Email Grew 16% While Organic Fell 13%, and It's the Only Channel You Own
Kevin Indig's Similarweb analysis of US and French websites, published November 2, 2026, found that email visits grew 16.3% in the US and paid search visits grew 20.8% from January to August, while organic search fell 13.2% over two years. Paid search offset only about 36.6% of the lost US organic visits, and every one of those visits now costs money.
That second number is the one I keep coming back to. Most "organic is shrinking, what now" plans I've read quietly assume some other channel absorbs the loss. Indig's data says no single channel does, and the one that comes closest is the one Google charges you for.
What Indig actually measured (and what he didn't)
This is part two of a series. In part one, published October 26, Indig used Similarweb's modeled traffic for 10,000 domains in each country and found US organic visits dropped from 19.9 billion to 17.3 billion between August 2024 and August 2026. France fell 24.4%. Publishers took the worst of it, down roughly 42% in both markets, while the median small site stayed close to flat.
He also made a point that gets lost in most coverage: AI isn't the only cause. Google started grouping up to four text ads under a single "Sponsored results" label in October 2025, ad blocks got bigger, and Google itself reported paid clicks up 13% in Q2 2026. Indig calls blaming AI for every lost click "a go-to avoidance mechanism," which is a little harsh, but honestly not wrong.
Part two asks where the traffic went instead. Comparing January through August 2026 against the same months in 2025:
- Email: +16.3% in the US (3.94 billion visits), +17.0% in France
- Paid search: +20.8% in the US (3.70 billion visits), +20.3% in France
- Direct: -2.7% in the US, -3.1% in France, even though it's still the biggest channel by median share (43.5% of a typical US site's visits)
Two caveats worth holding onto. These are modeled visits, not customers or revenue, and Indig says so plainly. And the paid search number is sensitive to one domain: including googleadservices.com drops US paid growth to 13.5%, and in the shared sample of 1,896 sites it falls to just +2.1%. Nobody seems to know whether that's real ad activity, routing, or a measurement artifact.
Paid search "growing" mostly means you're renting back your own clicks
On paper, paid search looks like the winner. And in a narrow sense it is: more visits arrived via ads this year than last.
But think about what that means for a single brand. Google moved ads higher and made them bigger, organic results slid down the page, and some of the people who used to click your organic listing now click your ad (or a competitor's). The visit still happens. You just pay for it now.
It's a bit like a landlord who knocks down the free parking lot next to your store, builds a paid garage on the same spot, and then reports that garage revenue is up 20%. Technically true. Not exactly a new source of customers.
The cost side matters here. WordStream's 2026 benchmarks, covered by PPC Land, put the median Google Ads CPC at $5.42 across 13,474 US search campaigns, with cost per lead at $66.69. So run the math on your own losses. If your site lost 10,000 organic visits a month, buying those back at $5.42 costs roughly $54,200 a month. Most teams I've seen don't have that line in the budget, and the ones that do usually can't justify it for top-of-funnel informational traffic that was never going to convert anyway.
Which is the real point. A lot of the organic traffic that disappeared was the cheap, low-intent stuff. Replacing it with paid clicks at a $5 floor is a bad trade. Replacing the high-intent slice is a reasonable one.
Email is the boring answer, and probably the right one
Indig calls email "severely underrated" and "the least volatile acquisition and retention channel there is." I agree with the second part more than the first. It isn't underrated so much as under-resourced: everyone knows email works, and almost nobody puts a growth person on list building.
The US retail and marketplace numbers make the case better than any argument. Email visits grew 26.6% year over year in that segment, paid search 24.9%, organic just 7.0%. That's the one industry where a direct sales motion exists, and email led.
Now, some of that email traffic is order confirmations and password resets, and Indig flags this himself. Transactional email isn't audience building. Still, a reader in the comments, Stephen M. Garrison, put the structural advantage well: someone chose to subscribe, and an algorithm can't revoke that.
Every organic visitor you don't convert into a subscriber is a visitor you'll pay Google to see again.
That line is why I'd reframe the whole channel question. The goal isn't to find a replacement for search traffic. It's to make each remaining search visit count twice.
Direct traffic is two different stories depending on your size
Direct is shrinking overall, but the split by site size is wild. Median small sites grew Direct 14.4% in the US and 21.2% in France. Median large sites lost 13.0% and 7.1%. News and editorial publishers were hit hardest, with editorial sites down 34.4% in the US on a provisional classification.
Part of this is measurement noise, and you've probably seen it in your own GA4. When OpenAI's update stripped referrers from ChatGPT links in October, Adobe's Cedric Huesler measured about 31,000 fewer daily ChatGPT visits and 26,000 more daily Direct visits across 500 pages, according to Search Engine Journal. We wrote about that UTM change and what it did to AI traffic reporting at the time.
Another commenter on Indig's post, Mike Harty, shared FunnelFuel data showing educational content as a first touch dropped from about 30% in 2023 to under 7% in 2026, with pricing and demo pages taking over. His read is that Direct is increasingly "a bag of different behaviours," people who did their research in a chatbot or a Slack thread and then typed your URL. I think that's right, and it's a little uncomfortable, because it means some of your content's value now shows up in a channel that gives it zero credit.
The reallocation test I'd run before moving a dollar
Indig's original framing (and the angle I'd push hardest) is that you should only move budget into channels that can actually absorb it. Here's a version you can run in about an hour:
- Size the hole. Pull organic sessions for January to August this year versus last year in GA4. Split by landing page type: informational content, commercial pages, homepage. Most of the loss will probably sit in informational content.
- Price the paid replacement. Multiply only the lost commercial sessions by your actual CPC (or $5.42 if you don't run search ads). If that number is bigger than the gross profit those sessions used to generate, paid isn't your replacement.
- Measure your email capture rate. Subscribers gained divided by organic sessions, same period. If it's under 1%, that's the cheapest fix on the list, and it compounds. Add a subscribe prompt to the 20 posts that still get the most organic traffic before you write anything new.
- Split homepage Direct entrances by next page. This one comes straight from Indig. If returning customers keep clicking through to login or pricing, put those one click away instead of making them wade through a first-time visitor flow.
A rough benchmark: if your email capture rate on organic traffic doubles, you've likely offset more long-term value than a 20% paid search increase would buy, at a fraction of the cost. I'd treat that as a hypothesis to test rather than a law, but the math tends to land there for content-heavy sites.
A prediction, and one thing I'm unsure about
My guess: by August 2027, Indig's next round of this data will show email growing faster than paid search in the US for the first time, mostly because rising ad load pushes CPCs up and teams finally start budgeting for owned audiences. I'd put that at about 60/40.
What I can't tell from this data, and what Indig admits he can't either, is whether any of these channel shifts created more customers or just moved the same customers around. We've seen HubSpot shrug off an 80% blog traffic drop without much revenue damage, which suggests a lot of the lost organic traffic was never doing much in the first place. It's possible the real lesson here is less about which channel replaces search and more about how little some of that search traffic was worth.
Either way, I wouldn't spend Q1 trying to win back every lost click. I'd spend it making sure the visitors who still show up have a reason to come back without Google in the middle.