AI tools are getting better at a lot of things. LSA advice is not one of them.
I ran a simple test. I asked Google's own AI, Gemini, to give me a bullet list of Local Services Ads best practices. The output came back confident, organized, and structured like something you would find in a training deck. It looked authoritative.
It was also wrong in several places. Not subtly wrong. Wrong in ways that would cost a business money, damage their ranking, and send an agency chasing problems that do not exist.
I am going to go through the output point by point. Not to dunk on Gemini. AI hallucination is a known problem and this is not unique to Google. But the specific errors reveal something important about how LSA actually works, and how far the public understanding of the platform lags behind reality.
If you are using AI tools to inform your LSA strategy, this is worth reading carefully.
The Badges Have Been Consolidated
"Earn the Google Guaranteed or Google Screened badge."
Both badges have been consolidated into a single credential: Google Verified.
The Guaranteed and Screened distinction no longer exists in the platform. If an agency is still presenting these as separate tiers to clients, they are working from a version of the product that has already changed.
On Service Area: Wider Is Better, Up to a Point
"Only select the zip codes and cities you can realistically and profitably serve. Extending your area too far leads to ignored leads."
Google rewards wider service area configuration. Set it as wide as you can operationally support, not as narrow as you feel comfortable with.
The intent is right but the framing is wrong in a way that leads to bad decisions. A broader coverage map signals to Google that you are available to more users, which improves your ranking and impression share. The limiting factor is your actual capacity to serve, not a conservative estimate of where you want leads from.
The concern about ignored leads is real, but the fix is operational: make sure you can actually answer and service what comes in. The answer is not to artificially shrink your footprint.
On Job Types: Most Searches Are Not for Specific Services
"Only toggle the specific services you actually want leads for. Turn off low-profit or over-capacity services."
80 to 90 percent of LSA searches are for the general category. Enable every job type you can genuinely service.
Someone searching "plumber near me" is not searching for "water heater installation" or "pipe repair." They want a plumber. The specific job types within a vertical account for a much smaller share of total search volume, somewhere around 15 percent for most home service verticals.
Restricting job types to only your preferred services is effectively narrowing your keyword coverage for no good reason.
On Photos: Show the User Journey, Not Just the Business
"Upload clear, professional photos of your team, vehicles, and completed work."
True in principle, but sequencing matters. Lead with the truck, then the technician at the door, then the technician at work. Logo shots come last.
The goal of LSA photos is to show a potential customer what it looks like when your business shows up at their door. Google rewards complete, relevant profiles. A well-sequenced set of photos signals legitimacy to both the algorithm and the person reading the listing. Generic team headshots miss the point.
On Messaging: Do Not Turn It Off
"If you cannot maintain fast response times to message leads, it is better to turn the messaging feature off."
Opting into message leads is a ranking signal. Turning it off actively works against your ranking.
Google rewards businesses that give users multiple ways to make contact. Unless you are completely unable to respond to messages, and I mean genuinely incapable, not just slow, keep messaging on. The right fix for slow response times is an operational one: build a process for handling message leads. It is not to remove the feature and accept the ranking hit.
On Lead Categorization: The CRM Is Optional
"Regularly mark leads as Booked, Completed, or Archived so Google's machine learning can understand which leads are converting."
Moving leads through the CRM sends no ranking signals. It is an internal reporting tool only.
Many businesses do not use the LSA CRM at all. They pull lead data through the API and manage it in their own systems. Google does not penalize this.
One important note on archiving: marking a lead as Archived removes your ability to rate that lead. If a lead comes in that was invalid, wrong service area, spam, or wrong category, and you archive it before rating it, you have lost the ability to flag the quality issue. Archiving is not a neutral action.
On Lead Disputes: They No Longer Exist
"Dispute invalid leads promptly within the dashboard within 30 days to get your money back."
Lead disputes were removed from LSA. Google now uses autocrediting to handle invalid leads automatically.
If an agency is still telling clients to submit lead disputes, they are describing a workflow that does not exist. There is nothing to submit. The system handles it automatically. This change has been live for a while and is documented in Google's own help center.
On LSA Reviews: They Have Been Migrated to GBP
"Use the LSA review link generated within your LSA dashboard so reviews tie directly to your Local Services Ads profile."
There is no LSA review link. All reviews have been migrated to Google Business Profile.
LSA reviews were fully migrated to Google Business Profile. All reviews now live on GBP and feed into LSA from there. The separate LSA review system is gone. This matters because it changes where you direct customers for reviews, and it underscores the importance of maintaining a healthy, active GBP as part of your LSA strategy. It is something that gets underdiscussed precisely because the two platforms look separate when they are not.
What This Tells Us
Gemini was not making things up randomly. Most of what it produced was grounded in how LSA worked at some point. The problem is that LSA has changed, significantly and repeatedly, and AI models trained on general web content have no reliable way to know which version of the platform they are describing.
The badge consolidation happened. The review migration happened. The dispute system was replaced. Job type search volume distribution is a nuance that rarely makes it into published guides. These are the kinds of details that only come from working inside the platform, or working closely with clients on it every day.
If you are using AI tools to research LSA strategy, use them to generate questions, not answers. Then go find someone who actually knows the current state of the platform to answer them. That is what we are here for.
- Google Guaranteed and Google Screened no longer exist as separate badges. Both have been consolidated into Google Verified.
- Wider service area configuration is better for ranking, up to your operational capacity. Do not artificially narrow your footprint.
- 80 to 90 percent of LSA searches are for the general category. Enable every job type you can service.
- Keep messaging leads enabled. It is a ranking signal. The fix for slow response times is operational, not turning the feature off.
- Archiving a lead removes your ability to rate it. Do not archive leads you may need to flag.
- Lead disputes no longer exist. Google uses autocrediting to handle invalid leads automatically.
- There is no LSA review link. All reviews have migrated to Google Business Profile.
Want the current state of LSA, not last year's version?
Book a call and we will walk through exactly where your account stands and what actually moves the needle today.