> For the complete documentation index, see [llms.txt](https://docs.adalysis.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.adalysis.com/manage/campaigns/campaign-experiments.md).

# Campaign experiments

Track your Google Ads experiments and see what to do next

Go to **Manage** > **Campaigns** > **Experiments** to see every campaign experiment in your Google Ads account in one place. Adalysis shows how each trial is doing on its goal, whether the difference is significant, and what to do next.

All results come from Google Ads. Adalysis doesn't run its own significance calculations on experiments. It shows Google's figures in a format that's easier to act on.

{% hint style="info" %}
Campaign experiments are a Google Ads feature, so this screen isn't available for Microsoft Advertising accounts.
{% endhint %}

#### How the data is refreshed

Adalysis downloads your experiments from Google Ads the first time you open the screen each day. The time of the last refresh is shown at the top of the screen.

#### The experiments list

Use the dropdown to choose which experiments you see:

* **All experiments**
* **Running**
* **Needing a decision**: experiments where you need to act, for example a trial that has ended and met its goal but hasn't been applied yet
* **Ended, applied or converted**

For each experiment you'll see:

* **Experiment**: name and experiment type (for example Search custom, Performance Max vs Shopping, AI Max, Broad match, Ad variation)
* **Status**: In setup, Scheduled, Not started, Running, Ended, Applying, Applied or Converted to campaign
* **Dates**: start and end dates, with days left for running experiments or days since the experiment ended
* **Campaigns and split**: the base and trial campaigns, and how traffic is split between them
* **Goal and Google's result**: the change in the goal metric (for example *Cost / conv. -12.4% (±5.1%)*) and whether Google reports it as **Significant** or **Not significant**
* **Next step**: what to do next, in plain language

**Next step**

The **Next step** column turns Google's result into a recommendation. For example:

* *Trial is meeting the goal… Consider applying it.*
* *Ended and not applied. Trial met the goal… Apply it, or convert it to a campaign.*
* *Ended. Trial missed the goal… No need to apply it.*
* *Ends in 3 days with no significant difference in Conv. rate yet.*
* *The trial campaign isn't serving (Not eligible: budget constrained), so no results can build up.*
* *Applying it to the base campaign failed. Check the experiment in Google Ads.*

If an experiment has no goal set in Google Ads, Adalysis reports a significant change as it is, without calling it better or worse. Some tests aim for fewer clicks, for example.

Experiments that ended more than 90 days ago stay on the screen as history, with their final result.

#### Experiment details

Click an experiment's name to open its details.

**Setup and goals**

See the experiment type, description, dates, base and trial campaigns, the traffic split, and the goals set in Google Ads.

**Google's results**

A table compares the control and trial for each metric:

* **Control** and **Trial** values
* **Change**: the trial against the control, as Google estimates it. This is a percentage for most metrics and a number of conversions for Conversions.
* **Margin of error**
* **p-value**
* Whether the result is **Significant**

CTR, Avg. CPC and Conv. rate are marked with \*. Adalysis calculates them from each arm's own totals, because Google doesn't give a change or significance figure for these metrics.

**Day by day**

A line chart shows control vs. trial for each day, as Google reports them. Choose the metric (Impr., Clicks, Cost, Conv. or Conv. value), and switch on **Running total** to see the cumulative figures.

Each arm's totals follow its own share of traffic. With a 50/50 split the two lines are directly comparable. With other splits, keep the split in mind.

**What's different**

See exactly what differs between the base and trial campaigns, read live from Google Ads:

* Campaign **Settings**: bid strategy and targets, budget, networks, location targeting, AI Max, text customization, final URL expansion, tracking template and final URL suffix
* **Ad groups**, **Keywords**, **Ads** and **Negative keywords**
* **Locations**, **Ad schedule**, **Devices**, **Languages** and **Audience lists**

Each section lists what's **Only in base**, **Only in trial**, and **In both, but different**. If everything matches, you'll see "Same in both".

{% hint style="info" %}
**What's different** isn't available for ad variations, experiments without a control arm, AI Max experiments where control and trial share one campaign, or trials that are still drafts.
{% endhint %}

#### Applying or ending an experiment

The Experiments screen is read-only. To apply, convert or end an experiment, go to Google Ads. Adalysis picks up the change the next time the screen refreshes.