ai-overviews-traffic-recovery

We work with a lot of local service businesses across Maryland, and every so often one of them calls with the same worried question: “our rankings look fine, so why has the phone stopped ringing?”

That’s exactly how this project started, with an HVAC and plumbing company based in Anne Arundel County that had spent two years building a solid organic presence, only to watch their lead volume slide without any obvious explanation.

This isn’t a hypothetical. It’s a real client, details adjusted slightly for privacy, and a real problem that a lot of Maryland business owners are going to run into as Google’s AI Overviews keep expanding into more search results. Here’s what actually happened, how we figured out the cause, and what we did to turn it back around.

The Client and the Problem

This client had ranked in the top three positions for their most valuable keywords, things like “furnace repair cost Annapolis” and “emergency plumber near me,” for well over a year. Rankings hadn’t changed much. But starting in late winter, their call tracking numbers told a different story. Organic leads were down noticeably, and nobody on their end could point to a single thing that had changed on the site.

They’d already ruled out the usual suspects on their own before calling us. No recent site redesign, No pricing changes that would scare off leads, and No negative reviews piling up. The owner told us flat out that his ad spend and referral numbers were steady, which meant whatever was happening was isolated to organic search specifically, and that narrowed things down fast.

That disconnect, strong rankings paired with weaker results, is one of the clearest signs something is happening above the traditional results, not within them. We see it more and more often now, and it almost always points back to the same culprit.

The Drop We Almost Missed

The first thing we checked was the obvious stuff: technical errors, a Google Business Profile suspension, a bad backlink or a manual action. None of that was it. The site was healthy. Rankings, when we checked manually, hadn’t dropped at all.

What tipped us off was pulling click-through rate by query inside Search Console and comparing it month over month. Several of their top-performing pages were still ranking in position two or three, but the click-through rate on those same queries had dropped by close to 40 percent. Rankings hadn’t moved. Clicks had.

That’s the pattern an AI Overview leaves behind. The page is still there, still ranking, still technically doing everything right, but a chunk of searchers never scroll past the AI-generated summary sitting above it.

Digging Into the Data

We took the client’s top twenty keywords by traffic and manually checked each one in Google to see which queries now triggered an AI Overview. Twelve of the twenty did. Every single one of those twelve had lost click-through rate over the previous ninety days. The remaining eight, the ones without an AI Overview showing, had held steady or even improved slightly.

That comparison made the cause pretty hard to argue with. This wasn’t a ranking problem, a content quality problem, or a seasonal dip. It was a visibility problem sitting entirely above where their rankings lived. Once we could show the client that side-by-side breakdown, the conversation shifted from “what did we do wrong” to “what do we do about it,” which is a much more useful place to start.

Why Their Pages Were Losing to the Overview

Once we knew where the damage was, we needed to understand why Google’s AI Overview wasn’t pulling from this client’s content in the first place, since being cited inside the Overview is a very different outcome than being buried beneath it.

A few patterns showed up fast. Their service pages opened with a paragraph of company background before ever answering the actual question a searcher typed in. Pricing information, when it existed at all, was vague, phrases like “affordable rates” instead of an actual range. And there was no structured FAQ content on any of these pages, which meant there was nothing clean for an AI system to lift and quote.

We also noticed their competitors weren’t necessarily doing anything more sophisticated. In a few cases, the business now getting cited inside the AI Overview had a noticeably weaker backlink profile and a lower domain authority score than our client.

What they had instead was a page structured around a direct answer, with a clear cost table and a short, specific FAQ sitting right below it. That was a useful, if slightly humbling, thing to show the client. Beating a competitor with fewer resources wasn’t about outspending them. It was about being easier to summarize accurately.

In short, the content wasn’t wrong. It just wasn’t built to be extracted. That distinction matters a lot right now, and it’s the core of the AI search optimization work we ended up doing for this client.

What We Changed, Page by Page

We didn’t rebuild the entire site. We focused on the twelve pages tied to the affected keywords and rewrote each one with the same approach.

Every page now opens with a direct, two or three sentence answer to the exact question in the title, before any company background or extra context. We added specific numbers wherever we’d previously used vague language, actual price ranges for common repairs, actual response time windows, actual service area boundaries down to the county and zip code level.

We also added a structured FAQ section to each page, written around the real questions this client’s customer service team said they heard on the phone every week.

For the furnace repair page specifically, that meant replacing a single vague sentence about pricing with an actual breakdown: what a basic repair typically runs, what a part replacement adds, and what pushes a job into emergency pricing territory.

It felt uncomfortable for the client at first, since pricing transparency isn’t something every contractor is used to putting in writing. But it’s exactly the kind of specific, checkable detail that AI systems tend to favor over softer marketing language.

None of this was complicated. It was closer to editing than writing from scratch. But it changed how easy these pages were for an AI system to summarize accurately, and that turned out to matter more than any single ranking factor we could have chased instead.

Rebuilding Trust Signals Across the Board

Content changes alone weren’t going to be enough, so we also cleaned up the client’s broader online presence. Their Google Business Profile was missing several service categories they actually offered, and their business hours hadn’t been updated after a schedule change nearly a year earlier.

We audited their citations across major directories and found three with an outdated phone number still active, which we corrected. We also added author credentials to their blog content, since a lot of it had been published without any indication of who wrote it or what qualified them to answer HVAC and plumbing questions in the first place.

For a company whose technicians are all licensed and have been in the trade for years, that was an easy fix once we knew to look for it, just a short author bio at the bottom of each guide noting years of experience and license number.

Consistency like this rarely gets attention until something forces the issue, and it’s a big part of the work we now build into local SEO services for every client, not just the ones already dealing with a drop.

It’s easy to treat this kind of cleanup as busywork, but AI systems are essentially trying to verify a business is real and current before recommending it, and stale information is exactly what makes that verification fail quietly, without ever showing up as an obvious error anywhere.

Getting Referenced Inside the Overview Itself

About six weeks into the changes, we started manually checking those same twelve queries again to see if anything had shifted. Four of them now showed this client’s business named directly inside the AI Overview text, a plumber in Annapolis being cited by name for an emergency service query, and a cost breakdown for furnace repairs being pulled almost word for word from their rewritten page.

That’s the actual goal here, and it’s worth being clear about. The point was never to make the AI Overview disappear. It’s not going anywhere. The point was to go from being buried underneath it to being one of the sources it actually pulls from and credits.

Where Things Stand Now

Ninety days after we started, this client’s organic traffic had climbed back above where it sat before the drop, not just recovered but about 14 percent higher than their previous baseline. Call tracking data showed a similar pattern, with lead volume from organic search up noticeably over the prior quarter.

Just as important, the click-through rate gap on those twelve affected keywords had closed almost entirely. A few queries never fully bounced back to their old numbers, and we’re honest with clients about that; some AI Overviews are simply going to absorb a portion of clicks no matter how well a page is built. But going from a 40 percent click-through drop to something closer to a 5 percent gap is a real, measurable win, not a marketing talking point.

The owner mentioned something during our final check-in call that stuck with us: his front desk staff had started noticing more callers say things like “I saw your pricing online” before ever being asked. That’s not a metric that shows up in an analytics dashboard, but it’s a pretty direct sign that the specific, transparent content was doing its job even beyond the traffic numbers.

The Bigger Lesson for Maryland Businesses

The takeaway here isn’t that AI Overviews are unbeatable, and it isn’t that they’re harmless either. It’s that ranking well is no longer the finish line it used to be. A page can hold its position perfectly and still lose the customer if it isn’t written in a way that earns a place inside the answer sitting above it.

For any Maryland business watching their rankings stay steady while calls quietly drop, this is usually where to look first. It’s rarely a single big fix. It’s closer to a series of small, deliberate ones, rewriting how a page answers its core question, tightening up business information that’s gone stale, and giving Google’s systems a clean, specific answer instead of a vague one.

If your traffic has been sliding in a way that doesn’t line up with your rankings, that pattern is worth checking before assuming something bigger is wrong. Our team walks through this exact audit with new clients regularly, and you’re welcome to reach out and talk through what you’re seeing.