> ## Documentation Index
> Fetch the complete documentation index at: https://help.autoady.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Understand results and attribution

> Compare numbers only after their definitions, reporting windows, and sources agree.

Two reports can be accurate and still disagree because they count different events or use different reporting rules. Check the measurement definition before treating a difference as an error.

## Use Reporting details

In **Meta Ads**, open **Reporting details** to inspect the reporting dates, account timezone, currency, fetched time, attribution configuration, and limitations. **Today in progress** means the current day is incomplete.

| Term | How to interpret it |
| - | - |
| **Results** | The Meta configured event represented for the entity. Check its event or objective context. |
| **Cost per result** | Spend divided by the relevant result count. It is only comparable when the result definitions agree. |
| **Conversions** in Google Ads | Counted conversions from the connected Google Ads reporting source. They are not automatically the same events as Meta results. |
| **Attribution** | The provider's rules for assigning a counted outcome to advertising. Mixed settings need extra care. |
| **Fetched time** | When the source report was obtained. It is different from the reporting dates or the time you opened the page. |
| **Budget** | Current configured allowance on its owner. It is different from spend incurred during the selected period. |

## Compare with a provider report

Check these in order: account or customer ID; start and end dates; timezone; currency; attribution settings; result or conversion definition; and data completeness. Then compare the same entity level.

Avoid adding reach across rows as if every row represented different people. Likewise, rows with different objectives should not be ranked by an unlabeled result count or cost per result.

## Leads and revenue need their own proof

A configured Meta event with a qualified-sounding name does not verify a CRM outcome or the configuration that existed historically. Use a connected source with the corresponding outcome evidence when evaluating qualified leads or won business.

A website click is not a successful form submission. A browser purchase-like event is not confirmed paid revenue. [Website events](/websites/events) explains the difference.

## Read impact claims carefully

A before/after result is an observational comparison. Other changes can affect it. Use the comparison as evidence for review, with its period and limitations, rather than a guaranteed causal return.

If the definitions agree but a persistent discrepancy remains, [contact support](/help/contact) with both periods, entity IDs, and reporting settings.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.