A client sent me his Google Business Profile analytics last March. His profile views had dropped 38% over four months. Rankings for "plumber [city name]" — his main money term — had slipped from position 3 to position 9. He'd just spent €1,200 on a new website.
The problem wasn't the site. It was his citations.
His old address was still live on fourteen directories. Two listings had a phone number he'd cancelled in 2021. One had his business name spelled with an apostrophe that didn't exist. Google had been getting contradictory signals for over a year, and eventually it stopped trusting the whole picture.
That's what local citations do. They don't rank you on their own. They corroborate you. And when the corroboration is messy, you pay for it in the map pack.
Key Takeaways
- Local citations are any online mention of your business name, address, and phone number (NAP) — structured or not.
- Consistency beats volume every time: 20 clean citations outperform 200 messy ones.
- Most listing errors can be fixed for free; paid services mostly buy you time, not magic.
- Quality checks matter — some directories actively hurt you through spam association.
- Track citation work like a campaign: fixed baseline, fixed review date, before/after comparison.
What are local citations, and why they matter more than most people think
A local citation is any online reference to your business that includes identifying information — usually name, address and phone number, though citations with a website, opening hours, category or photo also count.
Two families exist. Structured citations live on directories and platforms that have dedicated fields for your details: Google Business Profile, Yelp, Bing Places, industry-specific directories. Unstructured citations are mentions in blog posts, news articles, chamber-of-commerce pages, event listings — anywhere your NAP shows up in body text rather than form fields.
Why Google cares at all
Google's local algorithm weighs three broad signals: relevance, distance, and prominence. Citations feed the third one. Each consistent mention is a small independent confirmation that this business exists, operates at this address, and answers that phone. Twenty of them pointing the same direction make the entity legible. Twenty pointing in different directions make it noise.
I've watched this play out in reverse. A dental clinic I worked with two years ago had eleven duplicate Google-adjacent listings across aggregator-fed sites, three with an obsolete suite number. After cleanup, map pack visibility for "dentist near [neighbourhood]" moved from position 7 to position 4 in roughly six weeks. No other change. Same site, same reviews, same photos.
How to build local citations without wasting money
Here's the workflow I actually use, refined over too many wasted hours on directories nobody reads.
Step 1: audit before you build
Search your business name in quotes. Then your exact phone number in quotes. Then your address. Do it on Google, Bing, and DuckDuckGo — the result sets differ slightly and you'll catch listings the first engine buried.
Log everything in a spreadsheet: platform, URL, name variant, address variant, phone variant, whether you have login credentials, and status. Expect to find between 10 and 40 existing listings for a business older than three years. Most will be wrong in some small way.
Step 2: prioritise by something that isn't domain authority
Domain authority is a proxy metric. What actually matters is whether the directory gets real humans and whether it's indexed. My qualification checklist:
- Does the page appear in Google's index? (Search the business name plus site:domain)
- Does it attract genuine traffic, or is it a scraper feed of other listings?
- Is the niche right? A local plumbing directory beats a general business directory with a higher score.
- How many ad placements does the page carry? Heavy ad walls usually signal low editorial standards.
- Any spam association — check whether the directory hosts adult or gambling listings alongside your category.
Four checks, five minutes. That filter alone removed about 60% of the directories on my last client's list.
Step 3: claim, fix, then add
Existing listings come first. Claiming and correcting an old listing is faster and more valuable than creating a new one from scratch. Fix the NAP, upload a logo and a couple of photos where the platform permits, and pick the closest matching primary category.
Only after the existing set is clean should you add new listings. Otherwise you're pouring clean water into a dirty bucket.
How many citations do you actually need?
- Start with 15–25 high-quality, category-relevant listings. That covers most local businesses.
- The core set: Google Business Profile, Bing Places, Apple Business Connect, Yelp, Facebook, plus one or two industry-specific directories.
- Beyond 40–50 citations, returns flatten hard. Chasing hundreds is a time sink.
- One authoritative niche directory can outweigh fifteen generic ones.
The part nobody writes about: duplicates and dead ends
Duplicate removal is tedious and, in my experience, the single most underrated citation task. Here's the reality.
Google Business Profile duplicates resolve through the "suggest an edit" flow, or through support if the duplicate has been verified by someone else. Expect a few days to a couple of weeks. Aggregator-fed directories — the ones that pull data from large data providers like Data Axle or Foursquare — can take four to twelve weeks to propagate a correction, because your update has to travel down the supply chain to every site that consumes it.
I once waited nine weeks for a phone number change to appear on a single aggregator-fed listing. During that time, I kept the old number active on a forwarding service. Small cost, no lost calls.
And sometimes you just can't win. Some directories have no claim flow, no contact form, and no response to email. My advice: document the attempt, move on, and make sure the listing at least doesn't contradict your core NAP in a damaging way. Not every battle is worth the hours.
Are citations good for SEO?
Yes — but not in the way most people assume. Citations are not a direct ranking lever the way a backlink is. Their job is corroboration: they help search engines confirm that your business exists at a given location with a given contact method, which supports your eligibility for local pack visibility.
The nuance matters. A hundred citations that all contradict each other are worse than ten that align. And citations won't rescue a weak profile, thin content, or a poor review profile. In my experience, the payoff of a clean citation set is most visible when everything else is already decent and you're trying to break into the top three of the map pack.
Real talk: citation work is a hygiene task, not a growth hack. Treat it as table stakes, not as a lever you pull to leapfrog competitors.
Is ChatGPT a good citation generator?
For research, it's useful. For citation generation, no — and I'd go further: it's actively risky if you deploy it without checking.
Ask an AI model to list local directories for a given industry and it will produce plausible-sounding names, some of which exist, some of which are outdated, and some of which are outright hallucinated. I tested this once: of a list of twenty directories generated for a niche trade, four didn't resolve to a live site and three had changed ownership or shut down registration. The list looked perfect. It wasn't.
Where AI does help: drafting your NAP description text, generating category variants to test, summarising your audit spreadsheet, or suggesting search queries to find directories. Use it as a research assistant, not an authority. Every single listing name it produces needs manual verification before you spend time on it.
Local citation services and whether to buy citations
You'll see packages promising "500 citations for $99". Here's my honest opinion: most of them are worth roughly nothing. They submit your NAP to a network of low-traffic directories that nobody visits, some of which are clearly built to sell links. In the best case, you get a file of URLs nobody will ever click. In the worst case, you inherit spam associations that take months to undo.
| Approach | Cost | Speed | Real value |
|---|---|---|---|
| Manual DIY (audit + core listings) | Free, time-heavy | Slow, 1–2 weeks | High |
| Local listing management tools (paid) | €20–€60/month | Moderate | Useful for ongoing monitoring |
| Managed citation service | €150–€500 one-off | Fast on paper | Mixed — depends heavily on the vendor |
| "Buy 500 citations" packages | €50–€150 | Immediate | Near zero, occasionally negative |
The mid-tier managed services make sense if you have multiple locations and no internal time. Just ask for the exact list of directories before you pay, and check whether they can show you a before/after audit of your NAP consistency. Any vendor that refuses to name the directories is selling you a mystery box.
What will local SEO look like in 2026?
Short answer: more entity-based, less directory-dependent. Three things I'm watching.
First, AI-generated answers in search results increasingly pull factual business data directly from profiles and structured data — Google Business Profile, schema markup, and consistently maintained listings. That means the value of a clean structured data layer is rising, not falling. If your NAP is messy, the AI answer is messy, and the user gets a wrong phone number.
Second, generic directories are losing ground to niche and platform-native ones. Review platforms, vertical marketplaces, and social profiles carry more weight than they did five years ago. The old "submit to 200 directories" advice is already obsolete.
Third, expect more emphasis on the "consistency audit" as an ongoing service. Citations aren't a one-time project anymore. Data decays — a phone number changes, a suite number changes, an aggregator corrupts a field. The businesses that win are the ones that check every six to twelve months.
If you build the audit into your routine now, none of this will surprise you. If you don't, you'll be the plumber who lost 38% of his profile views to an apostrophe.
