# Why Google Ads and GA4 conversions don't match > If you raise budget on the platform number, you are scaling on sales your bank account never sees. Use each number for its own decision. Source: https://www.nashmarketinglabs.com/en/blog/google-ads-vs-ga4-conversions Author: Nash Marketing Labs, a paid media agency (Google Ads, Meta Ads, TikTok Ads, LinkedIn Ads). Date: 2026-10-12 --- **Say Google Ads reports 400 conversions for September. GA4 reports 280. Your backend shows 210 orders.** If you set next month's budget on the 400, you are paying to scale 190 sales that never reached the bank. All three systems can be working as designed. Each one counts something different, under its own rules. The decision that costs money is which number you use for which call. Below: what each one counts, two divisions you can run today on last month, and the signs that the gap has become a real fault. ## What exactly each of the three counts **The ad platform (Google Ads, Meta) counts attributed conversions.** These are the sales it claims for itself inside its attribution window, including view-through conversions with no click. Someone sees the ad on Monday, searches your brand on Thursday and buys: the platform can claim that sale. Google and Meta each claim their own share, so the same order can appear in both. **GA4 counts events inside sessions.** It needs the tag to load, consent to be accepted and the browser to let the signal through. Its attribution model splits each sale across channels, so it may credit Organic with a sale the platform claims in full. Some loss against the backend is expected for those reasons alone. **Your backend counts orders.** Real orders, with their value, net of cancellations, refunds and failed payments. This is the number that reaches the bank. It tells you little or nothing about which channel brought the order. So: the backend measures the business. The platform and GA4 help you steer the channel. ## The two ratios that turn the mess into information Take one closed month and run two divisions: - **Signal ratio = backend orders / GA4 conversions.** In the example, 210 / 280 = 0.75. It shows how much measurement you lose to consent, blockers, cross-device purchases or incomplete tagging. If it jumps from one month to the next, look at the tagging first. - **Attribution ratio = GA4 conversions / platform conversions.** In the example, 280 / 400 = 0.70. It shows how much the platform claims above what your site records. If it falls, check for long windows, view-through counting or several platforms claiming the same orders. Multiply the two and you get backend / platform: 0.75 × 0.70 = 0.525. A platform ROAS of 4 becomes 4 × 0.525 = 2.1 against real orders. That corrected figure is the one to set against your margin. It will never be exact, but if it holds steady month after month, it is good enough to decide on. ## When the gap stops being normal Expect a gap every month. What you watch is whether it holds **steady**. If the platform reports 40% more than the backend every month, that 40% is your constant and you can work with it. A sudden move is the alarm. Signs that something is really broken: - **The ratio changes overnight** and no one touched the campaigns. Check the last site release, the consent banner and any checkout change. - **The platform reports more than the backend bills in total**, across all channels. Look for double counting or an event that fires more than once. - **Average order value does not match** between platform and backend. If the platform value is higher, check whether you send the amount with tax and shipping, or before refunds. The algorithm is then learning from an inflated number. - **Conversions match but revenue does not.** Check whether the purchase event sends a fixed value, or the cart value instead of the confirmed order. ## Which number for which decision - **The backend decides whether the business is growing.** Any report for leadership starts there, set against total spend. That is what [MER](https://www.nashmarketinglabs.com/en/blog/mer-vs-roas) measures. - **The platform decides inside the channel.** Which campaign goes up, which goes down, which creative stops. Its number is consistent with itself, so it works for comparing campaigns with each other. - **GA4 explains the path.** Which channels show up earlier, which pages lose people, where the session drops. Use it to find where to look, then confirm against the backend. The costly move is taking platform ROAS, the highest of the three, to whoever reads the bank statement as if it were business ROAS. Months later the two numbers have drifted apart, and the meeting ends up arguing about the agency instead of the campaigns. Two settings widen the gap quietly. The attribution model changes the report and also steers bidding. The attribution window decides how many days back the platform can claim a sale. In lead generation, also check how conversions are counted: a form set to "every" instead of "one" adds conversions your CRM will never see. ## The twenty-minute check - **Pick last month** and pull the three numbers with the same date range and the same time zone. A time zone offset or a month that starts on a different day is the first thing to rule out. - **Line up the dates.** Platforms assign the conversion to the day of the click, not the day of the purchase. If you compare against orders by order date, month ends will not match. - **Calculate both ratios** and write them down. One month on its own says little; the series is what calibrates. - **Repeat for three months.** If the ratios hold, your measurement is imperfect but usable. If they swing, that is where the technical fault is. - **Only then recalculate the target ROAS** of your campaigns with the correction factor, and compare it with the [minimum profitable ROAS](https://www.nashmarketinglabs.com/en/blog/minimum-profitable-roas) for your margin. That threshold decides whether you raise budget. ## Frequently asked questions ### Why does Google Ads report more conversions than GA4? Because they count different things. Google Ads attributes conversions to its own ads inside its window, including view-through conversions with no click, using its own attribution model. GA4 counts events in sessions and splits the credit across channels, so a sale Google Ads claims in full may go to Organic or Direct in GA4. Add the signal lost to consent and blockers, and the gap grows with nothing broken. ### How much difference between platform and backend is normal? There is no universal figure: it depends on the industry, the purchase cycle, the share of mobile and how consent is set up. What you can check is stability. A gap that holds month after month is a constant you can work with. A gap that moves suddenly with no campaign changes points to a measurement fault: a dropped tag, a new consent setup, a migrated checkout or a duplicated event. ### Which number should I use to calculate ROAS? Two, for two different decisions. To know whether the business makes money: backend revenue against total spend, which is MER. To decide which campaign goes up or down inside the channel: the platform number, which is consistent with itself. Platform ROAS is the highest of the three, so do not present it as business ROAS. ### How do I know if the problem is attribution or measurement? Compare the two ratios. If backend and GA4 are close but the platform reports much more, it is attribution: long windows, view-through or several platforms claiming the same order. If GA4 sits far below the backend, it is measurement: the signal is lost before it arrives. Attribution is fixed by reading the reports differently; measurement has to be fixed on the site. ### Can I make the three numbers match? No. They measure different things by design. The realistic goal is a gap that is known, stable and explainable: how far each system deviates, why, and the factor you apply when you decide. With that factor you can raise budget on numbers that hold up in the bank. ## More resources - [MER vs ROAS: which metric to use and when](https://www.nashmarketinglabs.com/en/blog/mer-vs-roas) - [What is the minimum profitable ROAS (and how to calculate yours)](https://www.nashmarketinglabs.com/en/blog/minimum-profitable-roas) - [ROAS vs ROI: what's the difference and when to use each](https://www.nashmarketinglabs.com/en/blog/roas-vs-roi) - [CAC vs CPA: the difference and which one to track](https://www.nashmarketinglabs.com/en/blog/cac-vs-cpa) --- Contact: https://www.nashmarketinglabs.com/en/request-audit - Your performance plan: concrete actions to improve your account, no commitment.