Here is the short answer. For established local businesses and nonprofits, website traffic did not collapse. It moved.
Our team manages Google Analytics for dozens of local organizations. So instead of guessing, we looked. We reviewed local businesses and nonprofits with solid traffic data and here is what we found in that group, comparing the year before AI Overviews to the year after:
- Website traffic from search grew a median of 6.6%.
- Calls and direction requests from Google Maps grew 21.4%.
- Traffic from people inside the local metro area grew 11.9%.
- Traffic from out of state fell 10.5%.
- Blog and article pages fell 20.0%.
Read those last two lines again. Something did leave. It just was not your customer. If your dashboard shows a drop, the visit may still be happening in a place your dashboard cannot see.
Why “My Traffic Is Dead” Feels So True Right Now
We hear a version of this on almost every call. The traffic is dead. Nobody is finding us. I am pouring money into a hole and I cannot tell if any of it is working.
And then, almost always, something quieter: I built this website myself and I never really understood how any of it worked.
If that is you, hear one thing before the numbers. This is not something you failed to learn. The way local businesses get found genuinely changed, and the tool most owners use to measure it did not change with it. Being confused by that is not a flaw. It is the right response to a broken gauge.
For the plain-English version of the change itself, our team wrote a separate guide on what Google’s AI Overviews mean for your local business. This post is about what the data shows.
The Headlines Measured a Different Internet
Those big loss numbers are real, and we are not going to wave them away. But look at who got measured. Ahrefs studied 300,000 keywords and found that when an AI Overview appears, the top result gets about 34.5% fewer clicks. Those are mostly broad questions. Google answers those on the results page now, so nobody clicks through.
A local plumber is not a recipe blog. When a neighbor searches for a plumber to fix the broken hot water heater, they do not want an essay. They want a phone number and someone who will show up. That is a different kind of search, and it behaves differently.
Nobody in those viral studies asked what happened to the family clinic, the community nonprofit, or the local real estate team. That is the gap our study was built to fill.
How We Kept Ourselves Honest
Analytics is full of traps that make bad math look impressive. So before comparing anything, we threw out four kinds of noise: businesses that closed or lost tracking, sites too new to compare, tiny sites where a few visits swing the percentage wildly, and sites that changed domains.
That last rule matters most. One clinic in our data showed 554% growth. It was two websites being merged into one, not a single visitor Google sent. So we cut it from every number in the study.
We also report medians, not averages. One lucky outlier can yank an average upward. A median just means half did better and half did worse. An honest plan is better than a hopeful promise, and the same goes for statistics.
Where the Traffic Actually Went
Into Google Maps
Calls, direction requests, and profile views from Google Business Profiles grew 21.4% in our group, while website visits grew 6.6%. Customers are calling you or driving to you straight from the search results, without ever loading your homepage.
That is a lost pageview and a won customer. Whitespark found AI Overviews in 68% of the local business searches it tested, which is exactly the layer where this happens. Google says local results depend on how complete and accurate your business information is. Your profile is doing more work than your website right now.
Into Your Name
Homepage traffic, which is the closest stand-in for people searching your business by name, grew 5.9%. Google rarely puts an AI summary in front of someone who typed your name.
BrightLocal surveyed 1,002 US consumers and found that 88% of people using AI for local recommendations check the sources or verify the reviews before acting. AI names you. Then they come looking for you. Your name is the moat. This is the power of brand!
We have watched this happen slowly with the Kim Mixon Real Estate Team here in Columbus. They stayed consistent with our team for years, and that steady presence is now showing up inside AI-generated results. It was not a trick. It was time.
Into Your Zip Code
Local metro traffic grew 11.9%. Out-of-state traffic fell 10.5%.
This is our favorite finding, because it explains a loss owners feel that the bank statement does not show. AI absorbed the faraway readers who were never walking in your door. The neighbors kept clicking. Fewer visitors, better visitors.
What Really Did Leave
Blog and article pages fell 20.0%, while service pages grew 5.9%.
So if your plan was writing general how-to articles and waiting for Google, that plan took real damage, and we are not going to smooth that over. The pages explaining what you do and how to hire you are holding. The pages answering generic questions are being eaten, on your site the same as anywhere else.
If You Run a Nonprofit, Read This Part Twice
Here is something the cheerful version of this study would skip. In our group, the local businesses mostly gained. The nonprofits mostly did not. Three of the four nonprofit and community organizations lost traffic, between 8% and 16%. Only one grew.
Four organizations is a small number, so treat that as a warning sign and not a proven rule. But it fits the pattern above. Nonprofit websites lean heavily on explaining and educating, and that is the exact kind of content AI is absorbing.
We saw the other side of this with the National Ranger Memorial Foundation. Their old site could not tell the story of the Memorial, and that was costing them awareness and donations. Rebuilding around the story, rather than around information, is what changed it. In an AI search world, that is not a design preference. It is survival.
Count These Five Things Instead
This is the practical part. If your only number is website sessions, you are reading one gauge on a dashboard that now has five. Track these:
- Phone calls from your Google Business Profile. Available free in your profile dashboard.
- Direction requests. Same place. For anyone with a front door, this is a foot-traffic forecast.
- Searches for your business by name. Rising branded search is the healthiest signal you have.
- Form fills and calls sorted by city. Ten leads from Columbus beat a hundred readers from Ohio.
- Service page visits, counted separately from blog visits. Blend them and a blog decline will hide real growth.
If sessions wobble but calls, directions, and local inquiries hold steady, your presence is working. Do not let a vanity number talk you out of a strategy that is producing customers. Our team put together a longer walkthrough on measuring local SEO success if you want the step-by-step.
What This Study Cannot Tell You
We would rather be trusted than impressive, so here are the limits. This is one agency’s client list, concentrated in the Southeast. Eighteen clean properties is a real group, not a national census. We measured visits, not revenue. Year-over-year windows always carry some seasonal noise. And these findings describe established local organizations with real websites and working analytics, not publishers, national retailers, or brand-new sites.
The Bottom Line
AI Overviews did not erase local search. They filtered it. The casual traffic went to the machines, and the customers kept coming, more often through your Google profile and your name than your blog.
AI opens the door. Your website still closes the deal.
We have been a Columbus digital marketing agency since 1998, and after nearly 30 years of watching this story repeat, our read is that this is a measurement problem far more often than a demand problem. Which side of these numbers you are on is a question your own data can answer.
Why the Traffic You Do Get Matters More Than Ever
Here is a number worth sitting with. A separate Microsoft Clarity study of more than 1,200 websites found that visitors referred by AI assistants like ChatGPT sign up at almost eleven times the rate of visitors from a regular search click, 1.66% versus 0.15%.
That lines up with what we found in our own data. The traffic AI sends you through your name and your Google profile is not just holding steady, it is arriving pre-sold. The AI already answered the easy questions and pointed a real buyer your way before they ever loaded your homepage.
Which raises the harder question. If an AI assistant only names two or three businesses when someone asks who does what you do, and the traffic behind that recommendation converts at eleven times the normal rate, not being one of those names is not a small miss. It is the best leads in your market going to someone else by default.
That is exactly why your Google Business Profile, your reviews, and the name recognition we talked about above are not busywork. They are what gets you on the short list AI reads from.
Get the full study. The white paper includes all five findings, the property-by-property table, our exclusion rules, and how our numbers compare to published industry research. Download the full report here.
Prefer to skip ahead and just find out where you stand? Our Digital Checkup reviews your analytics, your Google Business Profile, and your site’s AI readiness, then hands you a plain-language report on what to fix first. It is $149, credited toward your first project or plan. Schedule a call and get clarity.
Frequently Asked Questions
Did AI Overviews really cut traffic by 80%?
For some websites, yes. Those were mostly national publishers, recipe sites, and affiliate review sites that earn traffic by answering broad questions. In our group of 18 established local businesses and nonprofits, the median change was a 6.6% gain.
Why does my Google Analytics still show a decline?
Two likely reasons. First, blog and informational pages are genuinely down about 20%, and if you have a lot of them they can drag your total. Second, Analytics cannot see a customer who called you straight from Google Maps. That visit never touched your website, so it never got counted.
Should I stop writing blog posts?
No, but change what they are for. Generic how-to articles are being absorbed by AI. Content that answers the specific questions your actual customers ask, in your actual service area, still earns its keep and gives AI something to cite.
Is my Google Business Profile more important than my website now?
They do different jobs. Your profile is increasingly where people find you, since Maps actions grew 21.4% in our data. Your website is where they decide. Neglecting either one breaks the chain.
What should nonprofits do differently?
Watch your numbers more closely than a for-profit business would. Most of the nonprofits in our group lost traffic while the businesses gained. Lead with story and mission rather than general education, and track donations and inquiries rather than pageviews.
How do I know if this is happening to my business?
Compare this year to last year in three places, not one: website sessions, calls and direction requests in your Google Business Profile, and searches for your business name. If two of the three are healthy, you are in better shape than your traffic chart suggests.
Does traffic from AI search actually convert better than regular search traffic?
Yes. A Microsoft Clarity study of more than 1,200 websites found that visitors referred by AI assistants signed up at almost eleven times the rate of visitors from regular search, 1.66% compared to 0.15%. By the time an AI sends someone your way, it has usually already answered their basic questions and compared their options.
