A baseline is the current performance level of a page, element, or user flow before any changes are made — the metric value your control is producing right now. In the context of A/B testing, the baseline conversion rate (or baseline AOV, baseline revenue per visitor) is the starting point from which lift is calculated, and it is the single most important input when sizing an experiment. Without an accurate baseline, you cannot calculate how long a test needs to run or what effect size you can realistically detect.
Baseline Conversion Rate = (Conversions / Total Visitors) × 100
To get a reliable baseline:
- Pull data from the last 30–90 days for the specific page or funnel step.
- Exclude any promotional periods, flash sales, or traffic anomalies.
- Segment by device if your mobile and desktop conversion rates differ significantly — your test may need separate baselines for each.
Why Baseline Matters for Ecommerce
The baseline is the denominator that makes lift meaningful. A 10% relative lift means something very different if your baseline is 1% (you'd go from 1% to 1.1%) versus 5% (you'd go from 5% to 5.5%). For a D2C Shopify brand, a more precise baseline directly improves experiment planning: you avoid running tests for too long (wasting time) or too short (producing unreliable results). Brands that track baselines at the device, traffic-source, and page level can also identify which segments have the most room to grow — those are the most valuable areas to test first.
Real-World Example
Nykaa's growth team was planning a test on their skincare category page. Before designing the variant, they pulled 60 days of data and found that the category page's add-to-cart rate was 4.3% on desktop and 2.9% on mobile. Rather than testing a single variant across all traffic, they ran separate experiments for desktop and mobile, each with a properly sized sample based on its own baseline. The mobile test required nearly twice the traffic to reach significance because of the lower baseline rate — a planning insight they would have missed had they averaged the two segments together.
How to Use Baseline Effectively
- Measure baseline on the specific metric you plan to optimize, not a proxy — if you're testing checkout, use checkout completion rate, not overall site conversion.
- Use at least 4 weeks of historical data to account for weekly traffic cycles.
- Recalculate the baseline if significant site changes occurred in the lookback period (new promotion, nav redesign, traffic source shift).
- Feed the baseline into a sample size calculator before starting the test so you know the minimum run time.
- Document your baseline alongside the test hypothesis so post-test analysis is clean.
Baseline in A/B Testing
The baseline feeds directly into power calculations. A lower baseline conversion rate means each visitor carries less statistical signal, requiring a larger sample size to detect the same relative lift. Tools like Evan Miller's sample size calculator or built-in calculators in testing platforms all ask for the baseline rate as their first input.
Run smarter A/B tests with CustomFit.ai — 14-day free trial, no credit card required.