How to Run A/B Tests on Email Campaigns in UniLink (Test Subject Lines and Content)

Stop guessing what resonates with your audience. Split your list, test subject lines or full email content, set a winner criterion, and let UniLink automatically send the winning version to the rest of your list.

Most email marketers pick subject lines based on intuition. A/B testing replaces intuition with data. When you test a direct subject line against a curiosity-based one — or a plain-text email against an image-heavy version — you find out what your specific audience actually responds to, not what a marketing blog says should work. Even a modest improvement in open rate compounds significantly over time: a 5-point open rate lift across 12 campaigns per year is a meaningful difference in revenue and engagement.

UniLink's A/B testing is built directly into the campaign builder, so you do not need a third-party tool or a statistics background to run a clean test. This guide covers the full A/B testing workflow from setup to interpreting results.

What A/B Testing Does

When you enable A/B testing on a campaign in UniLink, you create two variants: Variant A and Variant B. UniLink splits a portion of your recipient list randomly — by default 20% goes to Variant A and 20% to Variant B. The remaining 60% of your list is held back. After the test duration ends (you choose between 1 hour and 72 hours), the system evaluates both variants against your chosen winner criterion: open rate or click rate. The variant with the better result is then sent to the remaining 60% of your list automatically.

You can test two types of variables. Subject line testing keeps the email body identical across both variants and only changes the subject line — isolating the impact of your subject line on open rates. Content testing uses the same subject line but sends different email bodies to each group — useful for testing a promotional offer against a value-first email, or a text-heavy layout against a visual one. Testing one variable at a time is the correct methodology; testing both simultaneously makes it impossible to know which change drove the result.

Results are reported in the Campaign Reports section with per-variant breakdowns of sends, opens, clicks, and the winning margin. The winning variant, the margin of victory, and the time at which the winner was automatically sent to the remaining list are all recorded for future reference.

How to Get Started With A/B Testing

1. Create a new campaign — Go to Email Marketing → Campaigns and click New Campaign. In the campaign setup, fill in the campaign name and select your recipients (segment or full list).
2. Enable A/B testing — In the campaign settings panel, find the A/B Test toggle and turn it on. The composer expands to show two variant tabs: Variant A and Variant B.
3. Choose what to test — At the top of the A/B section, select Subject Line Test or Content Test. If you select Subject Line, you will write the same body but enter two different subject lines. If you select Content, you write two separate email bodies with the same subject line.
4. Compose both variants — Click Variant A and write your first version completely. Then click Variant B and write the second. For subject line tests, only the subject field differs. For content tests, keep the subject field identical across both tabs.
5. Set the test split — The default 20/20/60 split (20% to A, 20% to B, 60% held for winner) works well for most list sizes. If your list is small (under 500), consider increasing the test split to 40/40/20 to get statistically meaningful results.
6. Choose the winner criterion — Select Open Rate (best for subject line tests) or Click Rate (best for content tests). The winner criterion should match the variable you are testing.
7. Set the test duration — Choose how long the test runs before a winner is picked. Common choices are 4 hours (time-sensitive campaigns), 24 hours (standard campaigns), or 48 hours (weekend sends where behaviour patterns differ by day). Click Schedule or Send Now to start the test.

How to Use A/B Testing

1. Monitor results during the test period — Go to Email Marketing → Campaigns and open your active A/B test campaign. The results panel shows live open and click counts for both variants as they accumulate. Do not manually pick a winner early — let the test complete its full duration for a fair result.
2. Wait for automatic winner send — When the test duration ends, UniLink compares both variants on your winner criterion and sends the winning variant to the remaining held list automatically. You will receive a notification confirming which variant won and when the winner send completed.
3. Review the results report — Go to Campaign Reports and open the A/B test campaign. The full report shows variant-level metrics: sent count, open rate, click rate, unsubscribe rate, and the winning margin. Save or screenshot this for your testing log.
4. Log your findings — Keep a simple spreadsheet tracking what you tested, which variant won, and by how much. Over 10–15 tests you will start to see patterns specific to your audience: your readers prefer question-based subject lines, or they open more when you personalise with their first name, or they convert better from plain-text emails.
5. Apply winning patterns to all future campaigns — If curiosity subject lines consistently beat direct ones by more than 5 points, use curiosity framing as your default going forward. A/B testing is only valuable if the learnings change how you write every campaign, not just the one you tested.
6. Test one thing at a time — Resist the temptation to change multiple elements between variants. If Variant A has a different subject line, preview text, and opening paragraph compared to Variant B, you cannot know which change caused any difference in performance.
7. Re-test important variables after list growth — A test result from when your list was 500 contacts may not hold at 5,000. As your audience changes, revisit key tests — subject line tone, plain vs HTML email — to see if conclusions still hold.

Key Settings Explained

How to Get the Most Out of A/B Testing

The most common reason A/B testing fails to deliver useful insights is running tests on lists that are too small. With a 20/20 split on a list of 200 subscribers, each variant goes to just 40 people. A difference of 2 opens between the variants is statistically meaningless — it could be random noise. As a rule of thumb, each variant should receive at least 200–300 emails for the results to be meaningful. If your list is smaller, use a 40/40 split or wait until your list grows before investing heavily in formal testing.

Treat A/B testing as a long-term learning programme rather than a single experiment. One test tells you what worked once. Ten tests across different campaign types, subject line styles, content formats, and audience segments tells you what your audience consistently prefers. Create a testing calendar — for example, running one structured A/B test per month — and track all results in a single document. After six months you will have a data-backed style guide for your email programme that is specific to your audience, not borrowed from someone else's audience.

Do not stop at subject lines. Content A/B tests are underused but often more valuable. Test a long-form narrative email against a short punchy one. Test an email with one clear call-to-action against an email with three options. Test starting with the offer versus starting with the story. Click rate differences in content tests can reveal a great deal about how your audience prefers to be sold to — information that is hard to get any other way.

Use your A/B test results to personalise at scale. If you discover that one audience segment responds to curiosity subject lines while another responds to direct benefit statements, create separate campaigns for each segment with the appropriate framing. This is not additional work — it is applying what the data already told you, segment by segment.

Troubleshooting Common Issues