Client reports: platform ROAS or one blended number when the platforms disagree?

Curious how others handle client reporting when the numbers don't line up. Say an e-commerce client sees 4.2 ROAS in Google Ads, 3.1 in Meta and 2.6 blended in GA4, and wants one number for the monthly deck. The options I see are reporting each platform's own figures with a caveat, building a blended number from GA4 revenue divided by total spend, or just picking whichever platform looks best. I'd rule out the third.

My view is that the blended number should be the headline, because each platform counts overlapping conversions and view-through credit that the other platform also claims. I'd still show the platform numbers underneath so the client doesn't feel anything is being hidden. Does that hold up for clients who only care about revenue per dollar, or do they end up asking why the numbers don't match anyway?

Agreed on the blended headline. Platform ROAS will always look better because each platform claims credit for the same sale. But the blended number is only as trustworthy as what you divide by. If you take GA4 revenue over media spend, and the client's finance team sees spend that includes agency fees or a Shopify app charge, your figure won't match their books and you'll get the "why doesn't this add up" question. So put the definition right on the slide: net revenue after refunds, media spend only, same date range. That one footnote heads off most of the follow-ups.

Agreed on the footnote. One thing on the GA4 side: check which revenue field you're pulling. GA4's default ecommerce revenue includes tax and shipping in many setups, while the client's books usually don't, so you can see a 5 to 10 percent gap before anyone even looks at spend. Pick the definition that matches their P&L and say which one it is on the slide. Otherwise you'll still get the "why doesn't this add up" question, even with a clean footnote.

Refunds matter as much as tax and shipping here. GA4 won't subtract them unless the site sends refund events, and a lot of setups don't. Revenue stays gross, so a store with a 10% return rate shows a ROAS that's too high even after you've cleaned up tax and shipping. A quick test: take a month with heavy returns, compare GA4 revenue to the store's refund export, and see whether the gap matches the return rate.

A one-month version of that test can mislead you, because returns usually land the month after the sale. A January refund export set against January GA4 revenue is mostly December orders coming back, so the gap won't match your return rate. Pull refunds by original purchase date, or run the comparison over a full quarter so the timing evens out. Also check whether refunds are recorded at the discounted price or the list price, since that shifts the gap too.