Mastering Google Maps Scraping to Identify Closed Businesses

Most growth teams think about closed businesses as a hygiene problem. Remove them from the CRM. Stop emailing them. Update the territory list. Fine. That is table stakes.

The more interesting angle is that closures create second-order demand. When a restaurant closes, the landlord needs a new tenant. A POS provider may lose an account but can target nearby operators expanding into the space. A commercial cleaning company can sell turnover cleans. A signage vendor can watch for rebrands. A local SEO agency can find competitors gaining review velocity after a closure. If you sell into local businesses, closures are not just dead ends. They are market movement.

This matters because most lead generation channels are already leaky. Sitewide B2B visitor-to-lead conversion usually lands in the low single digits, around 2-5%. Demo pages and gated asset landing pages can do better, often 5-15%, but that is still dependent on intent. Cold outbound campaigns to 1,000 well-targeted prospects might produce 30-100 replies, yet only 10-30 may be genuine opportunities. And paid search? Average B2B lead conversion rates often fall around 3-8%, while high-intent branded or vendor-comparison campaigns can reach 10-20%. Broad educational terms can sit below 3% and happily burn cash while doing it.

So the SPENDTHRIFT view is simple: before buying more attention, waste less of the attention you already paid for. Closed-business identification helps in three ways. First, it suppresses dead accounts from outbound and ads. Second, it creates trigger-based lists around nearby openings, competitors, landlords, franchisees, and category churn. Third, it improves market sizing because your TAM stops pretending every pinned location is alive and ready to buy.

Google Maps can surface several signals that imply a business is closed, inactive, or at risk. The obvious labels are Permanently closed and Temporarily closed. Those are the cleanest fields to capture when available. But operators know the world is messier than labels.

A business may show stale hours, no recent reviews, a dead website, disconnected phone number, missing booking links, or duplicate listings where one location is closed and another is active. A restaurant may close for renovations and return under the same name. A medical office may move across town but leave the old listing hanging around. Chains are even more annoying: Google may mark one branch closed while the corporate website still shows it live, or vice versa.

That means you should avoid treating a single Maps status as absolute truth. A better scraping workflow collects a bundle of fields: business name, category, address, place ID if available, phone, website, rating, review count, latest review date, hours, status label, coordinates, category, photos count, and any visible status text. If you can enrich with state business registrations, Yelp, Facebook pages, domain status, or local permit data, your confidence score gets stronger.

In practice, I like a closure confidence model with three tiers. Tier one is directly marked permanently closed. Tier two is likely closed: temporarily closed for more than 90 days, dead phone, inactive website, and no recent reviews. Tier three is needs review: conflicting signals, duplicate locations, or suspicious category changes. This prevents your sales team from doing the classic spreadsheet faceplant where they call a business marked closed and get an annoyed owner saying, no, we are very much alive.

When you look across US metros, closure data rarely spreads evenly. It clusters by neighborhood, category, rent pressure, commute patterns, tourism exposure, and local regulation. The point is not to create a doom map. The point is to understand churn pockets so your sales motion matches reality.

In dense coastal cities like New York, San Francisco, Boston, Los Angeles, and Washington DC, closures often show up in high-rent retail corridors and food service categories. These markets have huge demand, but they also punish weak unit economics. A cafe with average reviews, rising labor costs, and a lease reset can disappear quickly. If you sell to restaurants, hospitality, local services, commercial real estate, payment processors, or insurance brokers, these closures can signal both loss and replacement opportunity.

In Sun Belt metros like Austin, Phoenix, Tampa, Nashville, Charlotte, Dallas, and Atlanta, the story is different. You often see a lot of openings and closures happening at the same time. Population growth attracts new operators, but fast expansion produces sloppy site selection. A suburb can look amazing in a population chart and still underperform if traffic patterns are wrong. For B2B teams, that means closed-business scraping should be paired with new-business detection. The delta between openings and closures is more valuable than either number alone.

In tourism-heavy cities such as Las Vegas, Orlando, Miami, New Orleans, and parts of Southern California, category swings can be sharper. Restaurants, attractions, short-term rental services, tour operators, and event vendors can see rapid status changes when travel demand shifts. Here, a quarterly scrape may be too slow. Monthly monitoring gives a cleaner read, especially if your product sells into seasonal staffing, cleaning, bookings, signage, payments, or local advertising.

In older industrial or mid-market metros like Cleveland, Detroit, St. Louis, Pittsburgh, Milwaukee, and Buffalo, the patterns often vary block by block. Downtown revival zones may have new openings while legacy retail strips decline. Scraping by city boundary is too blunt. You want grids, ZIP codes, census tracts, or radius searches around commercial corridors. Otherwise, you average out the signal and end up with a shrug.

The mistake I see often is treating USA city data as a single leaderboard. Closed restaurants in Manhattan do not mean the same thing as closed salons in suburban Phoenix or closed auto repair shops in Cleveland. The category, neighborhood, and replacement cycle matter. A closed independent gym may become a physical therapy clinic. A closed cafe may become another cafe within 60 days. A closed dental practice may indicate retirement and patient list acquisition opportunities. Your lead strategy should follow the local business lifecycle, not just the closure label.

There are two ways to approach Google Maps scraping. The chaotic way is to search a city, export everything visible, and spend days cleaning duplicate rows. The operator way is to define the market, category, refresh cadence, and confidence rules before collecting anything.

Start with your ICP. If you sell payroll software to restaurants with 10-50 employees, scraping every business in Chicago is wasteful. You need restaurants, bars, cafes, bakeries, caterers, and maybe hospitality groups. If you sell commercial HVAC, you may care about closed gyms, clinics, restaurants, warehouses, and retail spaces because those locations need equipment checks during tenant turnover. Same data source, completely different lens.

Next, define geography. City-level searches are often too vague. Use ZIP codes, neighborhood names, coordinate grids, or radius-based queries around commercial zones. A grid approach is usually cleaner for larger metros because Maps results can cap out or bias toward popular listings. Break Los Angeles, Dallas, or New York into smaller tiles. Store latitude and longitude so you can map churn over time.

Then choose the fields. At minimum, capture name, category, address, phone, website, rating, review count, status, hours, coordinates, and collection date. If your tooling supports stable identifiers, keep place IDs or listing URLs. This is crucial for longitudinal tracking. Otherwise, you will think Joe's Pizza closed, reopened, and duplicated itself when Google merely changed the display name or address formatting.

Set a refresh cadence based on category volatility. Restaurants, salons, fitness studios, and retail should be checked monthly in active markets. Professional services can be quarterly. Industrial and medical categories may need slower but deeper verification. The cadence matters because a single scrape is a snapshot. The business value comes from change detection: newly closed, still closed, reopened, moved, replaced, or duplicated.

Finally, build compliance and politeness into the process. Google has terms governing use of its services, and aggressive scraping can create legal, ethical, and operational problems. Do not hammer endpoints, bypass access controls, or collect personal data you do not need. Consider official APIs, licensed datasets, or providers that handle collection responsibly. If you use a tool like GeoLayer.io, the value is not that it magically removes every compliance question. It is that it can make structured geo-data collection less painful, especially when you need repeatable location workflows instead of one-off exports. Still, your team owns how the data is used.

Closed-business data needs quality checks because local listings are messy. Google may show a business as temporarily closed during renovations. Owners may forget to update hours. A franchise may consolidate listings. A practice may relocate but keep the same phone number. In some categories, especially healthcare and legal, the practitioner moves while the business entity remains active.

A solid QA process has four layers. First, dedupe aggressively using name, phone, address, coordinates, and stable IDs. Second, compare current scrape data with prior collection dates. Third, enrich against external signals such as domain status, business registry records, Yelp activity, social profiles, and recent reviews. Fourth, create a human review queue for high-value accounts before sales outreach.

The human review part sounds old-fashioned, but it saves face. If a target account is worth $20,000 annually, spending two minutes to verify the closure context is rational. If the account is worth $200, automate more and accept a bit of noise. This is where lead gen ROI gets practical. Not all leads deserve the same verification cost.

I also recommend labeling the reason for the closure flag. Instead of a single binary field called closed, use fields like status_label, last_seen_open_date, first_seen_closed_date, closure_confidence, conflicting_signals, and verification_source. This makes the data explainable. Your reps will trust a lead marked permanently closed on Maps, phone disconnected, website offline much more than a vague closed equals yes row.

One more caveat: do not confuse closure detection with intent. A closed competitor nearby may create urgency for surviving businesses, but it does not mean they want your product. The trigger is an opening line, not a contract. Use it to make outreach relevant, not creepy.

Closure data can sound morbid if used badly. Nobody wants an email that says, I saw your neighbor failed, want software? That is not sharp. That is LinkedIn cringe with a CSV attached.

The useful angle is market context. If a restaurant closes nearby, other restaurants may be absorbing demand. If a retail location turns over, vendors serving new tenants have a timely reason to reach property managers, brokers, contractors, and local operators. If several salons close in a ZIP code, surviving salons may care about reputation, booking, staffing, and local ads because customer behavior is shifting.

The best teams turn closure data into account segmentation. For example, create a list of active businesses within a half-mile of newly closed competitors. Sort by rating, review count, category, and website quality. A five-star business with 800 reviews probably does not need the same pitch as a 3.7-star operator with no online booking and a competitor disappearing down the street. One is defending demand. The other may need help capturing it.

You can also use closure trends to improve ad targeting. If paid search CPLs in B2B software and professional services often run $75 to $300+, wasting clicks on dead or irrelevant accounts is painful. Upload suppression lists where appropriate, adjust geo targeting around active commercial corridors, and build landing pages around specific category shifts. A generic local business software page is forgettable. A page for dental practices acquiring patients after nearby office closures is at least trying.

For sales ops, closed-business scraping also helps CRM hygiene. Remove closed accounts from sequences, flag relocated companies for enrichment, and alert account owners when a customer location appears closed. This reduces pointless touches and improves reporting. Pipeline conversion rates look less mysterious when your denominator stops including businesses that no longer exist.

GeoLayer.io is useful when teams need structured location data workflows without building every scraper, parser, and geo-normalization step from scratch. That can matter a lot if your growth team is small and allergic to wasting engineering cycles on brittle internal tools.

The main advantage is operational. You want repeatable city and category coverage, clean exports or API access, coordinates, status fields, and enough structure to plug into a CRM, warehouse, or enrichment process. If you are manually copying Google Maps results into sheets, you are not doing research. You are donating your afternoon to entropy.

That said, I would not position any tool as a complete answer. You still need ICP logic, QA rules, consent-aware outreach practices, and a point of view on which closure signals matter. A data tool can reduce collection waste. It cannot decide your sales strategy. That part is still on you, annoyingly.

The buying decision should come down to ROI. If a competitor costs more, requires heavier setup, gives you generic lists, or lacks verification fields, then a leaner option wins. If you need enterprise-only governance, custom legal review, and deep integrations, a larger vendor may be justified. Spend where the bottleneck is. Do not buy a battleship when you need a bike courier.