# Compare personalization with a holdback

You'll compare your personalized page with the original for the same audience, then use the campaign's decision state to judge whether the change helped.

Available on the Scale and Pro plans.

Find out whether your personalized page helps more people convert than the original. A holdback keeps some visitors on the original page, so you can compare people eligible for the same campaign. Holdback groups require **Pro**.

For example, [WorkForge](https://workforge.rightmessage.com) could compare agency-specific homepage copy with its original homepage and measure demo requests. Audience variants decide which content fits someone. Holdback decides whether they receive the campaign's changes at all.

You'll need editing and publishing access, RightMessage installed on your site, and a [campaign with tested page changes](https://rightmessage.com/help/create-a-personalization-campaign). Campaign holdback is separate from [question, form, and flow tests](https://rightmessage.com/help/split-testing-questions-and-offers).

## 1. Choose what you'll measure

Attach a [conversion goal](https://rightmessage.com/help/goals-and-conversion-tracking) so RightMessage can compare the action you care about, such as a demo request.

1. Open the campaign's **Conversion Goals** tab.
2. Turn on **Tracking** for your goal, or choose **New goal** to create and attach it.
3. If you track several goals, choose **Make primary** for the one you want to use to judge the campaign.
4. Under **Judge this campaign on**, keep **Primary goal** when that choice appears.

Decide what matters before starting. Choosing whichever goal looks strongest afterward changes the question you're testing. **Any goal** is also available when you want to count a visitor once for completing any attached goal.

## 2. Set the holdback

1. Open **Split Tests** in the campaign.
2. Turn on **Run an A/B test**.
3. Set **How many visitors should see your default content?** to the share you want to hold back.
4. Review **What to expect** for the approximate traffic in each group.

![Campaign Split Tests settings with Run an A/B test enabled and 20 percent of visitors assigned default content](https://rightmessage.com/help/screens/personalize-holdback.light.webp)

New campaigns start with testing enabled and 10% holdback. RightMessage can recommend a higher share for lower-traffic campaigns so both groups collect evidence.

> **Don't use 0% to turn testing off**
>
> An enabled test set to 0% currently uses a 10% fallback. To stop new holdback assignments, turn off **Run an A/B test** and publish. Existing browser assignments stay in place.

## 3. Publish the test

[Review and publish](https://rightmessage.com/help/publishing-your-changes) the full project draft once your page changes and test settings are ready.

The debugger shows the personalized page without changing your browser's actual holdback assignment. Use it to check your content, but remember that an ordinary visit can still receive the original page.

## 4. Read the comparison

1. Open **Results** after the campaign is published and enabled.
2. Choose the dates you want to inspect.
3. Compare the personalized and holdback visitor counts, conversion rates, and goal conversions.
4. Expand **Who it converts for** to see the combinations of personalized content visitors matched.
5. Read the decision state before you change the campaign.

![The first four campaign experiences with personalized and holdback conversion rates, visitor counts and lift](https://rightmessage.com/help/screens/reporting-campaign-experiences.light.webp)

You'll see the comparison build as visitors arrive and convert. A large observed lift isn't automatically a **Winner**. Decisions use scheduled checks of published experiment versions, not the dates you select in the report.

Next, read [what campaign decisions mean](https://rightmessage.com/help/campaign-experiment-decisions) before deciding whether to keep the personalized page for everyone.