What is A/B Test Duration?
This converts a required sample size into a timeline in days at your actual traffic level. It is the step that turns an abstract statistical requirement into a calendar commitment, and it is frequently where a test plan quietly dies — a sample that sounds achievable can imply four months of running, during which the page is frozen and the seasonality has shifted underneath it. Knowing that before you start is worth far more than discovering it in week six.
Why it matters
- It prevents planning a test that would take months to reach significance without realizing it upfront.
- It helps decide whether to test a bolder change (needing less traffic to detect) or wait longer.
- It sets an honest timeline expectation for stakeholders before the test even starts.
The formula
Days Needed = Required Sample Size per Variant ÷ Daily Visitors per Variant
- Required Sample Size per Variant
- — from the A/B Test Sample Size calculator
- Daily Visitors per Variant
- — your current daily traffic to the page, split across variants
How to use this calculator
- 01Run the A/B Test Sample Size calculator first to get your required sample size per variant.
- 02Enter your current daily visitors, divided evenly across variants.
- 03The result is the number of days the test needs to run.
- 04If the result is impractically long, consider testing a bigger change or running the test on higher-traffic pages.
Worked example
A test needs 8,146 visitors per variant, and the page gets about 500 visitors per variant per day.
- Required sample = 8,146 visitors per variant
- Daily visitors reaching each variant = 500
- Days needed = 8,146 ÷ 500 ≈ 16.3 days
- Round up to 21 days so the test covers three full weekly cycles
This test needs about 16-17 days to run — a reasonable, practical timeline for most testing programs.
Industry benchmarks
Compiled August 2026
High-traffic pages
Under 2 weeks
Fast iteration is possible.
Moderate-traffic pages
2 – 4 weeks
Standard testing window for most sites.
Low-traffic pages
1 – 3+ months
Often impractical — consider testing bigger changes instead.
Practical upper limit
About 4 weeks
Beyond this, cookie loss and seasonal drift start corrupting the result.
Common mistakes
- Ending a test at an arbitrary calendar date (e.g. 'two weeks') instead of the calculated required duration.
- Not accounting for weekly traffic patterns — running a test for less than one full week cycle can bias results.
- Running a test on a low-traffic page when a higher-traffic page could answer the same underlying question faster.
- Letting a test run for months to reach its sample — visitors clear cookies and get recounted, and the market itself changes, so a very long test measures something other than the change you made.
How to improve your test duration
Run tests on higher-traffic pages when possible
Testing a checkout-flow change on your highest-traffic entry point reaches significance far faster than a low-traffic page.
Test bigger, bolder changes on low-traffic pages
Since larger expected effects need dramatically less sample size, low-traffic pages should prioritize bold hypotheses over subtle tweaks.
Recheck required sample size →Decide the stop date before launching
Fixing the end date in advance removes the temptation to stop the moment a variant looks ahead, which is the single most common way a testing programme produces confident wrong answers.
Frequently asked questions
Should I round up to full weeks?+
Yes — running for full week increments (e.g. 14 or 21 days, not 16) avoids bias from weekday/weekend traffic pattern differences.
What if my traffic is highly seasonal?+
Avoid testing across a major seasonal shift (like a holiday) unless that's specifically what you're studying — it distorts the baseline.
Should the duration count all site traffic or just this page?+
Only the traffic that actually reaches the tested page and gets bucketed into a variant. Using a site-wide figure is the most common reason a test that was supposed to take two weeks is still running after two months.
Is there a minimum duration regardless of sample?+
Run at least one full week even if the sample arrives sooner. Buying behaviour differs sharply between weekdays and weekends, and a test that only saw Tuesday to Thursday has measured a specific slice of your audience rather than all of it.
Related calculators
Put this calculator on your site
Free to embed, no attribution required beyond the link the widget already carries. It posts its height to the parent window as marketing-calculators:height if you want to size the frame exactly.
<iframe src="https://markcelerate.com/embed/ab-test-duration-calculator" title="A/B Test Duration Calculator" width="100%" height="520" style="border:1px solid #ddd;border-radius:6px" loading="lazy"></iframe>