Review conversion measurement, search intent, landing pages and bidding in that order. Learn why two search terms with the same cost can require different decisions.
Start Google Ads optimization with measurement rather than budget. In a Search campaign, confirm the conversion you want, inspect the searches that triggered ads, then review the landing page. Only after those checks should you decide whether to change bidding or spend. Optimizing toward the wrong conversion can make an incorrect signal more expensive.
1. Confirm what counts as a conversion
A store may need purchases, while a service business may need qualified enquiries. Check that the conversion actions used for optimization represent the outcome you actually want. Counting the same order through two primary measurement sources can mislead both reporting and bidding. Review the source, value and role of each conversion action. Google’s guide to creating conversions from Analytics events explains how imported events fit into that setup.
2. Separate expensive searches from irrelevant searches
An expensive search term is not automatically a poor one. “Office chair repair” is irrelevant to a store that only sells new chairs. “Adjustable office chair” may be relevant, even if it has few clicks or recent clicks have not yet produced reported purchases. Check the actual search rather than relying on the name of the keyword it matched. Google’s search terms report documentation describes the report; do not assume every search will appear in it.
Negative keywords also need the right scope. An exclusion for one campaign can be inappropriate for another campaign that offers a different service. Inspect the match type and where the exclusion applies before saving it. The useful outcome is less irrelevant demand without removing good searches, not simply adding more exclusions.
3. If the search is relevant, inspect the landing page
Open the destination on mobile. Does it show the product promised in the ad? Is the item in stock, and are price and delivery terms clear? A recorded click does not prove the checkout works. Fix broken measurement before judging reported results; fix a broken product page before changing bids to send more visitors to it.
4. Make bidding and budget decisions separately
Target ROAS uses conversion values in bidding. Setting a target is not a guarantee of achieving that return. Verify the values and allow for the normal time between a click and a reported conversion. Google’s bidding learning guide explains conversion cycles, and its target ROAS guide describes value-based bidding. Changing the target, budget and landing page together makes it harder to interpret the result.
Worked example: Two terms, two decisions
Suppose two hypothetical queries each spend TRY 600 without a reported purchase. One is a repair request for a service the store does not offer. The other names a product the store sells, received clicks in the last two days and leads to a page showing that product out of stock. For the first, investigate an appropriately scoped negative keyword. For the second, check stock or destination accuracy and account for conversion delay. Identical spend and zero reported sales do not establish an identical cause.
Keep a short decision record
Write down the problem, evidence, one planned change and the next review date. Review purchase volume and conversion value alongside cost per acquisition. A lower CPA with very little useful volume is not a complete success measure. The AdLens analytics overview explains how campaign reporting supports that review. Changes to queries, bidding and budgets remain separate decisions in your Google Ads account. The Turkish version of this checklist covers the same review sequence.
