GSC clicks and GA4 organic sessions never match: which one do you report for SEO?

Every few weeks someone asks why Search Console shows 1,800 organic clicks last month while GA4 shows 1,200 organic sessions. Take a dental clinic site as the example. The gap is normal. GSC counts clicks on Google results, but GA4 only records a session when the tag fires and consent allows it, so blocked scripts, consent rejections and very fast bounces never make it through.

My view is that you should pick one number as the SEO scoreboard and stop trying to reconcile the two click for click. Use GSC for query and page performance, since it's Google's own count of impressions and clicks, and use GA4 only for what happens after the landing: conversions, engagement, revenue. Trying to force them to match is where people lose a week. How do you split them, especially when a big share of visitors reject the consent banner?

I'm not sure that split holds up. If a lot of visitors reject consent, GA4 is missing the conversions you'd use to judge those pages too, so its post-landing numbers have the same blind spot. Check whether consent mode modeling is on before you trust the conversion rates.

Kai's right that the blind spot carries over to conversions, but I don't think that kills Jun's split. It comes down to how big the consent-rejected share of your organic traffic is, and whether it's similar across landing pages.

If 15 to 20 percent of organic visitors reject, GA4 can still rank pages against each other reasonably well. The trouble starts when rejection swings from page to page. A blog post that mostly attracts people who decline tracking will look weaker than a service page that doesn't, even if it's just as valuable. That's the point where I'd stop comparing raw conversion rates across pages.

If the rejection share is high or uneven, patching GA4 with more SEO reporting won't help. Pull leads or orders from your CRM or backend and attribute them to the landing page instead. That way the conversion number doesn't depend on the banner at all.

With Consent Mode v2, visitors who decline still send cookieless pings, and GA4 uses those for modeling. That does recover some of your totals, but the modeling needs enough traffic to work, and it won't rescue a page where most visitors say no. In that case backend attribution is still the cleaner option.

Backend attribution isn't automatically the cleaner option. A CRM lead only knows its landing page if the form passed it along, and plenty of forms drop the landing URL before it reaches the CRM. Phone calls are worse. For a dental clinic, a big share of bookings can come in by phone, and those vanish from SEO reporting unless call tracking is tied to the page. So you'd be swapping a consent gap for an attribution gap that's much harder to spot. Before you trust the CRM over GA4, check what share of organic leads actually arrive with a page field.

Agreed, and the form field is only half of it. Organic has no click ID to carry through, so you have to capture the landing page the moment the visitor arrives. A first-party cookie that stores the first page they hit, copied into a hidden field on the form, usually survives the trip to a "thanks" page where the referrer doesn't.

Phone is harder. A per-page tracking number only counts people who actually saw that page's number. Say a clinic's ad-free service page gets 40 calls a month: you'd want to know how many of those callers were organic, not just how many numbers rang.

I wouldn't go the cookie route. Storing the first landing page is still writing to the visitor's device, so the people who rejected consent are exactly the ones it won't cover, and you'd end up rebuilding the same gap. Capturing the landing URL on the server at first request is less exposed, though it has its own limits.

I don't think the server-side approach closes that gap. The first request gives you the landing page, but the form submit comes a few requests later, and the server can't link the two without an identifier the visitor carries. With no cookie, that usually means a URL parameter or a session token in the form, and both break when someone opens the page in a new tab or moves from phone to laptop. So server-side still leans on client-side state, just less visibly.

For a clinic with a short booking funnel, a required "how did you hear about us" field with a clear organic search option is often more useful. It's self-reported, so consent isn't an issue, and it's crude, but it survives every device switch. It won't tell you which page drove the booking, but it will tell you whether organic is producing bookings at all. And when a booking comes in by phone, is anyone logging the source, or does it end up as "walk-in"?

Agreed on the self-reported field, but build in one fix from the start. A lot of people who say "Google" mean whatever Google showed them, and for a clinic that's often the Maps pin rather than the organic results. If the only choices are "Google" and "Other", Maps and paid traffic inflate your organic count and you can't see it happening. Split the option into "Google search results", "Google Maps or Business Profile", "Google ad", and "friend or family" at minimum.

The Maps line matters most for dental, because a lot of local bookings come from the pin and never touch a blog post. Then compare the self-reported split against GSC clicks and your Business Profile insights for the same month. If they point the same way, the field is doing its job even if the exact numbers are soft.