If you have talked to a marketing company in the last six months, you have probably been told that AI now answers most local searches. The number that gets quoted is 68 percent. That number is real and it comes from a legitimate study. But it is an average across three very different kinds of searches, and once you pull them apart, the picture for a fireplace store looks close to the opposite of the warning.
This matters because dealers are being sold expensive fixes for a problem they may not have, while a real change goes unnoticed. Both halves of that are worth your time.
Where the 68 percent comes from
The figure traces back to research published by Whitespark in May 2025. They ran 540 searches by hand across Houston, Phoenix, and Denver, covering six industries: plumbers, personal injury lawyers, dentists, optometrists, medical clinics, and real estate agents. Averaged across every search they ran, Google showed an AI Overview 68 percent of the time and a traditional local pack 39 percent of the time.
Two things are worth knowing before anyone builds a strategy on that. The study is from May 2025, so it is more than a year old and Google has not stood still. And not one of those six industries is retail. Nobody has published the same test for hearth, patio, or barbecue. Anyone quoting 68 percent at a fireplace dealer is applying a dentist's number to a showroom.
The average hides the part that matters
Whitespark sorted their searches into three types, and the results split so sharply that the overall average stops being useful. Here is what they found.
Read that top row again. For the search where somebody is looking for a store, the map pack showed up 93 percent of the time and the AI answer showed up 15 percent of the time. That is the search that ends with a person walking into a showroom, and it has barely moved.
Your map pack is not the thing under threat
When a homeowner in your market types “fireplace store near me” or “wood stove dealer” plus a town name, Google reads that as local intent. It wants to show a map with three businesses on it, because that is what answers the question. An AI paragraph about what a wood stove is would be a worse answer, and Google knows that.
So the fundamentals that have always driven the map pack still drive it. A complete and active Google Business Profile, real reviews arriving steadily, accurate hours and address information everywhere Google checks. If your dealership is not showing up for those searches today, the cause is almost certainly one of the ordinary ones rather than anything to do with AI. We wrote about diagnosing that separately.
Not sure whether your dealership shows up for the searches that matter in your market? That is a specific, checkable thing and it takes about twenty minutes to find out. Book a free consultation and we will look at your market with you.
What you actually lost
Now look at the other two rows, because that is where something real happened. Informational and hybrid searches now return an AI answer 92 and 97 percent of the time. In the hearth business, those searches look like this:
- “wood stove vs pellet stove which is better”
- “how much does it cost to install a gas fireplace insert”
- “do I need a liner for a fireplace insert”
- “is a direct vent fireplace worth it”
- “what size wood stove do I need for 1500 square feet”
- “gas logs vs gas insert cost”
Every one of those is a question a homeowner asks before they ever search for a store. It is the part of the process where they figure out what they want, what it costs, and whether the project is realistic. For years that research happened on somebody's website, and a dealer who answered those questions well got to shape the answer and pick up the visit.
That conversation now happens inside a Google answer box, and often the homeowner never clicks through to anyone. They arrive at “fireplace store near me” already knowing what they want and roughly what they expect to pay. The map pack still shows your name. It just shows it to somebody whose mind is more made up than it used to be.
Why that costs more than it sounds like
Losing research traffic is easy to shrug off, because that traffic never converted well anyway. Somebody reading “wood vs pellet” is months from buying, and dealers have always known that.
The cost is not the traffic. It is that preference forms during research, and you are no longer in the room when it happens. If the AI answer explains that pellet stoves need electricity and a yearly service visit, and your showroom is the one in town with the strongest pellet lineup, that framing shapes what walks through your door. When the buyer arrives with a fixed idea and a price expectation set by a national average, the conversation starts further from the sale than it used to.
There is a second effect. Google's AI answers are built from sources, and being one of those sources is now a form of visibility. In one plumber dataset from the same research, 60 percent of the sources cited were third-party publishers and 40 percent were individual businesses. Local businesses do get cited. It is simply not automatic.
What to do about it
Answer the research questions on your own site, specifically and locally. Most dealer websites have a product page for every brand they carry and nothing that answers a homeowner's actual question. A page called “What a gas insert costs to install in [your county]” with real ranges and the local variables that move the number is more useful than another brand page, and it is the kind of page that gets cited. Generic content pulled from a manufacturer will not do this. Specific local knowledge will. This is the highest-value thing most dealer sites are missing.
Keep the local fundamentals strong, because the pack still shows. The map pack appears on 93 percent of the searches that bring buyers to showrooms. Your Google Business Profile, your review flow, and your listings across the web are still doing the heaviest lifting in your entire marketing program. Nothing about AI changes that.
Be skeptical of AI SEO packages sold on the 68 percent number. If somebody opens with that figure and does not mention that it is an average across six non-retail industries from May 2025, they either have not read the study or they are counting on you not having read it. Ask which of the three intent types they are proposing to fix, and ask what they expect it to do for showroom visits. A good answer to that exists. It just is not the panic pitch.
Do not stop advertising on the assumption that search is finished. Paid search results sit above and around all of this, and for a dealer they still put you in front of somebody with local intent at the moment they have it. The mechanics of that have not changed.
The honest limits of this data
We would rather give you a number with its caveats attached than a clean one that falls apart later. So: the study behind everything above ran in three cities across six industries, none of them retail, and it published in May 2025. Google ships changes to search constantly. The intent split is the most useful thing anybody has measured on this question so far, and it is still a proxy for the hearth business rather than a measurement of it.
What holds up regardless is the shape of the finding. Searches where somebody wants a business still show businesses. Searches where somebody wants an explanation now get an explanation. If your marketing depends on being found by people who want a fireplace store, you are in better shape than the headlines suggest. If it depended on being found by people who wanted to learn about fireplaces, that ground moved, and the response is to be useful and specific enough that the answer engines quote you.