Local rank tracking by city: how to measure your local ranking

In brief: measuring a local ranking by city means taking multiple measurement points with identical settings (keyword, surface, device, date), and never relying on a single search or a global average that would mask differences between areas.

A restaurant that ranks well in a city centre can vanish from the local pack (the block of geolocated results displayed at the top of Google Maps and certain Google searches) just a few streets away. This is not an anomaly: it is simply how local ranking works. For a brand present in several cities, or for a business covering a large urban area, a single search therefore tells only a tiny part of the story.

Prerequisites before tracking your local ranking by city

Before running any readings, a few decisions need to be made once and for all. They determine the reliability of everything that follows.

• A verified, active Google Business Profile listing for each business tracked: without verification, no performance data is available.

• A stable list of cities or areas to compare, paired with keywords with local intent; this scope must remain identical from one reading to the next.

• The choice of surface to track: Google Search, Google Maps, or both, as these two products display distinct local results.

• A reference visibility threshold set in advance, such as the top 3 or top 10, and kept for the whole duration of tracking.

• A measurement method chosen between manual point-by-point readings and a dedicated grid tool, with a budget set aside if the chosen tool is paid.

• Access to Search Console and the Google Business Profile Performance report, so that you can later cross-reference the observed rank with real interaction data.

What a local ranking really measures

The local ranking refers to a business's position in the results Google displays for searches for shops or places near the searcher. This ranking is not a single, universal value for a city: according to the Google Business Profile Help Centre, these local results are contextualised by the user's proximity or by the location known to Google. Two people in different parts of the same city may therefore see a different order for the same search.

To build this order, Google combines three official factors: the relevance of the listing to the search, the distance between the business and the searcher (or the simulated point), and the prominence, that is, the business's reputation, including reviews. This combination is documented by Google in its dedicated local ranking Help Centre.

A direct consequence for tracking: since distance affects the result, a single search from a single address represents only one point of presence. On its own, it cannot support conclusions about the ranking of an entire city. Measuring several points is therefore a mechanical necessity, not a convenience.

Good to know

A grid of points is a set of geographic positions spread across an area, from which the same search is simulated to observe how the ranking varies depending on where you are looking from.

Steps 1 to 3: scoping and setting up measurement by city

The full protocol takes seven steps. The first three are for scoping the measurement; the next four, detailed in the following section, are for carrying it out and interpreting it.

1. Select the local-intent queries to track for each business.

2. Separate the cities or areas and choose, for each, a single point or a grid of several points.

3. Fix a constant set of parameters (exact keyword, Search or Maps surface, device, date) for each series of measurements.

4. Record the results on Google Search and/or Google Maps using the same settings for each point.

5. Systematically record the rank obtained, the visible competitors and the date of the reading.

6. Compare the results with a reference period measured using the same method.

7. Compare the observed changes with Business Profile report data and organic traffic.

Step 1: choose the local queries to track

This step involves drawing up the list of searches that reflect genuine local intent, the ones a customer types when looking for a nearby business rather than general information. The number of queries should remain limited and stable over time: adding or removing a keyword mid-tracking breaks comparison from one period to the next.

Step 2: divide the city into a single point or a grid

For a hyperlocal business or a compact city, a single point, such as the business's premises, may be enough. For a sprawling city or a wide catchment area, a grid is necessary: you need to choose a centre, an extent, a grid size and a spacing between points, then exclude any points that fall outside the area actually targeted, such as a neighbourhood with no customer base.

Step 3: lock the measurement parameters

One reading is comparable with another only if a set of parameters stays identical: the exact keyword, the surface observed (Search or Maps), the device used and the date of the reading. This is what the Google Search help reminds us about how results vary by location and device. Each reading must also be tied to the exact business being tracked, and the frequency of scans should be planned in advance rather than changed from one reading to the next.

Steps 4 to 7: running the protocol and reading the results

Once the scoping is done, execution becomes mechanical. What matters here is discipline: applying the same settings to every point, without exception.

Step 4: run the reading on Search and/or Maps

Each point in the grid, or the single point chosen, is the subject of a simulated search on the chosen surface or surfaces, with exactly the same settings as in the previous reading.

Step 5: record rank, competitors and date

For each point, you need to note the rank obtained, the list of competitors visible in the results at that precise point, and the exact date of the reading. These three pieces of information make it possible to reconstruct, later on, what changed and why.

Step 6: compare with the reference period

Comparison only makes sense if the earlier period was measured using the same method: the same points, the same keywords, the same surface, the same device. Comparing a current reading with an older one taken differently amounts to comparing two measurements that are not about the same thing.

Step 7: cross-reference with Business Profile and organic traffic

The final step is to compare changes in rank with the real data generated by the listing: direction requests, call clicks, website clicks and, depending on the business's profile, messages, bookings or menu clicks. A rank that improves with no effect on these actions deserves to be questioned, just like a drop in rank that comes with no fall in real activity.

Three indicators sum up a local ranking by city and by keyword: the rank observed at each point, the share of points where the business appears within the chosen threshold (top 3 or top 10, for example), and the list of competitors present. These three indicators are complementary: rank alone says nothing about geographic coverage, coverage alone says nothing about competitive pressure, and competitors alone say nothing about progress over time.

Key takeaway

A grid is read as a coverage map and how it evolves over time, never as a single average: the rank at each point, the share of points within the chosen threshold and the competitors present together make up the correct reading of a local ranking.

Interpreting ranking differences between two cities

A business that is highly visible in one city and much less so in a neighbouring one is not necessarily a sign of a measurement error. Since ranking combines relevance, distance and prominence, the comparison with local competitors naturally changes from one city to another. A brand may have few strong competitors in one conurbation and face well-established players a few dozen kilometres away.

Before concluding that there has been a real drop, you must first check that the query, measurement points, surface observed, device and period compared are strictly identical. A simple change of device between two readings, or a grid point moved slightly, is enough to produce a difference that has nothing to do with a loss of visibility.

Once this check is done, you still need to examine the competitors present in the results over the period concerned: a newcomer that is very active with reviews, or the disappearance of a long-established competitor, often explains a difference better than a penalty hypothesis.

Point to watch

Before attributing a change to an SEO action, check the following in order: the exact query, the points or grid used, the surface (Search or Maps), the device, the comparison period, and then the competitors present. A single differing parameter invalidates the comparison.

What Search Console, Google Business Profile and Google updates do (and do not) contribute to tracking

Two free Google tools provide useful data for this kind of tracking, but neither replaces a point-by-point reading. Understanding exactly what each one covers helps you avoid misinterpretation.

The limits of Search Console for city-level tracking

In the Search Console Performance report, the position shown is an average position across impressions of the topmost result for the property or page, according to Google's Search Console Help. It is not a fixed local ranking for each city: two users in two different cities may have seen the site at different positions, and Search Console only reports an overall average.

The report lets you segment data by query, page, country, device and date, but it offers no "city" dimension for positions, as Google's help documentation on performance reports explains. It is therefore impossible to isolate a city's ranking directly from this tool alone.

What Google Business Profile does (and does not) reveal

The Google Business Profile Performance report provides the search terms that led to the listing appearing, as well as user views and actions (calls, directions, website clicks) over a chosen period, according to the Google Business Profile Help Centre. This data measures discovery of and interaction with the listing. It is a valuable signal, but it is no substitute for a point-by-point reading or a grid when it comes to knowing where the business actually ranks in a given city.

Tracking frequency and checking Google updates

For routine grid monitoring, a weekly cadence is suitable in most cases, with additional scans before and after a major change to the listing or website, or when an unusual fluctuation occurs. Before attributing a broad change to a local action, it is worth consulting the Google Search Status Dashboard, which publishes incidents and ranking updates that may explain a general movement rather than an effect specific to the business.

Another development worth knowing about: Google now allows you to link Google Business Profile to Google Analytics, which brings local interactions (calls, directions, messages, bookings) into GA4 and makes it easier to connect visibility with conversions. This way of measuring ranking by city is part of a comprehensive approach to local SEO on Google, which also covers the business listing, reviews and the content of local pages.

Common mistakes to avoid when tracking local position

Most tracking errors do not come from using the wrong tool, but from a protocol that is not maintained over time.

• Mistaking a single point for the whole city: a reading taken in the city centre, however flattering, says nothing about the outlying neighbourhoods.

• Changing a parameter between two readings being compared: a changed keyword, device, surface or point invalidates the comparison.

• Treating the average position in Search Console as a fixed local ranking for each city, when it aggregates impressions without any geographic distinction.

• Systematically attributing a drop to an SEO action without having checked the Google Search Status Dashboard or the competitive context.

• Ignoring competitors that have appeared or disappeared of the results over the comparison period, when they often explain the difference observed.

• Failing to specify whether tracking covers Search, Maps, or both, which makes any later data difficult to interpret.

Practical tips for reliable long-term tracking

A well-designed protocol loses its value if it is not documented and followed over time. A few simple habits prevent this drift.

• Document the protocol (queries, points, visibility threshold, frequency) in a shared file that anyone taking over the tracking can access.

• Keep the same visibility threshold (top 3 or top 10, for example) throughout the tracking period, without tightening or widening it along the way.

• Add one-off scans around major changes, in addition to the regular cadence, to isolate the effect of a specific change.

• Always compare the observed ranking with the real actions generated by the listing (calls, directions, clicks), rather than stopping at the displayed ranking alone.

A local ranking that improves without ever translating into calls or directions deserves as much attention as a drop in position: ranking is only one indicator among others, never an end in itself.