# Product recommendations on Underoutfit product pages: +3.2% value

- Outcome: Won
- Page type: Product page (Ecommerce)
- Elements: Upsells & cross-sells
- Test run by: [Visually](https://liftarchive.com/practitioners/visually/)
- Commentary by: LiftArchive (our analysis, not theirs)
- Original source: https://143410811.hs-sites-eu1.com/en/underoutfit_casestudy_visually
- Page: https://liftarchive.com/test/underoutfit-pdp-product-recommendations/

Adding product recommendations to Underoutfit's product pages lifted per-session value by 3.2%. This test ran on Visually's platform with the Underoutfit team and was published as a Visually case study. The recommended products were chosen automatically for each shopper by an AI-based engine.

## Result

**+3.2% value**

- Per-session value: +3.2%.
- Conversion rate and average order value were not reported on their own, so it's unclear whether the gain came from more orders or bigger ones.
- Sample size, test length and confidence level were not shared.

## Control

Product pages showed no product recommendations.

## Variant

A product recommendations section was added to product pages, with items picked automatically for each shopper by an AI-based engine.

## Why it likely won (our reading)

- Lingerie shoppers often buy matching or related pieces, so relevant suggestions can add items to the order.
- Recommendations give shoppers who don't like the current product another option instead of leaving.
- Picking items per shopper likely made the suggestions more relevant than a fixed list.

## Caveats

Vendor-reported in a Visually case study; sample size, test length and statistical significance were not shared.

## How to run this test

- Hypothesis: If we add personalized product recommendations to product pages, then revenue per session will rise because shoppers find related items to add or better options to buy.
- Primary metric: Revenue per session
- Guardrail: Conversion rate
- Guardrail: Add to cart rate for the main product
- Guardrail: Page load time
- Who should test it: Stores with a wide range of related products, such as sets, colours or complementary items. It may do little for stores with only a few products.

---

Credit: Visually ran this test and owns the result. Summary and commentary by LiftArchive (https://liftarchive.com/test/underoutfit-pdp-product-recommendations/).
Usage: Free to use, quote and build on, with credit and a link: name the practitioner or organization that ran the test and link to its LiftArchive page. The results belong to the practitioners who ran and shared them; the summaries and commentary are LiftArchive's.