Landing Page A/B Test: Outcome vs Capability Positioning
A landing-page A/B test template that compares outcome-focused positioning ("2x pipeline in 30 days") against capability-focused copy ("AI experimentation platform"). The test isolates positioning as the cause of any conversion difference, with Day-7 activation as the lead-quality guardrail so a higher signup rate does not mask a worse cohort.
Landing pages have the widest variance of any experiment category. A single positioning change can double conversion or kill it. Two disciplines protect you from a misleading result: a real lead-quality guardrail, and segment-level analysis before averaging.
Copy the template
Use it in Notion, a Google Doc, or wherever your team already works.
# Landing Page A/B Test: Outcome vs Capability Positioning
## Hypothesis
Because [observation about conversion or message-market fit], we will replace capability-focused headline copy with outcome-focused copy, and expect signup conversion rate to lift for [audience] with equal or better Day-7 activation. We will ship at +15% conversion with no Day-7 regression.
## Variants
- Control (A): Capability-focused headline ("AI experimentation platform").
- Variant (B): Outcome-focused headline ("Hit your growth targets without hiring").
## Metrics
- Primary: Visitor-to-qualified-signup conversion rate.
- Guardrails: Day-7 activation, bounce, time on page, scroll depth, form abandonment.
- Segments: paid vs organic, desktop vs mobile.
## Math
- Sample size: ~500 conversions per variant for 15% relative effect, 80% power, 95% confidence, 3% baseline signup rate.
- Duration: 3-4 weeks, full business cycles.
## Common failure to avoid
Averaging across paid and organic. A variant that wins overall can lose decisively in one segment.
The variants
Capability-focused headline describing what the product does.
Example: AI experimentation platform for growth teams
Outcome-focused headline describing the result the buyer wants.
Example: Hit your growth targets without hiring
Metrics, math, and success criteria
Visitor-to-qualified-signup conversion rate.
Day-7 activation rate, bounce rate, time on page, scroll depth, form abandonment. Segments: paid versus organic, desktop versus mobile, ICP-fit score where available.
About 500 conversions per variant for a 15 percent relative lift at 80 percent power and 95 percent confidence on a 3 percent baseline signup rate. Translate to traffic with your current conversion rate.
At least a 15 percent relative lift in signup conversion rate with equal or better Day-7 activation for the resulting cohort.
Three to four weeks, or 500 conversions per variant, accounting for day-of-week effects.
Expected outcome range
Landing page positioning tests show the widest range of any category. Wins commonly land between 10 and 40 percent on B2B SaaS, but flat or negative is the modal result; treat any single win as a hypothesis to be replicated.
Common failure mode
Averaging across paid and organic traffic. They behave so differently that a variant that wins overall can lose decisively in one segment. Always segment before declaring a winner.
What this unlocks next
- If outcome positioning wins, propagate the same frame to ad creative, sales email subject lines, and the pricing page hero.
- If outcome positioning loses, the issue is usually in how outcome is defined. Re-test with a sharper outcome derived from customer research, not internal language.
- The losing variant still teaches you what your traffic does not respond to, which trims a class of future ad and email tests.
Running this template manually vs in GrowthLab
| Step | Manual (spreadsheet) | In GrowthLab |
|---|---|---|
| Draft variants | Write a new headline from scratch, often with internal language no buyer uses. | AI drafts outcome-focused copy from your customer research, with the audience and segment pre-filled. |
| Prioritize | Compete the landing test against other ideas with no shared scoring. | ICE pre-scored (Impact 9, Confidence 6, Ease 5). Reach-adjusted view shows the audience size next to the score. |
| Run and track | Drop a tracking pixel, hope the numbers come out clean, forget to segment by source. | Paid vs organic and desktop vs mobile segmentation wired automatically. The Day-7 activation guardrail tracks alongside the primary metric. |
| Capture the learning | The losing variant is forgotten in two weeks. | The learning, win or flat, is stored with the test's hypothesis and segment data, so the next positioning test inherits the prior. |
Inside GrowthLab
Inside GrowthLab the template is pre-wired with paid-versus-organic segmentation, the Day-7 activation guardrail set as a tracked metric, and sample-size math computed against your live signup rate. ROTI weighs the learning against the build time on the variant page.
Frequently asked questions
How long should a landing page A/B test run?
Three to four weeks or 500 conversions per variant, whichever is longer. Landing-page tests are sensitive to day-of-week and traffic-source mix; running shorter than a full business cycle invites false positives.
Why segment landing page tests by traffic source?
Paid and organic visitors have different intent, expectations, and conversion baselines. A variant can win on organic and lose on paid, or vice versa. Averaging masks the difference and leads to the wrong decision.
What is a good baseline for landing page signup conversion?
On B2B SaaS, 2 to 5 percent is a common range for a cold-traffic landing page leading to a self-serve signup; warm or branded traffic runs higher. Calculate sample size against your own baseline rather than an industry number.
Go deeper
- What is an A/B test
- What is an experiment hypothesis
- Growth experimentation: the 2026 ultimate guide
- Experiment prioritization guide
About GrowthLab
GrowthLab is an experiment management tool where AI drafts the hypotheses, ICE and ROTI prioritize them, and every learning compounds into the next batch.
