is the single largest predictor of revenue per recipient in the dataset. We weight it higher than any individual lever in our audit framework because the combination is what does the work. A strong CTA on a vague offer underperforms. A clear offer with a weak CTA underperforms. The two together are what move the number.
Length doesn't matter. Density does.
Most teams optimize email length: “is shorter better, or longer?” The data says neither. Long emails win (+4.5pp), medium emails win (+13.9pp), short emails lose (-7pp), and very long emails lose (-11.5pp). The signal is messy because length isn't the variable. Density is.
Length buckets
| Bucket | Delta |
|---|---|
| Short | -7pp |
| Medium | +13.9pp |
| Long | +4.5pp |
| Very long | -11.5pp |
Mixed signal. Length isn't the lever.
Density buckets
| Bucket | Delta |
|---|---|
| Very low | -4.5pp |
| Low (sparse) | -15.2pp |
| Medium | +28.2pp |
| High (verbose) | -8.5pp |
Medium density wins by +28.2pp.
Medium density (right ratio of text to product/image elements, paragraphs broken up by visuals and CTAs) beats both verbose walls of text and sparse decoration-only emails by a wide margin. This is what to A/B test, not length.
What we excluded from the leverboard
Two patterns we expected to find didn't make the cut. We excluded them from our audit grading because the data was too noisy to be actionable.
- Layout (designed vs plain text): No meaningful delta. Designed emails win, plain text emails win, designed emails lose, plain text emails lose. Layout isn't the lever. Stop A/B testing it as if it were.
- Email length in isolation: Mixed signal as shown above. Length is correlated with both wins and losses depending on density, so we don't grade it directly.
Why this matters
Most retention teams we audit have an A/B testing backlog that includes “short vs long email”, “designed vs plain text”, “more product vs less product”. The data says those tests have low signal. The high-signal tests are CTA copy strength, density, and the offer + CTA composite. Reallocate the testing calendar to the levers that actually move the number.
The most leveraged single change most teams could make tomorrow: audit the CTAs across the last 30 days of campaigns. Count how many say something like “Shop the Sale” or “Get Yours” versus “Learn More” or “Discover”. If the latter outnumbers the former, that's the biggest revenue lever sitting unaddressed.
What to do Monday
- Audit your CTA copy across the last 30 days. Count strong, medium, weak. If strong is below 50% of your campaign CTAs, that's your highest-leverage rewrite.
- Apply the composite test to your top 10 campaigns. Did the campaign have a clear offer AND a strong CTA? If only one of the two, that's the most likely revenue leak.
- Remove length from your test backlog. Replace it with density tests. Run a medium-density variant against your sparse and verbose variants. Measure RPR.
- Test header/nav removal on promotional sends. Run a no-nav variant against your standard send. The +21pp delta on header presence suggests removing it for promo could be a meaningful lift.
- Stop running giveaway-only campaigns. Giveaway as offer (-12.7pp) underperforms percent-off campaigns. Run the percent-off variant and keep giveaways for list-growth contests, not revenue sends.
How we use this in our audits at Sweat Pants Agency
When we audit broadcast creative, we score each campaign against the leverboard. Strong action-verb CTA gets the largest weight. Composite (offer + CTA) gets the second largest. Density, urgency, scarcity, founder voice, personalization, and social proof fill out the rest. We don't grade layout or length in isolation. The audit output is a heatmap of which levers are pulling and which aren't.
Want to see how your account stacks up?
Get a free retention audit from the team behind 4.3 billion sent emails and $2B+ in client revenue.
Get Your Free AuditRelated Reports