Campaign settings and lifecycle reference
You'll find campaign controls for audiences, page URLs, goals, reports, and visitor activity, including what happens when you disable or delete a campaign.
Use these settings to manage who sees your page changes, maintain the pages they run on, and measure conversions. Your changes save to the project draft and reach visitors after you publish.
Campaign list
Campaigns shows each campaign's name, goal, conversion rates, lift, and status. Each lift names the dates it covers, such as "Whole test, since Oct 5". The summary across campaigns includes conversion lift and the number of live campaigns. It says "Whole test" when every measured campaign's figures cover its whole test, and shows your dates otherwise. The date selector changes the report figures. Each sparkline always shows the last seven days.
Until you pick dates yourself, the list starts at the earliest test among your campaigns, so each row shows the same lift as that campaign's Overview. See which dates the lift covers.
Draft and disabled campaigns explain why they have no live reporting. If a report is loading or unavailable, that doesn't mean it has zero conversions. New campaign opens the creation page whether the list is empty or already has campaigns.
Website personalization campaigns require Scale or Pro on the current plans, and a trial includes them until it ends. Legacy entitlements can differ. When a project's plan doesn't include campaigns, publishing leaves them off your live site and keeps them in the draft. If the plan drops, such as a trial ending on Grow or a downgrade, RightMessage takes campaigns off your live site automatically; after you upgrade, publish to bring them back. An ended trial blocks publishing. A visible draft is not evidence that a campaign is live. See publishing for the project's plan and validation checks, and project billing for what happens when a plan changes.
Creation defaults
| Setting | Default and limits |
|---|---|
| Campaign name | Required. |
| Which page should it personalize? | A valid absolute URL. Starts with HTTPS and your first configured domain. |
| Segment groups | Optional. Choose up to three groups under Create an audience for each segment in these groups. Adding a group selects every segment in it. Deselect any segments you don't need. Skip this to A/B test the page itself. |
| Audiences to create | One audience per selected segment. You can't create more than 100 audiences. The preview shows the first 12 names. |
| Holdback | On at 10% when the campaign has audiences. Held-back visitors see your page as it is. |
Create campaign saves the draft and opens the campaign editor in the same tab. A campaign with no audiences is a plain page A/B test. There is no unrestricted audience: every audience needs at least one segment. Original page in the editor is what visitors see when no audience matches; you can A/B test it.
The editor's Add audiences dialog lets you add more later. Its Select all and Clear all controls apply to the group shown. On the creation page, choosing a group selects all its segments automatically.
Create a campaign walks you through choosing your first audience.
Rename, disable, duplicate, or delete a campaign
Edit the title to rename your campaign. The status switch enables or disables it in the draft. Debug on site opens an authorized debugger session focused on that campaign.
Edit on page opens the campaign editor in the same tab. You can also rename the campaign by clicking its name in the editor's breadcrumb. See the campaign editor controls.
Under More actions, Duplicate and Delete both ask for confirmation. A duplicate is a separate campaign, so review its settings before using it. Publish a deletion to remove the campaign from your live site. A missing or deleted ID shows Campaign not found.
Campaign tabs
| If you'd like to... | Open |
|---|---|
| Check your setup | Overview, with the campaign's lift, Righty's summary, and links to page changes, goals, and testing. Its count covers configured rules and changes, not every URL behind a wildcard. |
| Compare personalized and original content | Results, with published, enabled campaign performance, goal comparisons, combinations of matched audiences, and engagement. |
| Edit pages and maintain their URLs | Experiences, with page targeting descriptions, change counts, and Edit on page links. |
| Inspect individual visitors | Activity, with the campaign changes they received. You need access to visitor profiles. |
| Measure the campaign and run A/B tests | Testing (formerly Split Tests; old links redirect), with the holdback card, page and audience A/B test cards, and deployed winners. See the Testing tab reference. |
| Choose what to measure | Conversion Goals, with attached goals, the primary goal, and goal setup. |
A/B tests, measurement periods, and Deploy
| If you'd like to… | What happens |
|---|---|
| Start a page or audience test | Start test is a draft change; the test starts when you publish. Starting or restarting needs Pro. |
| Change a running test | Restart with changes starts a new measurement period with a fresh split when you publish. The earlier results stay in the test's history as a run. |
| Change something that a test or the holdback measures | Publishing lists it under Restarts measurement. Once a test has a week of data, you confirm with a checkbox before publishing. |
| See the first verdict | The earliest verdict comes at the week 2 check, about three weeks after the test starts. Many tests take longer. |
| End a test | End test asks which version to keep. It's a draft change that goes live on Publish, and it doesn't need Pro. Tests never end on their own. |
| Keep an Original page winner until your website is updated | Choose Show it now, until my website is updated when ending. RightMessage keeps showing it, then you choose Undeploy once your website has the changes. Undeploy doesn't need Pro. |
| Back out a draft change | Cancel start, Cancel restart, or Undo end puts the test back the way it's published. |
If the project moves to a plan without Pro, such as a downgrade to Scale or a trial that ends on Scale, a running test stops on your live site when the plan changes, or at your next publish at the latest. Visitors get control. The test stays in your draft, and you can still end it and keep a version. Grow doesn't include campaigns, so campaigns and their tests come off your live site. See end a test and deploy the winner and how verdicts work.
Edit page URLs
Edit on page opens a listed page in the campaign editor in the same tab. New URL opens Personalize another page and checks the URL. Choose Open visual editor to continue; this button opens the campaign editor despite its older label. Opening a page doesn't save a change by itself.
Edit URL and Bulk update URLs update page rules across matching changes. The bulk dialog has separate Domain and Path fields. It previews the update and shows how many changes it affects, with up to three examples. A blank new path becomes the homepage.
Delete removes the page from matching changes but keeps their other page entries. If a change has no page entries left, RightMessage removes it. Publish these changes to apply them to your site.
Choose your conversion goals
Conversion Goals lists your project's goals, their definitions, and a Tracking switch. Turn on Tracking to attach a goal to the campaign. New goal opens a drawer and attaches the goal you create. Edit changes the shared goal. Remove detaches it from this campaign without deleting it from the project.
With at least two attached goals, Judge this campaign on offers Any goal or Primary goal. New campaigns start with Any goal. Older campaigns without a saved choice use Primary goal. Select Primary goal, then use Make primary on a tracked goal to choose the deciding goal.
Any goal counts a visitor once if they complete any attached goal. Primary goal uses only that goal to decide Winner or Loser.
Warnings flag goals with no events, goals attached to no campaign, and duplicate event definitions. A duplicate warning doesn't merge goals. See goal setup for the supported event types.
Results and lift
Results shows conversion lift, visitors reached, goal conversions, daily conversion rates, and personalized-versus-holdback comparisons for each goal. With Stripe connected and the holdback on, Revenue per visitor compares payments per personalized and held-back visitor. Who it converts for expands combinations of matched audiences. Engagement expands visitors, pageviews, and pages per visitor. Revenue and engagement describe behavior but don't decide the experiment's winner.
Each Who it converts for row shows both groups' conversion rates with their counts, such as "4 of 28". Lift appears only when the held-back visitors converted at least once. Otherwise the row says why, such as "holdback 0 of 4" or "no holdback visitors". A group with fewer than 10 visitors adds Small sample, and the row stays. Each person counts once, in the row for the audiences they matched together, so an audience can appear on more than one row and the rows don't add up to the campaign total.
How lift is measured explains the comparison. With the holdback on, RightMessage compares with it once both the personalized and held-back groups have at least 10 converters; until then the campaign shows no lift. An unchanged rollout can show a previous test's findings. With the holdback off, similar unexposed visitors on the same pages can provide a descriptive baseline once there are at least 10 conversions to compare. That comparison isn't randomized proof.
The headline lift on the Campaigns page and the dashboard's Personalization lift average only controlled lifts, from a holdback or a previous test, weighted by visitors over your chosen reporting range. A baseline lift stays on its campaign's row but never sets the headline. When the only lifts available are baseline lifts, the headline says "No campaign has a holdback with enough conversions to compare. Turn on A/B testing to measure lift".
The report can be computing, refreshing cached data, unavailable, or too large for the chosen range. Wait for computation or narrow the dates as requested. An error isn't evidence that personalization performed worse.
Which dates the lift covers
Until you pick dates yourself, the campaign list and a campaign's Overview and Results cover the whole test: from the day its first measurement run started through today. That includes runs from before a publish restarted the holdback's measurement. The automatic start stays within the last 90 days.
The words beside the lift name the window.
| Label | What the figures cover |
|---|---|
| "Whole test, since Oct 5" | Every run of the campaign's test, through today |
| "Oct 23 only" | One day |
| "Sep 4 – Oct 3" | Any other dates, including the last 90 days of an older test |
Dates you pick, including dates carried over from a dashboard link, stay as you set them. When they don't cover the whole test, the summary strip adds the whole-test lift beside them, such as "Whole test is +12.4% since Oct 5". For a test older than 90 days, it reads "+12.4% since Jul 26", covering the last 90 days. If an earlier run within your dates held back a different share, the lift is shown without a better or worse color, with "Lift mixes runs with different holdback shares."
Righty's summary on the Overview names the same window in its heading and first sentence, so you can tell which dates its figures describe.
Activity and visitor details
You'll need access to visitor profiles to use Activity. Filter by All visitors, Personalized, Holdback, or Converted, or search for an email or visitor ID. Rows show the group, matching personalization, first and last seen dates, and exposure count. Reporting ranges can't exceed 90 days.
Open a row to see identity, conversion history, acquisition context, device, location, matched audiences, segments at first exposure, and a page timeline. Full profile opens the visitor profile.
For holdback visitors, Would have received shows the personalization they would have seen. Visitors seen in both comparison groups have an exclusion warning because they aren't included in Results.
The drawer shows up to the first 500 pageviews and first 200 exposures. A notice tells you when there are more. If no exposure was recorded, there are no exposure details to show.
For testing controls and verdicts, read the Testing tab reference, holdback setup, and how verdicts work.