# Understanding Rematches

Rematches run within a CampaignMatch. If you haven't set one up yet, start with [Understanding CampaignMatches](https://messagematch.gitbook.io/messagematch-docs/introduction/key-terms/understanding-rematches)

### What is a Rematch?

Rematches are MessageMatch’s built-in split tests for a single UTM rule set. Turn them on to evenly test multiple copy variants for each CampaignMatch, then use Analytics to see which variant wins.

### When to use?

* You already have a working **CampaignMatches** and want to improve them.
* You want to test underlying core motivators for your audience segments.
* You need clean, per-campaign learnings without rebuilding pages.
* You want the dashboard to surface variant-level insights and recommendations.

### How Distribution & Caching Work

* **Variants:** Up to **5** versions (A–E) per UTM rule set.
* **Traffic split:** Even distribution across active variants.
* **Return visits:** A visitor who saw a variant for a given page link continues to see that same version on later visits to that link (keeps experience consistent).
* **Important:** Rematches only run **after the required UTM rules match,** though they can also be set to operate on the base website copy as well.

### Enable in a MatchAgent

1. Dashboard → \[**MatchAgents tab**] → select your MatchAgent → click **Manage**
2. Confirm UTM rules (and any referrer qualifier).
3. Toggle **Rematches ON** → add up to 5 variants → write copy → **Save**.
4. **Deploy** the MatchAgent.

### QA & testing

* Use a test URL with matching UTMs (e.g., `?utm_source=google&…`).
* To see a different variant during QA, use the dashboard’s variant switcher or test in a fresh/incognito session.
* If nothing changes: check required UTM keys/values, element mapping, and that variants are **enabled**.
* **Do I need to rebuild pages?** No—variants update mapped elements only.
* Rematches only update the elements mapped in your MatchAgent — no page rebuilding required. All other content remains unchanged.

### Analytics & recommendations

* **Variant reporting:** view count, CTR, conversions, and a simple **win-rate** per variant.
* **Roll-ups:** Views → Eligible Views → **CampaignMatch Views** with a filter for variant A–E.
* **Insights:** the dashboard highlights top/underperforming variants and suggests **next actions** (e.g., extend/pause, rebalance, or rotate variants) based on lift and traffic patterns.

### Plans & Pricing

* Rematches are available on Pro plans and higher; max **5** variants per UTM rule set.&#x20;
* Usage still counts against monthly **Views**; variant impressions are shown for diagnosis, not billed separately.
* See [**Plans & Pricing**](https://messagematch.gitbook.io/messagematch-docs/plan-limits-and-billing) for caps/overages.

### Best practices

* Establish a working CampaignMatch with measurable traffic before enabling Rematches. Split testing without a baseline produces inconclusive data.
* Change only one element per variant (e.g., only the H1, or only the CTA) to isolate what's driving lift. Testing multiple elements simultaneously makes it impossible to attribute results.
* Wait for statistically meaningful traffic before drawing conclusions. Low-volume results will show in Analytics but should not drive decisions.
* Use the dashboard's variant switcher during QA — don't rely solely on incognito sessions, which can be inconsistent.
* When a clear winner emerges, pause losing variants rather than deleting them. Historical data is retained and useful for future reference.

### Troubleshooting

* **Only one version shows repeatedly:** expected for return visits to the same page link; use the dashboard to preview others or test in a fresh session.
* **Uneven traffic:** confirm all variants are **active** and the UTM audience is large enough.
* **No data in Analytics:** ensure the MatchAgent is deployed correctly and the page has traffic that matches the UTM rules.

💬 **Need help?** Submit a support ticket from the dashboard — click the support icon in the upper right corner.
