What Google calls “local” and how it calculates it
A local query triggers a different display from classic results: a map, three businesses, their ratings, and their opening hours. This block is called the Local Pack, and it takes up the most visible screen space on mobile. For local service queries, it captures most clicks before the first organic result even appears.
The first thing to understand is that the Local Pack and organic results are two separate rankings, calculated by different systems. A perfectly optimized website can rank #1 organically and still be absent from the pack. The reverse is just as common. Working on one is never enough to get the other.
The second is that Google triggers the local display even when the user has not typed a city name. “Plumber” from a phone in Toulon returns the same results as “plumber Toulon” typed from a computer elsewhere. Local intent is inferred from context, which greatly expands the number of queries concerned.
Google publicly documents three factors for local ranking: relevance, distance, and prominence. None is officially weighted, but distance imposes a hard limit that nothing offsets: a business located 30 km away will not appear for a search made downtown, regardless of how good its profile is.
The three Local Pack factors
These three criteria reinforce each other and are worked on separately. Addressing them in the wrong order wastes months.
Relevance: the declared match
Relevance measures the gap between what the user is looking for and what your business claims to do. It is determined first by the primary category of the business profile, which carries more weight than anything else. Choosing “general contractor” instead of “roofer” when you are a roofer is enough to disappear from roofing searches.
Next come secondary categories, the description, listed services, and attributes. Each adds semantic hooks. The rule is to declare what you actually do, without stacking categories to broaden coverage: a profile that claims twelve trades dilutes its signal and loses on each one.
Distance: the constraint you cannot bypass
Google calculates the distance between the user and the declared address. This factor explains most situations where a well-known professional ends up behind a lower-rated competitor: the second is simply closer. No optimization moves an address.
What can be worked on, however, is the service area. A business that operates at customers’ premises can declare a service perimeter instead of a visible address, which changes how Google presents it. And above all, a multi-location business benefits from having one profile per real location, each with its own landing page on the website.
Prominence: the part you build
Prominence aggregates what the web says about your business: the volume and quality of reviews, mentions of your name and address on other sites, inbound links, and overall consistency. It is the only one of the three factors that improves continuously with the work you put in.
Its progression is not linear. The first reviews count a lot, the next ones less and less, until a credibility threshold is reached. Going from three to thirty reviews changes everything; going from three hundred to three hundred and thirty changes almost nothing. Marginal effort should therefore shift to mentions and links once that threshold is crossed.
Local queries are not targeted like the others
A local search combines three phrasings that tools treat separately even though Google often groups them. The explicit query pairs the service and the city. The implicit query contains only the service, with local intent inferred from location. The proximity query adds “near me” or “open now”.
The volumes shown by tools systematically underestimate the last two, because they are spread across thousands of geo-localized variants. Relying only on the explicit query volume leads you to massively undervalue the potential of a local market. In small areas, like those we serve from our web agency in Andorra, the gap between the reported volume and the traffic actually captured commonly reaches a factor of three.
NAP consistency: the foundation almost nobody maintains
NAP refers to the name–address–phone trio. The principle is trivial to state and hard to maintain: these three pieces of information must be strictly identical everywhere they appear on the web.
Why a comma is enough to create a problem
Google matches mentions of your business found on different sites to consolidate its knowledge of the entity. When variants diverge, matching becomes uncertain and trust drops. “SARL Dupont Couverture” on a directory, “Dupont Couverture” on the profile, “Couverture Dupont” on the website: three possible entities instead of one.
The same logic applies to the address, where “avenue” and “av.” are not equivalent, and to the phone number, where international and national formats create two different strings. The chosen form matters little; consistency is everything.
Inventory existing mentions
For a business established for several years, mentions often number in the dozens: professional directories, federations, former partners, quote-request sites, social networks, review platforms. Many were created by third parties and contain outdated information—sometimes an old number or an old address.
The inventory is done by searching the exact name, phone number, and address in quotation marks on Google. Corrections are then made one by one, starting with the most visible sources. It is thankless work with no immediate effect, and yet it is the work that unlocks situations where nothing else works.
Geo-targeted pages: the condition under which they work
Creating one page per city is everyone’s first idea—and it is also the most common source of failure in local SEO.
The difference between a local page and a duplicate page
A page that repeats the same text while changing the town name is near-duplicate content. Google detects it easily, indexes only part of it, and the presence of dozens of weak pages degrades the perceived quality of the entire domain. The cure is worse than the disease.
A local page that works actually talks about the place. It mentions neighborhoods, the local economic fabric, constraints specific to the area, references to projects or clients in that sector, and real response times. It answers the question a resident truly asks—not “do you exist?” but “do you serve my area and do you understand my constraints?”
How many pages, and which ones
Five strong local pages are better than forty empty ones. The selection criterion is the area’s real commercial potential, not the search volume shown by a tool—often very approximate for small towns.
Each local page must be reachable within two clicks from the homepage, receive links from the corresponding service page, and point to the relevant business profile. This geographic silo structure is the same as the one described in our article on the semantic cocoon, applied to a territorial axis.
Customer reviews: an asset you build
Reviews matter on two levels at once: ranking in the pack and click-through rate once displayed. A business in third position with eighty reviews often captures more leads than the first with six.
Getting reviews without buying them
The only sustainable method is to ask—systematically, at the right time. The right time is when the customer expresses satisfaction, not three weeks later via an automated email. A short link sent by message immediately after the job gets a response rate that is incomparable with a delayed follow-up.
Bought reviews are detectable: profiles with no history, uniform writing, clustered arrival. The penalty is not limited to removing fake reviews; it affects the trust granted to the business. The risk is disproportionate to the gain.
Replying—especially to negative ones
Replying to a negative review is not about convincing its author, who will not come back. It is about showing the next readers how the business reacts when something goes wrong. A factual, brief, non-aggressive reply that offers a concrete next step turns a bad review into a demonstration of professionalism.
A perfect rating is also less credible than a very high rating with a few addressed reservations. Buyers are wary of an absolute five out of five across several hundred reviews.
Photos and posts: two neglected signals
Photos on a profile are not just decoration. Their freshness, their number, and the fact that they are taken on site rather than pulled from a stock library factor into the assessment of the business’s real activity. A profile whose last photo is four years old sends an abandonment signal.
Posts, often neglected because they disappear after a few days, maintain the freshness signal and take up space on the displayed profile. A short weekly post announcing a recent job or availability is enough. It is a few minutes of work that most competitors do not do.
LocalBusiness markup and the website itself
The business profile does not remove the need to work on the website. The two reinforce each other, and Google systematically cross-checks one against the other.
Declaring the entity in Schema.org format
JSON-LD markup of type LocalBusiness, or one of its more specific subtypes such as Restaurant, Dentist, or HomeAndConstructionBusiness, explicitly declares the name, address, phone number, opening hours, service area, and geographic coordinates. This information must match the profile down to the character.
A specific subtype is better than the generic type when it exists, because it conveys additional semantic information. The areaServed property also makes it possible to declare a service area that goes beyond the headquarters address alone.
The landing page matters as much as the profile
The business profile link should not point to the homepage when there is a page dedicated to the location or the requested service. A landing page aligned with the query improves visitor behavior—and that behavior is measurable.
This page must load fast, especially on mobile, since most local searches happen there. Core Web Vitals thresholds apply here as elsewhere, with a stronger effect: a user on the move, on a degraded network, abandons faster than a user sitting at a desk.
Local citations: what still matters in 2026
Citations are mentions of your business on third-party sites, with or without a link. Mass submission to hundreds of directories was common ten years ago; it is not anymore: a directory with no real audience has zero value, and mass submission creates an artificial profile.
What remains useful falls into three families. Top-tier general directories that Google actually cross-checks. Professional bodies and trade federations, which also provide industry credibility. Real local players—regional press, tourist offices, merchants’ associations—whose mention matters both for SEO and for direct brand awareness.
A dozen solid, accurate citations are better than two hundred automated listings. The selection criterion is simple: does the site receive human visitors? If the answer is no, the listing adds nothing and creates one more source to keep consistent.
The method for tackling an area
The order of operations determines how quickly results arrive. Here is the one we apply in a local SEO service.
- Merge duplicate profiles and verify ownership of the main profile. As long as two profiles coexist, nothing else will produce a measurable effect.
- Correct the primary category. This is the single setting that creates the biggest visibility gap—and it takes five minutes.
- Standardize the NAP across all mentions found. Long, thankless, decisive.
- Complete the profile 100%. Services, attributes, special hours, recent and geo-tagged photos.
- Build or rebuild local pages with truly specific content, and interlink them from the service page.
- Implement LocalBusiness markup consistent with the profile, and validate it.
- Start collecting reviews by making the request systematic in the sales process.
The first two steps sometimes have an effect within two weeks. The next ones are measured over three to six months. A prior SEO audit makes it possible to know which ones truly apply to your situation, rather than doing everything blindly.
What local SEO will not fix
An honest service means stating what is beyond optimization. Three situations recur, and in each, spending on local SEO yields little.
The first is geographic distance from the target market. A business located on the outskirts that wants to capture downtown runs into the distance factor, which nothing offsets. The answer is not technical; it is real estate: open a second location or accept working a different area.
The second is a trade with no local search demand. Some services are chosen for expertise regardless of proximity, and users never type a city name. Investing in the Local Pack then amounts to optimizing a door nobody uses.
The third is a damaged reputation. An average below three out of five with a meaningful volume of reviews is not fixed by gaining positions: additional visibility simply amplifies the problem. The work is operational before it is digital, and telling the client is part of the job. This candor about limits follows the same principle we apply to our website redesign quotes, where we regularly decline projects whose problem is not the website.
The case of businesses without a visible address
Many businesses operate at the customer’s premises and do not have a public-facing location: tradespeople, emergency repair services, in-home providers. Google covers this case with service areas, but it imposes constraints that few professionals know.
A service-area profile hides the address and shows a perimeter instead. This setup reduces the proximity advantage, since Google no longer has a precise point to calculate distance from the user. Ranking then relies more on relevance and prominence, which makes work on reviews and mentions even more decisive.
Declaring an unrealistic perimeter harms the business. An area covering three departments dilutes the signal without improving visibility in the center. It is better to declare the area you truly serve, even if that means opening a second location later if the business justifies it. The same trade-off arises when choosing between a single site and multiple sites—a question we address on the website creation side.
The case of multi-location networks
A business with multiple locations raises an architecture issue that goes beyond the business profile alone. Each real location deserves its own profile, and each profile deserves its own landing page on the website. The classic mistake is to point all six profiles to the same homepage.
Each location page must carry its own markup, its own address, its own opening hours, and ideally content that differentiates it: the on-site team, specific services, access details. Without that, the six pages cannibalize each other and Google favors one at the expense of the others.
A national or regional page usefully sits above the whole set and captures queries with no geographic intent. It lists the locations, links to each one, and serves as the pillar page for the geographic silo. This structure works just as well for a retail network as for the trade-based variations we build by industry sector.
The last point is management. One profile per location means review collection per location, review replies per location, special hours kept up to date everywhere. For a network of more than five locations, this workload becomes a function in its own right, and ignoring it produces abandoned profiles that drag the whole set down. The organic SEO services we run for this type of network systematically include this operational component.
Key Takeaways
The Local Pack is driven by three factors, only one of which can be worked on freely. Distance is a constraint, relevance can be adjusted in a few hours on the profile, and prominence is built over months. It is this asymmetry that should govern budget allocation.
Two workstreams produce most of the result: consistent business information across the web, because a fuzzy entity does not rank; and regular collection of authentic reviews, because no other signal improves as reliably with the effort invested.
The rest is honest execution: the right primary category, local pages that truly talk about the place, markup consistent with the profile, and a fast landing page. Nothing spectacular—and that is precisely why most competitors do not do it.
One last note on measurement. Local ranking varies with the user’s exact position, which makes any check done from your own office useless. A business can rank first in the pack 500 meters away and disappear at 3 km. Measurements must therefore be taken on a grid of points distributed across the target area; otherwise, the conclusions drawn are wrong. The projects we track in our case studies are measured this way, not with a single search from the agency.
