# Out of Stock: Remove the Button or Notify Me A/B Test

Page: https://alidemirbas.com.tr/en/lab/ab-test-library/tests/out-of-stock-remove-the-button-or-notify-me-ab-test

## Hypothesis

If the purchase button gives way to a "notify me when it's back" option, interested visitors may follow the product instead of leaving it; a plain "out of stock" label explains the situation and leaves the next step to them. The notify arm could build up back-in-stock notification signups, but that demand only pays off if it turns into purchases after restock.

## When to run this test

Steady traffic to sold-out items, support requests asking when a product will be back, or quick exits from out-of-stock pages mean interest in those products may be going to waste. If the notify path works, visitors in that arm leave a signup, and it brings more purchases once stock returns. If visitors want a similar product rather than this one, look at the alternative-product recommendations first. Before launch, check that any restock date is labeled as an estimate when it is one, and that the flow carries no urgency message unsupported by live inventory.

## Primary metric

**Purchase rate after restock**: After the product returns to stock, compare purchases among all eligible assigned visitors over the same follow-up window.

## Other metrics

- **Back-in-stock notification signup rate**: Signups exist only in the notify arm, so they do not compare the two arms; they show how much the added path is used.
- **Alternative product page visit rate**: Measure visits to alternative product pages among assigned visitors who see an alternative-product link.
- **Notification complaints and unsubscribes**: Among users who consent to notifications, track complaints and unsubscribes over the same follow-up period.

## What to check before and during the test

- **Comparison**: Compare an out-of-stock status label with one defined back-in-stock notification option.
- **Channel**: Choose one notification channel and consent flow before launch.
- **Alternatives**: Keep alternative-product recommendations the same.
- **Stock timing**: If shown, use the same accurate availability estimate in both arms.
- **Device**: Check notification signup completion on mobile and desktop.

## What to watch when reading the result

- Signups exist only in the notify arm, so they cannot compare the two arms on their own. Make the main comparison on purchase rate after restock, across all eligible visitors assigned to each arm; a flow that collects many signups but brings notification complaints and unsubscribes is not a winner.
- If price, images, product details and alternative-product recommendations stay the same in both arms, the difference comes from the notify mechanism alone. Report alternative product page visits as a diagnostic, and do not narrow the main comparison afterward by who saw the link.
- Ask only for what is needed to send the stock notification; extra fields can lower signups for reasons that have nothing to do with the mechanism. Read form completion on mobile and desktop separately.
- Purchases after restock may be rare and can only be measured once the product is back. Find the sample each arm needs for that rate with the [sample size calculator](https://alidemirbas.com.tr/en/calculators/sample-size-calculator) up front, and plan the follow-up period around it. Should the gap still be small, leave the question open rather than naming a winning arm.

## Frequently Asked Questions

### Is the back-in-stock signup rate enough to compare groups?

No. The status-only control has no signup mechanism. Compare purchases after restocking across all eligible assigned visitors. Monitor notification complaints and unsubscribes too.

### What information should the notification form request?

Collect only what is needed. Record the user's request, provide the required privacy notice and send the requested notification. Handle any promotional-message consent and applicable Turkish IYS requirements separately. Do not merge disclosure with consent. Check completion by device and avoid unverified urgency.

### Should I show an expected restock date?

Show the same reliable estimate in both groups without presenting it as certain. Alternative-product clicks are diagnostic. Keep the common eligible entry population for the main purchase comparison rather than selecting people by later exposure.