Why Are Google Ads Conversions Being Double-Counted? A Practical Diagnosis
Google Ads conversion numbers can be inflated by repeat fires, overlapping primary actions or the wrong Count setting. Use this practical diagnosis before touching the tags.
Short answer: start by separating three different problems that are often called "double-counting": repeated fires for one action, several conversion actions recording the same action, and a counting setting that is correct for sales but wrong for leads. Do not change the Count setting or delete tags until you know which problem you have. Otherwise you can make the report look tidier while breaking useful measurement.
What double-counting can look like
If one person submits one form and Google Ads records three conversions, the likely causes are not all the same. A form success event might fire on the form submit and again on a thank-you page. A Google Ads tag and an imported Analytics event might both be marked as primary goals. A purchase tag might fire again when somebody reloads the confirmation page. Or the action may legitimately happen more than once after a single ad click.
Those situations need different fixes. Treating every high number as a tag bug is the measurement equivalent of replacing the car because the fuel gauge looks suspicious.
First check: which conversion actions are actually being counted?
Google Ads lets each conversion action use a Count setting of Every or One. Every counts each conversion for that action after an ad interaction and can make sense for purchases where each sale has value. One counts only one conversion per ad click and can make sense for a lead action where repeated forms do not represent separate prospects.
The setting is per conversion action. A single click can lead to more than one action, and changing the setting only affects future reporting. It does not rewrite historical data. Start in Goals, then Conversions, then Summary. List every action marked as a primary goal for the relevant campaign and write down what real-world action each one represents.
Do not assume the default is right. Google says defaults vary by conversion source. A lead form, a purchase, a phone call and an imported Analytics action may need different decisions.
Second check: can the same event fire more than once?
Use a clean test journey. Open the landing page with Tag Assistant or your chosen debugging tool. Complete one test action. Note every tag or event that fires and the page or event that triggered it. Then reload the confirmation page and check whether the purchase or lead event fires again.
For a form, check whether both the form handler and the thank-you page send a conversion. For a purchase, check whether the order confirmation is persistent and whether a refresh repeats the tag. For a booking flow, check whether the site event and an imported platform event represent the same completed booking. The question is not "did a tag fire?" It is "did one real business action create more than one primary conversion?"
Third check: are two systems measuring one action?
A Google Ads conversion tag, an imported GA4 key event and an offline CRM upload can all be useful. They are not automatically three separate outcomes. Map each action to its source, trigger, primary or secondary status, and business meaning. If two primary actions are both intended to represent one enquiry, choose the reporting design deliberately rather than letting both steer bidding.
This is where account notes matter. Record the action name, the trigger, the destination, whether it is primary, the Count setting and the person or system that owns the source. A screenshot of the conversion summary is useful, but it is not a substitute for testing the actual journey.
For purchases, use a unique transaction ID
Google recommends a unique transaction ID, such as an order confirmation number, for purchase conversion tracking. If a customer reloads the confirmation page, Google can recognise the same ID and avoid counting the repeat as another conversion for that action. The value must be generated dynamically, unique to the transaction and free from information that identifies the customer.
Do not hardcode one example ID into every order. That can cause undercounting because Google may treat later legitimate purchases as repeats. If the same order is sent through a browser tag and a server or CRM upload, keep the transaction ID consistent across both routes and verify it with a test transaction.
A hypothetical AU-dollar check on why the diagnosis matters
This is an explicitly hypothetical example, not a Nexus result or industry benchmark. Assume a campaign spends $1,200 and produces 40 genuine enquiries. The account reports 60 conversions because 20 enquiries were recorded once by the Google Ads tag and again by an imported event.
- Reported cost per conversion = $1,200 / 60 = $20.
- Correct cost per genuine enquiry = $1,200 / 40 = $30.
- The report overstates conversion volume by 20 / 40 = 50%.
The arithmetic was checked in code. It does not prove that the duplicate action is the cause in a real account. It shows why a cheap reported CPA deserves a measurement check before anyone increases budget.
What to fix, depending on what you find
If one action is firing repeatedly, repair the trigger so it fires on the genuine completion event once. If two primary actions represent one lead, redesign the goal structure and decide which action should guide bidding. If the Count setting is wrong for the business model, change it knowing the correction applies to future data. If purchase reloads are the problem, add a dynamic transaction ID and test the tag.
Do not solve duplicate reporting by deleting historical data. Keep an annotation of the date and the change. When reviewing performance, separate the affected period from the corrected period and explain the limitation plainly.
The five-minute duplicate-conversion checklist
- Export or screenshot the conversion summary and list every primary action.
- Map each action to one real business outcome and one technical source.
- Run one controlled test journey and watch the tags and events in sequence.
- Reload or repeat the confirmation step to test for repeat firing.
- For purchases, verify a unique transaction ID and its consistency across tag and server routes.
- Record the correction, the date it went live and the first clean period you will use for comparison.
For the wider setup, read the Google Ads conversion tracking guide and the practical guide to clicks with no conversions. If you want a second pair of eyes on the measurement map, Nexus Ads Consultancy can help structure the diagnosis. Final website changes, CRM follow-up and customer conversations remain the client's responsibility.
Next step: do one test conversion today and document every event that fires. A tidy conversion report is useful only when it corresponds to something that actually happened.
