The Untold Truth About Ranking in Google’s SGE (And What We Discovered)

Table of Contents

Our case study reveals surprising insights about ranking in Google’s SGE.

Can You Actually Rank in Google’s SGE? Here’s What Happened in Our Case

When Google quietly started rolling out its Search Generative Experience SGE for short it didn’t look like a big deal at first glance. You searched for something, and instead of just the familiar list of blue links, there was a big AI-generated answer sitting at the top, pulling in information from multiple sites. Some people saw it as a novelty. Others saw it as the future of search.

For us, it was an opportunity. A new front page had just opened up, right above the old front page. And the way to get there wasn’t exactly the same as ranking in traditional SEO.

This is the story of how we tested, adjusted, and eventually managed to get one of our clients consistently showing up in SGE’s AI summaries for their most valuable topics. It wasn’t instant. It wasn’t magic. It was a mix of understanding how Google’s generative answers work and shaping our content to fit their needs without losing the human touch.

The starting point for Rank in Google’s SGE

The client was in the home improvement space a niche where competition is fierce and the top search spots are often dominated by huge brands and content mills. They had been doing well with traditional SEO for years. Their guides on topics like “how to install a backsplash” or “best flooring for kitchens” were pulling in steady traffic.

But in early tests of SGE, we noticed something alarming. When we searched for their core topics in the SGE interface, their site wasn’t mentioned at all. The AI answer was pulling from big names, DIY forums, and even a few less polished blogs, but our client’s well-researched articles were invisible.

This was a wake-up call. It hit us that showing up in SGE wasn’t only about protecting the rankings we already had. The AI’s answer layer was becoming the new first impression. If you weren’t there, you risked losing awareness even if you were still on page one.

First round of analysis to Rank in Google’s SGE

Before we touched a single piece of content, we started watching patterns. We ran dozens of searches for different queries in their niche and looked at which sites SGE was citing. The first thing we noticed was that SGE didn’t always pull from the highest-ranking sites in traditional search. Sometimes it pulled from sources that were ranked halfway down page one, or even page two.

This told us something important: SGE wasn’t just mirroring Google’s normal rankings. It was picking sources based on how useful and complete they were for building a direct answer. And that meant we had a chance to compete even against sites with stronger backlink profiles if we could make our content more “answer-friendly.”

Testing the first batch of updates to Rank in Google’s SGE

We picked three articles to test: one on installing laminate flooring, one on kitchen lighting ideas, and one on preventing mold in bathrooms.

The first step was to read them as if we were the AI. Did they answer the main question completely? Did they cover related sub-questions without forcing a click away? Were they written in a clear, easy-to-rephrase style?

In all three cases, we found gaps. The articles were written for human readers who would happily click to another page for more detail. But SGE doesn’t do that because it wants to package the whole answer right there in its summary. Basically, we realized we needed to add more context right into the article itself.

Looking closer, you could tell right away the way things were laid out wasn’t helping. The points were scattered and the key details were tucked away where you wouldn’t expect them. We figured it made more sense to lay it out so every paragraph focused on just one point. That way, both readers and AI could follow along without wading through long, confusing blocks of text.

What we changed in the content to Rank in Google’s SGE

We didn’t rewrite everything from scratch. We added clearer explanations and made sure the main points were all covered together, so everything felt more organized. For the flooring article, that meant adding sections on tools needed, estimated time per step, common mistakes, and even brief safety notes.

For the kitchen lighting piece, we expanded it to include types of bulbs, how lighting affects color, and examples of different layouts. For the mold guide, we wanted to go past just removing it. So we also laid out how to keep it from showing up again and added why those prevention steps matter.

The tone stayed casual, though we stripped away the stuff that didn’t actually add anything useful. We also broke up some dense sections into smaller, clearer units so the AI could lift them without rewording too much.

Watching the first signs of movement to Rank in Google’s SGE

Within two weeks of updating, we started seeing small but encouraging changes. In test searches, our client’s site began appearing as one of the sources cited in SGE’s answers for two of the three updated topics. It wasn’t in every variation of the search yet, but it was enough to know we were on the right track.

The interesting part was that the article on preventing mold the one with the most thorough updates and the clearest structure showed up most consistently. That reinforced our theory: completeness plus clarity was the winning combination.

Expanding the strategy to Rank in Google’s SGE

We took what we learned and applied it across the client’s top twenty articles. This was a bigger lift. Each piece had to be reviewed for completeness, updated with fresh data if needed, and restructured where it wandered too much.

We also started including examples and scenarios where possible. AI-generated answers seemed to favor sources that included real-life applications, probably because they made the summary feel more natural. So instead of just saying “use a dehumidifier to prevent mold,” we added a short example of a homeowner using one during a particularly humid summer and seeing results in a few weeks.

Results after three months to Rank in Google’s SGE

By the three-month mark, we were seeing the client’s site cited in SGE’s answers for about half of their main keywords. In some cases, they were the only small-to-mid-sized site mentioned alongside much bigger competitors.

Traffic from traditional SEO didn’t drop, which was important this wasn’t about replacing old rankings but adding a new layer of visibility. And even though it’s hard to directly measure clicks from SGE, we saw brand searches go up. That told us people were seeing the name in the AI answer and then looking for the site directly later.

What didn’t work for Rank in Google’s SGE

Not every experiment paid off. We tried adding extra “SEO-style” keyword phrases to some articles, thinking it might help SGE match the content more easily. In practice, it didn’t seem to make a difference and in some cases, it actually made the writing clunkier, which might have hurt our chances.

We also tested creating separate “short answer” pages thinking they’d be easier for SGE to pull from. That didn’t work either. The AI seemed to prefer richer pages that could supply multiple parts of an answer, even if it only quoted a small piece.

Lessons we’re taking forward

If I had to sum up what we learned, it’s that ranking in Google’s SGE is less about traditional keyword battles and more about writing the kind of resource an AI would want to lean on if it had to answer the question without leaving gaps.

That means your content has to be:
Complete enough to stand alone.
Clear enough to lift without heavy rewriting.
Current enough to avoid looking outdated.

And it has to read in a way that works both for humans and for an AI blending multiple sources.

The bigger picture to Rank in Google’s SGE

SGE is still evolving. What works now might need adjusting in a year, just like traditional SEO has shifted over time. But the principle feels solid: make your content the easiest, safest, and most useful choice for an AI to recommend.

For this client, that meant going beyond what we thought of as “good SEO” and leaning into completeness, clarity, and conversational tone. The result was more visibility in the newest, most competitive spot in Google search the AI answer box.

And honestly, the best part was seeing the shift in how the client thought about content. They stopped writing only for humans or only for search engines and started writing for both at once with the AI acting as the bridge. That’s where I think the future is heading, and if you get there early, you’re not just keeping up. You’re setting the pace.