# Product Measurement Line on PDP A/B Test Page: https://alidemirbas.com.tr/en/lab/ab-test-library/tests/product-measurement-line-on-pdp-ab-test ## Hypothesis If a line showing the product's real dimensions sits next to the product image, visitors may judge its physical size more accurately before they buy. That could lower the scale-related return rate, measured on orders that are past their follow-up window. ## When to run this test Return reasons that point to size expectations, product questions about measurements, or reviews saying the item looks different from the photo are signs that products in a category arrive larger or smaller than shoppers expect. If the line works, visitors grasp the size before buying and scale-related returns fall. If returns come from quality, color or damage, a line will not fix them; and if the page has no written dimensions, add those first, because the line does not replace them. ## Primary metric **Scale-related return rate**: Compare scale-related returns among orders with a completed follow-up window in each arm. Review the result alongside conversion and revenue. ## Other metrics - **Purchase conversion rate (CR)**: Compare purchases among all assigned visitors who see the measurement line. - **Product measurement questions**: Compare support questions about product size using the same traffic denominator. - **Revenue per visitor (RPV)**: See how total revenue changes even if the line brings in more orders. ## What to check before and during the test - **Measurement line**: Add a fixed line that shows the product's actual dimensions. - **Reference**: Do not switch between a line, object or person as the scale reference. - **Image slot**: Choose one image slot and keep the number of images the same. - **Readability**: Make sure the line does not obscure the product and that the measurement remains legible. - **Category**: Review categories where physical scale is relevant. ## What to watch when reading the result - A drop in scale-related returns does not justify shipping if purchase conversion or revenue per visitor falls with it, since the line may also change who buys. Make the call together with revenue per visitor, and compare product measurement questions using the same traffic denominator. - Count returns only for orders whose follow-up window has closed; a return rate read early is incomplete. This finding covers the tested category and image slot; do not carry it to categories where scale is not an issue. Before moving to another category where scale matters, validate it there again. - Throughout the test, check that the line shows the real measurement accurately, does not make the product look larger than it is and does not cover the photo. A faulty line creates a new wrong expectation instead of reducing returns, and that spoils the result. - With a small or unclear gap, keep in mind that scale-related returns are a rare event and that you also have to wait out the follow-up window. Put the current return rate and the drop you want to detect into the [sample size calculator](https://alidemirbas.com.tr/en/calculators/sample-size-calculator), then divide the result by daily orders per arm, and allow for the follow-up window on top. Save a familiar object or person as the scale cue for a separate variant later. ## Frequently Asked Questions ### Are dimension lines and reference objects equivalent? No. Dimension lines provide measurements, while familiar objects or people provide visual scale. This test uses lines with written dimensions. Define other references as separate variants. Use the scale-related return rate as the primary outcome, and track conversion and revenue per visitor alongside it. ### Do fewer size-related returns prove the lines worked? Compare mature return outcomes alongside conversion and revenue. The lines may also change who buys. Revenue per visitor shows the combined revenue effect even when order volume changes. ### How can I check whether dimension lines are accurate and readable? Use real measurements, preserve readable written dimensions and avoid obscuring the photo or making the product look larger. Lines must convey information rather than act as decoration. ### Can I generalize the result to every category? No. Focus on categories where physical scale is uncertain. Compare size-related support contacts using the same traffic denominator.