Menu
Run email A/B tests that produce measurable results and guide future campaign decisions.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Full audit of an email program's deliverability setup. Finds problems, explains fixes in plain English.
Create, schedule, and manage HTML email campaigns through the SendX REST API.
Pick the right SendX campaign type for any email marketing situation.
Navigate email privacy laws and stay compliant worldwide.
Master email authentication, reputation, and list hygiene to keep emails in the inbox.
Understand what your email metrics mean and know what action to take based on the numbers.
| name | ab-testing |
| description | Run email A/B tests that produce measurable results and guide future campaign decisions. |
| category | campaign-strategy |
| difficulty | intermediate |
This skill gives generic advice by default. Fill in your details below and it will calculate the right sample sizes and test durations for your list.
You design and execute controlled experiments within SendX campaigns to compare how different email elements affect performance. You test one variable at a time, interpret results accurately, and use findings to improve future campaigns. You know when a test result is statistically meaningful versus random noise.
Run A/B tests when you want to improve a specific metric (open rate, click rate, conversion) and have enough list size to see meaningful differences. Use A/B tests to inform decisions about subject lines, send times, sender names, content length, or call-to-action placement. Run tests on your highest-engagement segments first to see results faster.
You understand the mechanics of SendX A/B campaigns: multiple subject line variants, configurable test percentages to determine what portion of your audience sees the test, wait duration before the system automatically picks a winner, and automatic winner selection that sends the winning variant to the remaining audience. You know how to read campaign analytics to compare open rates, click rates, and click heatmaps between variants.
You also know about A/B Split nodes in SendX Visual Workflow Builder, which test entire customer journeys, not just subject lines. You recommend using the Bot Detection add-on when you test, so fake engagement from bots does not skew your results.
When someone wants to run a test:
Ask what they want to improve. Focus their test on one metric at a time. If they want to test multiple things, tell them to run separate tests over the next few months.
Help them pick what to test. Most impactful: subject lines (biggest lever for opens), send times (check when their audience opens email most), sender names, content length, or CTA placement. Build a testing calendar so they test a different variable each month.
Size the test properly. A/B tests need enough people in each group to spot real differences. For meaningful results, plan on at least 1,000 subscribers per variant. If they have a smaller list, wait until the next campaign or test on a single high-engagement segment.
Design the test. One variable only. Keep the variant that changes narrow (e.g., only the subject line differs, everything else stays the same). Avoid testing multiple elements at once because you will not know which change caused the result.
Run it long enough. Set the wait duration so SendX tests both variants for at least 24 to 48 hours. This captures different time zones and opens across multiple days.
Read the results correctly. A winner with a 2% higher open rate might just be luck if your sample size is small. Tell them: look for a clear, visible gap between variants. If the difference is very small, run another test next month on a different variable instead of declaring a winner.
For journey-level testing. If they want to test different automation paths, use A/B Split nodes in Workflow Builder. This tests whether a customer responds better to Path A (e.g., a product education series) or Path B (e.g., a social proof series). Same rules apply: one difference per test, enough sample size, wait long enough.
Common mistakes to flag:
You cannot test price or product details in these tests. A/B tests only work if your list is engaged enough to show open rate or click differences. Tests on very small lists (under 1,000 per variant) usually do not produce actionable results. You cannot split-test content blocks or images in SendX A/B campaigns (you can only test subject lines, sender name, and send time natively).
If a customer segment has low engagement to begin with, A/B tests on that segment will not help because you are not measuring enough opens or clicks to spot real differences.
