homerunn Let's talk
← Back to blog

Why Your Organic Traffic Flatlined, and Why Writing More Won't Fix It Alone

A specialty ecommerce retailer I run marketing for grew organic traffic steadily for about five years. Over the last 12 months, it stopped: 30,500 to 35,600 visits a month, bouncing inside that band, no upward trend.

This doesn’t read like a broken site. It reads like the ground shifted underneath it.

What changed is where the answers live.


The search engine became the destination

The old deal with Google was simple. You answered a question well, Google sent you the click.

That deal is ending. ChatGPT answers the question itself. Google’s AI Overview answers it above the results. Reddit threads fill the first page, and YouTube answers it in a video that never links out.

People aren’t searching less. They’re asking more questions than ever. The answers just stopped requiring a visit to your site.

So traffic isn’t just harder to earn. It’s different too: the visitors who make it through are the ones AI couldn’t fully answer for, which changes what your site needs to do once they land.


The numbers tell you which problem you have

Here’s what this looks like in the data for that retailer.

39.6M impressions in Google Search Console. 0.8% average click-through rate. Average position around 16, the middle of page 2.

The site isn’t short on visibility. It’s short on clicks.

And the AI layer is already active. An AI Visibility score of 26. Roughly 1,300 mentions across ChatGPT, Google AI Overview, AI Mode, and Gemini. About 1,200 pages cited by AI engines.

But look at Google’s own SERP feature breakdown: AI Overviews are only 2.6% of the site’s positions, against 74.1% classic organic. The machines are reading the site and citing it, in a small, real way. The site just wasn’t built for that job: page 2 rankings and thin category pages that exist for browsing, not answering.

Some of that gap is just math. Page two of Google gets a fraction of the clicks page one does, no AI required. But this site is also already being read and cited by AI engines while sitting on page two, which means the content is trusted enough to be a source and still isn’t winning the human click. That’s two problems stacked on each other: a ranking problem old SEO already knew how to name, and a newer one about whether the page converts once someone actually lands on it.


Ranking was the old goal. Getting cited is the new one.

Ranking #1 rewarded backlinks and keyword coverage. You could win with a 2,000-word post that buried the answer under a long intro, because the click happened before anyone read a word.

Getting cited inside an AI answer rewards the opposite. A clear, extractable answer near the top. Terms that are actually defined. Structured content a model can parse. Specific data it can quote.

An AI engine doesn’t need the page ranking #1. It needs a page that’s already trusted enough to be in the running, then answers the question more directly than the alternatives. Authority gets you into the room. Clarity decides whether you get quoted.


A citation without a click still counts. Just not for everything.

When ChatGPT names your brand while someone is deciding what to buy, that’s exposure at the exact moment of decision. That has value. Roughly 1,300 mentions is 1,300 times this retailer showed up inside someone’s research without paying for it.

But a citation doesn’t collect an email address. It doesn’t build a remarketing audience, and it doesn’t create a customer you can sell to again.

Treat AI citations as a new kind of brand awareness. Don’t treat them as a replacement for traffic you own.


The real hedge is a channel no algorithm controls

Google organic, AI citations, Reddit visibility, YouTube reach. Every one of those is rented, and the terms can change in a quarterly update you don’t get a vote on.

The hedge is owned channels, and email is the strongest one. A list you built is yours. No model update or SERP redesign can take it away. Every result I can show you from email shares that trait.

Video belongs in this plan too, as the bridge. If you’re not creating video, start small: short-form clips repurposed from blog posts you already have. YouTube builds a presence that still sends real traffic to your store, and it doubles as a source AI engines cite. TikTok and Instagram add zero-click visibility on top, which is fine as long as you treat them as awareness and not as the system.


What to do with a flat traffic graph

  1. Pull Search Console. Compare impressions, click-through rate, and average position. Huge impressions with sub-1% clicks points to answer clarity, not visibility.
  2. Rewrite your highest-impression pages to answer the question in the first hundred words, with defined terms and specific numbers an AI can lift.
  3. Put a conversion path on those same pages: an email capture and a clear next step. If a human does land, the page should do something with them.
  4. Start repurposing your best posts into short video. YouTube first, short-form second.
  5. Build the email list like the traffic might stop, because some of it will.

That’s not a list of tactics. It’s a repositioning of the whole content system, from chasing clicks to earning citations while building channels nobody can switch off from outside.


If your organic traffic looks like this retailer’s, flat band, big impressions, weak clicks, I’m happy to look at your Search Console data and tell you what’s actually going on. Reach out. No commitment, just a conversation.

Ready to talk?

If this resonated, let's see if it makes sense to work together.

Get in touch