How to fix inconsistent business listings

I have audited somewhere north of 200 local business profiles in the past decade, and I can tell you the pattern is almost always the same: a tired owner shows me a spreadsheet they paid someone to build, half the entries are wrong, the phone number on Yelp still routes to the old office, and Apple Maps thinks they closed in 2019. The owner asks me to “just push the right data everywhere”. If only it worked that way.

This article gives you a framework I have been using with mid-market clients for years, called SCAR. It is not a tool, not a piece of software, and not something you can outsource for $49 a month. It is a sequence. If you follow the sequence, you stop playing whack-a-mole. If you skip steps, you will be back here in six months wondering why a Foursquare listing from 2014 keeps overwriting your hours.

Why most NAP cleanups fail

Before I introduce the framework, I want to be specific about why the usual approaches do not work. Most cleanup projects fail for three reasons that compound on each other.

The whack-a-mole problem

You fix Yelp. Two weeks later, an aggregator pushes the old address back into 14 downstream sites you have never heard of. You fix those. Then a customer “claims” your Google profile suggesting an edit because they remember the old suite number. Then Bing pulls the wrong hours from a Facebook page that an ex-employee created in 2017.

This is what happens when you fix listings in isolation. Each citation is connected to other citations, and many of them are downstream of data sources you do not control. Until you understand the plumbing, you are just mopping water while the pipe is still leaking.

Limits of bulk submission tools

I like Yext. I like BrightLocal. I like Whitespark. I use all three depending on the client. But I want to be blunt about what bulk tools do and do not do.

Bulk submission pushes data to a network of partner sites. It does not fix duplicates. It does not remove rogue listings on sites outside the network. It does not arbitrate when Google’s own knowledge graph disagrees with what you submitted. And the moment you stop paying for Yext, many of the listings it managed will revert (this caught a client of mine off-guard last year; they cancelled, and within four months 38% of their managed listings had drifted back to old data).

What aggregators actually do (and don’t)

The classic US aggregators (Data Axle, Localeze/Neustar, and Foursquare/Factual) feed hundreds of downstream directories. In the UK and Europe, the picture is messier; Google Business Profile, Bing Places, Apple Business Connect, and the major review sites do most of the heavy lifting, with sector-specific players filling the gaps.

What aggregators do well: seed new sites with your data. What they do badly: clean up the historical mess they already propagated. Submitting corrected data to an aggregator does not retroactively scrub the bad data they sent out three years ago. That is on you to chase.

Introducing the SCAR method

SCAR is the framework I developed after watching too many cleanup projects collapse under their own weight. The name is deliberately not pretty, because what we are dealing with is not pretty: you are healing over old damage, and the result is rarely invisible. It is just stable.

kanban
  Todo
    [Source phase]@{ priority: 'High' }
    [Quarterly mini-audit]@{ priority: 'Medium' }
  In Progress
    [Citation mapping]@{ assigned: 'consultant' }
    [Aggregator resubmission]@{ assigned: 'consultant' }
  Done
    [Lock truth record]@{ ticket: 'SCAR-01' }
    [GBP verified]@{ ticket: 'SCAR-02' }

Source, citation, audit, reconciliation

The four phases:

You can see why people skip to phase four. It is the only phase that feels like “doing something”. The first three feel like paperwork. The first three are also what separates a cleanup that holds for years from a cleanup that falls apart by Q3.

Origin of the framework

I started using a rough version of this in 2016 with a regional veterinary chain that had grown by acquiring single-location practices. Each clinic came with its own decade of listing baggage. Trying to fix them all at once was chaos. Working clinic by clinic, in this fixed order, took us nine months but the results held for years afterwards. I named it SCAR in 2019 because I needed a shorthand to teach junior consultants without saying “the thing we do, you know, the thing” in every meeting.

When SCAR outperforms manual fixes

SCAR is overkill if you have one location, fewer than 15 citations, and no history of address changes. For that, a careful afternoon with a spreadsheet will do.

SCAR earns its weight when any of these are true: multiple locations, a recent move, a rebrand, an acquisition, a merger, a name change with prior trading names still live, or a suspicion that competitors or former staff have created listings you do not control. In those cases, the sequence saves you from fixing the same listing twice (or fixing it before you have decided what “correct” actually means).

Source: locking down your truth record

You cannot fix inconsistency if you have not decided what consistent looks like. This sounds obvious. It is the step skipped most often.

Choosing a single authoritative listing

Pick one listing to be the canonical version. In almost every case I have worked on, this is the Google Business Profile, because Google’s data tends to override or influence what users see across the wider web (including Apple’s results in some regions, via partnerships). If you have a verified Google profile with a long history at the correct address, that is your anchor.

If your Google profile is itself a mess, then your truth record lives in a document you control, not on any platform. I use a simple Google Doc titled “NAP Truth Record” with a version number and a last-modified date. Boring works.

Formatting NAP for machine parsing

This is where most owners trip up. They think “consistent” means “looks the same to humans”. It needs to look the same to crawlers, which are fussier than humans about punctuation and abbreviation.

The “best” column reflects how I write it in the truth record. The “acceptable” column reflects what some platforms will force on you (they reformat your input). You learn to live with this. What matters is that your name does not vary in actual content (no “Inc.” here, no “Inc” there, no “& Co” creeping in from your accountant’s filing).

Handling suite numbers and DBAs

Suite numbers are a quiet nightmare. Google’s address parser handles “Suite 200” differently to “Ste 200” differently to “#200”. Pick one and stick with it. If you are in a shared building, your suite number is the only thing distinguishing you from neighbours; if it disappears from a listing, you have effectively merged your business with someone else’s in the eyes of the crawler.

DBAs (doing business as) are worse. If your legal entity is “Northwood Holdings Ltd” but you trade as “Northwood Dental”, the listings should use the trading name. Your truth record should explicitly note this with a one-line rationale. I have seen multiple cleanups derailed because a well-meaning bookkeeper updated Companies House and then a Yext sync pushed the legal name into 80 directories.

Citation: mapping the listing ecosystem

You cannot fix what you have not found. The Citation phase is reconnaissance. Skip it and you will spend the next year discovering rogue listings one customer complaint at a time.

architecture-beta
  group tier1(cloud)[Tier 1 Sources]
  service gbp(server)[GBP] in tier1
  service agg(database)[Aggregators] in tier1
  service socials(internet)[Socials]
  service longtail(disk)[Long Tail]
  gbp:R --> L:socials
  agg:B --> T:longtail

Tier 1 data aggregators

For US businesses, the tier 1 aggregators are Data Axle, Neustar Localeze, and Foursquare. Submitting clean data to all three propagates to a wide network. For UK and European businesses, the equivalent anchors are Google Business Profile, Bing Places, Apple Business Connect, Facebook, and the relevant national directories (Yell in the UK, PagesJaunes in France, Gelbe Seiten in Germany).

I keep a working list of roughly 60 sites I check in any UK audit. The exact list shifts every year as sites die, merge, or change their data sources. If you want a starting set of credible directories that still drive traffic in 2025, you can browse curated lists like business directory alongside the obvious players; the point is to know which still matter and which are zombie sites taking up audit time.

Industry-specific directories worth fixing

Vertical directories often matter more than horizontal ones for ranking inside that vertical. As Thryv, “many industries have their own targeted listing directories, like Angi for home services”. Examples I see come up repeatedly:

Discovering rogue citations you forgot

The way I find rogue citations is unglamorous. I do exactly what Third Marble Marketing recommends: “Start by Googling your business name, address, and phone number in different combinations.” Then I do it again with old phone numbers, old addresses, old trading names, and common misspellings of all of the above.

I also run reverse phone lookups on every number the business has ever used. The number you discontinued in 2018 is almost certainly still listed somewhere, attached to a now-defunct version of your business name. Those listings are not just wasted citations; they actively confuse Google about which entity is current.

Audit: scoring inconsistency severity

Now you have your truth record and your citation map. The audit is where you compare reality against truth and decide what is worth fixing first. Not every inconsistency is equal, and treating them as if they are will burn your budget.

erDiagram
  BUSINESS ||--o{ LISTING : has
  LISTING }|--|| NAP_RECORD : references
  LISTING ||--o{ INCONSISTENCY : contains
  INCONSISTENCY }|--|| SEVERITY : rated_by
  PLATFORM ||--o{ LISTING : hosts
  PLATFORM }|--|{ AGGREGATOR : feeds_from

High-impact vs cosmetic mismatches

Here is the severity scale I use:

I want to push back gently on something the sources I researched all glossed over. None of them distinguished between “Inc.” vs no suffix (cosmetic, in my view) and a wrong phone number (high-impact, will literally cost you customers today). Treating these as the same problem is why people give up halfway through a cleanup; they spend three days on a hundred cosmetic fixes and then run out of energy before they touch the five high-impact ones.

The duplicate listing detection workflow

Duplicates are a special kind of audit problem. They are not inconsistencies between platforms; they are inconsistencies within a single platform. Google Business Profile in particular is prone to duplicates when a business has moved, changed names, or had an over-eager customer “add” the business while a legitimate profile already existed.

My workflow:

1. Search Google Maps for the business name plus city. Note every result that could plausibly be yours.
2. Search for the phone number on Google. Any Maps result that returns is associated with that number.
3. Search for the address. Same drill.
4. Cross-reference. If two profiles share any identifier, one of them is a duplicate.

Document each duplicate with screenshots, place IDs (if you can extract them), and notes on which seems most established. You will need this evidence when you request a merge or removal.

Building a discrepancy heatmap

I build a heatmap as a spreadsheet: rows are platforms, columns are NAP fields plus hours and website. Each cell is colour-coded by severity (red, amber, yellow, green). At a glance, you see which platforms need the most work and which fields are most commonly wrong across the ecosystem.

The pattern usually tells a story. If your phone number is wrong on 14 platforms but right on Google, you probably changed phones and only updated Google. If your address is right everywhere except the aggregator-fed sites, your aggregator submission was never updated. The heatmap diagnoses the root cause, not just the symptoms.

Reconciliation: fixing in the right order

This is where most articles on this topic start, and that is why most articles on this topic are useless. By the time you reach reconciliation, the work is already 70% done in your head. What remains is execution discipline.

sequenceDiagram
  participant Owner
  participant GBP as Google Business Profile
  participant Agg as Aggregators
  participant Yelp
  Owner->>GBP: Submit corrected NAP
  GBP-->>Owner: Verification code sent
  Owner->>GBP: Enter code, listing live
  Owner->>Agg: Resubmit clean data
  Agg->>Yelp: Propagate downstream
  Owner->>Yelp: Manual claim and correct
  Yelp-->>Owner: Edit approved

Why sequence matters more than speed

You fix Google first. Always. Google is the most heavily crawled, most heavily referenced citation, and many other platforms either pull from Google directly or use Google as a validation signal. If you fix Yelp before Google, and Google still has the wrong data, Yelp’s confidence in your correction is lower and it is more likely to flag your edit for review.

After Google: the major social platforms (Facebook, LinkedIn), then Apple Business Connect and Bing Places, then the aggregators, then the long tail. Within the long tail, prioritise by traffic (use tools like Ahrefs or Semrush to check domain authority and estimated traffic on each directory before spending time on it).

Claim, correct, suppress, or delete

For each listing, you have four possible actions:

The hardest of these is suppress. Some platforms will not let you fully delete a duplicate; they will let you mark it as closed or merge it into another. Yelp is famously difficult. Yellow Pages variants in different countries each have their own appeals process. Budget time for this; in my experience, suppressing one stubborn duplicate can take 4-6 weeks of intermittent emails.

Verification windows by platform

Each platform has its own verification cadence. Plan around them rather than fighting them:

Walking SCAR through a relocated dental practice

Let me show this in action. The names and a few details are altered, but the numbers are real.

Starting state: 47 listings, 12 variants

Client: a three-dentist practice in a UK city, recently relocated from one part of town to another about 1.5 miles away. The move happened eight months before they engaged me. They had updated their website and Google Business Profile within the first week of the move. They assumed that was enough.

My initial audit found:

• 47 listings across various directories
• 12 distinct NAP variants in active circulation
• 2 duplicate Google profiles (one at the old address, marked open)
• 1 Apple Maps profile still pointing to old address
• 3 health-sector directories with old phone number (the line had been transferred but the practice was paying for forwarding to avoid losing patients)
• 17 sites with correct address but old practice name (they had also rebranded six months prior to the move)

So this was a compound problem: rebrand plus relocation, neither fully propagated. They were paying about £40 per month in unnecessary phone forwarding because they were afraid to cut the cord.

Applying each phase

Source (day 1-2): we wrote a truth record. Name format finalised after a 20-minute argument about whether to include “Dental Practice” in the name (we did, because it differentiated them from a similarly named beauty clinic two streets over). Address format finalised including suite number. Phone number set to the new line, with explicit instruction to retire forwarding by week 8.

Citation (day 3-6): I mapped 47 active citations. Found another 14 via reverse phone lookups on the old number that the initial audit had missed. Total surface area: 61 listings.

Audit (day 7-9): scored each listing. Heatmap revealed that the aggregator-fed sites were the worst offenders (had old name, old address, old phone). The high-authority direct sites (NHS profiles, Healthgrades equivalent, local council directory) were a mixed bag. Identified 4 high-impact, 19 high, 24 medium, 14 cosmetic.

Reconciliation (day 10 onwards): fixed Google first (merged duplicate, killed old-address profile). Then Apple, Bing, Facebook in week 2. Aggregator resubmissions in week 3. Long tail in weeks 4-8. Phone forwarding shut off at week 9 after we confirmed no remaining citations had the old number.

Results after 90 days

At day 90:

• 59 of 61 listings had correct, consistent NAP (two stubborn ones were still in dispute with the platforms)
• Google Business Profile views up 34% vs. The 30-day baseline before cleanup
• Direction requests up 51%
• Phone calls from GBP up 22%
• Local pack appearances for “dentist [neighbourhood]” went from rank 7-8 to rank 2-3 consistently

I will not pretend this was all attributable to NAP cleanup; they also started actively soliciting reviews at the same time, which helped. But the structural visibility gains (direction requests, local pack rank) are hard to explain without the citation work.

Edge cases SCAR doesn’t solve

SCAR is not magic. There are situations where the framework helps but does not fully resolve the problem, and I would rather tell you that now than have you discover it at week 8 of an engagement.

stateDiagram-v2
  [*] --> Unclaimed
  Unclaimed --> Claimed : claim listing
  Claimed --> Corrected : update NAP
  Corrected --> Verified : pass verification
  Verified --> Monitored : enter monitoring cycle
  Monitored --> Corrected : drift detected
  Claimed --> Duplicate : duplicate found
  Duplicate --> Suppressed : merge or suppress
  Suppressed --> Monitored : platform confirms

Service-area businesses without addresses

If you are a plumber, mobile dog groomer, or wedding photographer who travels to clients, you probably do not display a public address. Google supports this as a “service-area business” designation, but many directories do not. Some will refuse to list you without an address; others will list you with your home address visible whether you like it or not.

SCAR still applies, but the Source phase needs an additional decision: what address do you give to directories that require one? Options include a home address (with privacy implications), a virtual office (with possible policy violations on some platforms), or a PO box (often rejected). There is no clean answer. I usually recommend a real coworking address that you can document with a lease or membership; this satisfies most platforms’ “must be a physical location” rules without exposing your home.

Mergers and rebrands mid-cleanup

If your business is in flux during the cleanup, you have a problem. I have had clients announce a merger in the third week of a cleanup project, which essentially invalidates the truth record and forces a restart. If you know a rebrand or merger is coming within 6 months, defer the SCAR cycle until after the change is locked in. Otherwise you fix everything to point at “Smith & Co”, then immediately have to fix everything again to point at “Smith Jones LLP”.

The exception is high-impact fixes (wrong phone, wrong address). Those still need to happen, even mid-rebrand, because patients and customers are losing time today.

When to hire versus DIY

I am going to give you genuinely conflicting advice here, because the honest answer depends on factors only you know.

DIY makes sense when: you have one to three locations, you have time to spend 20-30 hours over six weeks, and you can stomach the platform paperwork. You will save somewhere between £1,500 and £6,000 depending on the scope. The work is not intellectually difficult; it is tedious and requires patience.

Hiring makes sense when: you have more than three locations, you have a history of mergers or rebrands, you have suspended profiles, you have active duplicates that have resisted self-merge, or your time is genuinely worth more than the £100-£200 per hour a competent consultant will charge. A specialist will also know the platform appeals processes from muscle memory, which can shave weeks off the suppress/delete actions.

The middle path I sometimes recommend: hire a consultant for the Source, Citation, and Audit phases (1-2 weeks of expert time), then execute Reconciliation in-house with a clear playbook. This costs maybe £2,000-£4,000 and gives you the strategic clarity plus the cost savings.

If you take one thing from this article, take this: stop reaching for the cleanup tool before you have written down what “correct” means. Open a document right now, title it with your business name and today’s date, and write out your name, address, and phone the way you want them to appear everywhere. That document is the foundation of everything else. The tools, the audits, the reconciliations all build on top of it. Without it, you are just rearranging the mess.