700 reviews and still stuck on page 2 of the map. Invisible in ChatGPT. Local visibility now happens on four surfaces at once, and most strategies only work on the first one. Nine steps, in the order that makes them work, with the ChatSEO prompt for each.
ChatSEO, the SEO agent for local SEO · 🎁 7 days free
Local visibility now happens on four surfaces at once:
The steps below run in a specific order. The order is the method. Each step feeds the next one.
For the sake of this masterclass, let's take a dentist: a practice in Ballwin, Missouri, 700 Google reviews, #20 in its own city's map pack. Here's why, and how to fix it.
The instinct when you open a new site is to start fixing. Don't. Not even the obvious mistake you just spotted.
Take a snapshot first. If you fix things before you measure, you can never prove what worked. That matters for your own learning, and ten times more if a client is paying you a retainer.
And the snapshot can't be "where do we rank on Google". For local, you track four surfaces: the map pack (with a geo-grid across the city), organic rankings, whether the business is named in AI answers, and which third-party sites those AI answers cite. That last list becomes your to-do list in step 3. Keep it.
Then install the three basics: Google Search Console, Google Analytics 4, and call tracking. For a local business, a call is the conversion. If you don't track calls, you're flying blind on the one metric the owner cares about.
Connect Search Console and GA4 to ChatSEO and your baseline is there before you change a single line:
Build my "before" snapshot for [site URL], dated today, from Search Console and GA4 (last 28 days): total clicks, impressions and average position, the top 20 queries and pages by clicks, and every query containing [city]. Do not suggest fixes yet. If a number cannot be read from the data, say so. Do not estimate.
For the dentist, the benchmark showed: #20 in the local pack, weak across Ballwin itself. Decent organic rankings, competitors ahead on several engines. Visible in Gemini, Perplexity, Claude and Grok, but not in ChatGPT, the one with the most users. And a clear list of sites cited in AI answers for local dentists.
The practice is located in Ballwin. It serves Ballwin. But the profile was optimized for Chesterfield, the neighboring city.
So Google had a practice in Ballwin telling it, in effect, "I'm about Chesterfield". Result: #20 at home. The fix takes seconds. Point the profile at the city where you actually are.
Nobody should promise results in SEO. But this pattern shows up again and again, and fixing it alone often moves a profile 5 to 10 spots in the map pack. Same for the primary category: the wrong category is the other one-second fix that moves rankings. Here the category was already right, so location was the whole problem.
Why it's so powerful: Google's local ranking comes down to relevance, distance and prominence. This practice already had prominence (700 reviews). It was losing on relevance to its own city. You don't need more reviews when the signal pointing Google at the right city is broken.
After the fix, the routine is simple: keep driving Google reviews consistently, not in bursts. Add photos of the office, the team, the equipment. Recheck location and category every time someone touches the profile. Most people overcomplicate the GBP. Get the location and category right, then feed it reviews.
This is where the playbook breaks from how local SEO worked for a decade.
The old way: build citations in bulk. Submit to 100+ directories, tick the box, move on. Honestly, it barely mattered. Citations matter again. But not as a volume game.
AI engines answer "best dentist in Ballwin" by retrieving pages and citing them. Go back to the AI citations from step 1 and look at which third-party sites appear. For this dentist, four came up: HealthGrades, Delta Dental (the insurer's dentist directory), Yelp, and ADA.org. That's the list. Not 150 directories. Four.
Across most local industries, 4 to 7 citation sources actually show up. The rest are probably not worth your time. And the list is different for every niche: a plumber's won't include HealthGrades, a lawyer's will include sites a dentist never needs. You don't guess it. You read it off the AI citations.
Analyze the AI search results for "best [business type] in [city, state]" and the 5 closest variations. List every third-party site cited in the answers, how many times each one appears, and whether [business name] is listed on it today. Rank them by citation count. Give me only the top 4 to 7 as my citation list. Do not add directories that were not cited.
Being listed on HealthGrades isn't enough. You need to rank inside HealthGrades.
Search the directory for dentists in the area and the practice is buried far down the list. So when an AI engine retrieves that page, it sees other brands first. Three moves to fix it.
1. Consider the sponsored slots. The HealthGrades results page had 3 sponsored listings at the top. AI engines don't care whether a listing is paid. They see the brands at the top of the page. If the directory is a top citation in your niche, a sponsored slot might be one of the cheapest AI visibility plays available.
2. Split your review effort 90/10. 90% of your review requests go to Google. The other 10% go to these secondary review surfaces. That's how you climb inside the directory and give AI engines fresh signals from a source they already trust.
3. Rewrite the profile like a web page. The practice's HealthGrades profile was thin: strip the boilerplate and one sentence was left. Treat it exactly like a page on your site. Get the topics and entities from the pages already ranking for "cosmetic dentist Ballwin", draft around 500 words, compress to 300 or 150 if the platform has a limit, and make sure each service (veneers, implants, Invisalign) is covered naturally.
In ChatSEO, that's one message. You paste the result into the directory yourself:
Analyze the top 10 results for "[cosmetic dentist Ballwin]" and list the topics, entities and services they cover. Then write a directory profile for [business name] of about 500 words that covers those topics naturally and mentions each service we offer: [veneers, implants, Invisalign]. Then give me a 300-word and a 150-word version that keep every key topic. Do not invent credentials, awards or numbers I did not give you.
Bonus: a well-optimized directory profile can rank in Google's organic results on its own. One more slot on page 1 that belongs to you, even though you don't own the domain. Repeat for every directory on your list.
AI engines build their picture of your business from everything written about it online. If that picture is contradictory, you lose trust and visibility.
For this practice, the audit found something no rank tracker would: the practice had changed hands. The current dentist took it over, and the web was still full of the previous dentist's info. Old names, old details, scattered across directories and profiles.
That's a massive to-do list from one prompt. Clean it up source by source, starting with the sites that appear in AI answers or on page 1.
Search for every mention of [business name], [address] and [phone number]. List every inconsistency you find in business name, address, phone, hours, practitioner or owner name, services and website URL, across directories, review sites, social profiles, maps and news. For each one: the source URL, what it says, the correct value, and a priority (high if the source appears in AI answers or on Google page 1). Flag anything linked to a previous owner, old address or old phone number.
This site wasn't in bad shape. Two issues stood out, and they're the same two you'll find on most local sites.
Over-optimized title tags. "Dental Bonding in Ballwin, Missouri" is fine. Then came variations stacking the city, the state abbreviation and the ZIP code. The rule: one clean version. "Dental Bonding in Ballwin, MO | Practice Name". LLMs and Google both understand that "Ballwin, Missouri", "Ballwin MO" and the ZIP refer to the same place. Stuffing all three adds nothing, and it probably hurts.
Boilerplate meta descriptions. Many pages shared the same generic one. In 2026 there's no excuse: every page can have a unique description in minutes.
One rule most people miss. Don't put your phone number in the title or meta description. If the searcher can grab your number from the snippet, they don't click, and organic CTR feeds rankings. This only applies to organic results: on your Google Business Profile, the phone number obviously stays.
This is the step nobody should do by hand anymore. With ChatSEO connected to Search Console and your WordPress or Shopify site, you review the rewrites, approve, and they're live. For a 40-page local site, minutes instead of an afternoon.
Crawl [site URL] and find every title tag that stacks more than one location signal (city + state + ZIP), every meta description shared by more than one page, and any title or meta that contains a phone number. For each page, propose one clean title in the form "Service in City, ST | Business Name" and a unique meta description under 155 characters with no phone number. Show me a before/after table with each page's clicks from Search Console. Do not publish until I approve.
The practice already had the right structure: "Invisalign in Ballwin, Missouri". Service plus core location. Keep that structure.
The problem was the optimization. The homepage had too many words and was aggressive without being well optimized. A good local page has a word count within the median range of the competitors ranking for that keyword (not double, not half), covers all the critical topics and entities for that service in that location, and is optimized for one primary topic and one location.
The workflow: look at what ranks for "cosmetic dentist Ballwin Missouri", list what those pages cover and how long they are, then write a page that matches the range and fills the gaps. ChatSEO runs that SERP analysis, drafts the page, and publishes it to your CMS once you've checked it. Design still needs a human pass.
Analyze the top results for "[Invisalign Ballwin Missouri]": median word count, the topics and entities every ranking page covers, and the page type Google rewards. Then draft a page for [business name] targeting that one service in that one city, inside the median word range, covering every topic listed, optimized for one primary topic and one location. Flag any claim about the business you could not verify from my site. Do not invent prices or results.
Priority order: main service pages first, then every secondary service. Only after that do you move to content.
Why these prompts need ChatSEO and not a chat window. Steps 1, 6 and 7 read your live Search Console rows, crawl the actual site and push the approved change back into your CMS. A prompt pasted into a plain chatbot can't see the rows you didn't paste, can't refresh them next month, and can't check whether last week's fix moved the number. That's the whole reason ChatSEO connects to Search Console, GA4 and your CMS. The first 7 days are free.
This is the mistake almost every local blog makes.
The practice's blog had articles like "How long do dental implants last?". That's a national question. Thousands of sites answer it, AI answers it directly, and it does nothing for a practice that only serves Ballwin.
The rule: every piece of content should build relevance for your location and your expertise. What that looks like in practice: statistics-driven, local linkbait. Content built on Ballwin-specific data that local news, bloggers and organizations would want to cite. Some ideas will be good, some won't. But ideas are no longer the bottleneck. Execution is.
I run [business type] in [city, state]. Using the queries my site already gets impressions for, give me 15 content ideas built on local data I can collect or compile myself (a local survey, public records, price comparisons across the area, local rankings, seasonal trends). Each idea must be specific to [city], something local media, bloggers and organizations would want to reference. For each: the headline, the data source, and 3 local sites likely to link to it.
Local links build local topical authority, the signal that says "this business belongs to Ballwin".
One well-written prompt produces a full list of city-specific link prospects in about 4 minutes. Work that used to take a full day of manual prospecting.
Find link opportunities for a [business type] in [city, state]. Include local news sites, chambers of commerce and business associations, local events open to sponsors, schools and sports clubs, local bloggers and podcasts, "best of [city]" lists, nonprofits, and complementary businesses. For each: name, URL, why they would link, and the exact angle to pitch.
Then you do the part AI can't: reach out, sponsor, show up.
It runs from cheapest and fastest to slowest and most compounding. Steps 1 to 4 can move rankings within weeks. Steps 7 to 9 are what keep you there when competitors wake up. And steps 3 and 4 only exist because you tracked AI citations in step 1.
Connect Search Console, GA4 and your CMS once. Then paste the prompts above, one by one, and approve what ChatSEO proposes. 7 days free, cancel anytime.
Start your 7-day free trial →