Did your Google Ads budget or bid change work?

How do I know if a Google Ads budget or bid change improved results?

Note the date and the old and new value, then compare equal periods before and after: clicks, conversions, cost per conversion and conversion value, for the changed campaign and the account as a whole. When your AI changes a budget or bid through Cogny, Cogny opens a ticket on its own and reads the result in Google Ads three days later, closing it as worked, partial, didn't work or inconclusive.

How long should I wait after a budget change?

Give the campaign a few days before you judge it. Smart bidding often needs time to adjust after a budget or target change, and conversions can be reported days after the click. Compare at least three full days, and keep the before and after periods the same length and on the same weekdays. If the campaign gets few conversions, wait longer, or accept that the result is inconclusive.

Which metrics should I compare?

Start with the metric the change was meant to move. For a budget increase: did conversions grow, and at what cost per conversion? For a bid or target change: cost per conversion or ROAS, alongside volume. Always check impression share lost to budget, and the account total, because a gain in one campaign can be spend moved from another. Clicks and CTR alone don't tell you whether it paid.

Can my AI track this for me?

Yes. Connect Google Ads to Cogny and work from Claude, ChatGPT, Cursor, Codex or the Cogny app. Budget, bid, status, keyword and targeting changes made through Cogny become tickets automatically, as do new campaigns and ads. Three days later Cogny compares the periods in your Google Ads data and records the verdict, misses included. This runs on every plan, Free included.

How to measure a Google Ads change, step by step

1. Record the change. Campaign, setting, old value, new value, date and time. Google Ads keeps this in Change history too.
2. Name the goal. More conversions at a similar cost, a lower cost per conversion, or a higher ROAS.
3. Pick equal periods. For example, the 7 days before the change and the 7 days after, so each weekday appears once in both.
4. Compare the campaign. Clicks, conversions, cost per conversion, conversion value and impression share lost to budget.
5. Compare the account. Did total conversions rise, or did spend just move between campaigns?
6. Allow for lag. Re-check conversions a few days later, once late conversions have come in.
7. Write the verdict. Worked, partial, didn't work or inconclusive, with the numbers. Keep the misses.

Frequently asked questions

Is three days enough?

For a campaign with steady daily conversions, three days shows the direction. For a low-volume campaign it often isn't, and Cogny will close the ticket as inconclusive rather than call a win.

What if I made several changes on the same day?

Then the result belongs to the set, not to one change. Change one thing per campaign at a time when you want a clean reading.

Does Cogny make the change without asking?

From your own AI over MCP, your AI client asks before a write; Cogny logs every write with who authorised it. In the Cogny app, write tools ask first by default. Any tool can be blocked.

Do I need a paid plan?

No. Measurement runs on every plan. Free includes 15 MCP calls a month; Solo is $9 a month with MCP unmetered under fair use.

Related: How to measure whether an SEO change worked · Connect Claude Code to Google Ads · Google Ads automation with Cogny