Ask Advisor in Google Analytics: A Guide for Google Ads Teams

Ask a question about the property you are in

In an eligible Analytics property, open Ask Advisor from the top-right icon or through the Analytics search box. Google's Help page currently describes English-language availability and responses based on that property's data. It can answer questions, visualize results, link to reports and inspect property configuration.

An example request: “Compare purchases from paid search for the last two complete weeks. Which campaigns and landing pages account for the difference? Show the reports behind the answer.” Adjust the question to the data your property actually collects. A lead-generation business might care about a submitted form or a qualified-lead event instead of purchases.

Follow the report links. Check the selected property, dates, filters and named metric before discussing the explanation. Ask Advisor reflects the data available in Analytics; it cannot resolve a missing business fact simply by producing a confident answer.

A GA4 drop is not automatically an Ads failure

Imagine a hypothetical account where paid-search purchases look lower in GA4, but the Google Ads conversion total is nearly flat. The immediate task is to understand the difference, not lower bids until the charts agree.

First confirm that you are comparing the same event and campaign population. A property's total purchases include more than the Ads-attributed purchases you may be looking at elsewhere. A session-based acquisition report also answers a different question from an attribution report. Keep the report names and dimensions visible.

Then compare counting settings, date basis and reporting delay. Standard Ads conversion reports credit the ad-interaction date, while Analytics records the event date. Try the Conversions (by conv. time) or All conv. (by conv. time) columns when comparing with Analytics. Counting one conversion per interaction can also differ from counting every event. Google's discrepancy guide advises allowing 24–48 hours for synchronization before comparing; a new link or significant setup change may take longer.

None of these differences proves the setup is healthy. They tell you what to reconcile before calling the discrepancy a broken tag or a campaign decline. Once the measures are comparable, an unexplained gap becomes a narrower and more useful investigation.

Check the connection and the measurement

Verify the intended Ads account is linked under Analytics Admin, Product links, Google Ads links. The person setting up the link needs Administrator or Editor access to the Analytics property and administrative access in Google Ads. Linking enables data sharing; it does not guarantee that the right event is used for bidding or counted as intended.

Open the conversion action the campaign actually uses. Compare its source and settings with the event in the Analytics report, and check the campaign's selected conversion goals. Do not change a primary or secondary designation merely to make two totals equal.

If tracking is the concern, bring evidence from the tag, event or import that is in question. An account-setting read can show a configuration without proving that a form fired correctly or that a CRM upload arrived. Preserve the relevant error or diagnostic information instead of replacing it with an AI-generated explanation.

If measurement is consistent and purchases really fell, return to the business question. Did traffic change? Did the site or checkout change? Was a promotion withdrawn or a product unavailable? Some answers are in the ad account, others are with the people running the site.

Send the evidence the next person needs

A useful message to a paid media teammate might say: “GA4 purchase events from these paid-search campaigns fell over this period. We checked the account link and report filters. We have not yet reconciled interaction-date versus event-date reporting. Here are the two reports; please check that comparison before we change the campaign.”

That hypothetical handoff is useful because it separates an observation, completed checks and an unanswered question. It does not need a special 'join template' or a mandatory set of fields for every investigation.

Once the reason is established, shorten the final explanation to what matters to the reader. If there is still uncertainty, say what evidence would settle it. Do not turn an attribution difference into a story about wasted spend.