Multi-city SEO: tracking keywords in Bogotá, Mexico City and Madrid
Your Google position is not a number — it is a number per city. How to measure it properly, how much quota it really consumes, and why most tools charge you twice for it.
By Alejandro Verjel · Founder of Rankiamos TL;DR
- ✓ Google personalizes the top 10 by city. The same keyword can sit at position 3 in Bogotá and 14 in Medellín.
- ✓ Country-level ≠ city-level. The
gl=coparameter gives you a capital-skewed average, not what your client sees in Cali. - ✓ Every city consumes one unit of your quota. What does not go up is the price: adding cities never changes your plan.
- ✓ Each device consumes its own unit. Mobile and desktop are measured separately: track both and that is 2 units per keyword and city, not 1.
- ✓ Start with 3 cities, not 8. If you would not act differently based on the data, that city does not deserve quota.
Why your position changes by city
Google stopped having a single set of results a long time ago. When someone searches from a phone in Medellín, the engine knows their approximate location and uses it as a heavy signal — not just for the map pack, but for the organic results below it too. On queries with local intent the effect is dramatic: the top 10 reshuffles, nearby businesses move in, and domains that dominate nationally drop out.
The practical consequence is uncomfortable. If your tool reports "position 4" without telling you where it measured from, that number is an average matching no real user. You might be at 3 in Bogotá and 14 in Cali, and the average keeps you calm while you lose your second market. It is the same problem as Search Console's average position, which we cover in detail in the guide to connecting GSC to a rank tracker: an average hides the volatility you need to see.
❌ Measuring one location
One number per keyword. Convenient, cheap and misleading if you sell in more than one city. You find out there is a problem when that region's revenue drops.
✓ Measuring by real city
One number per keyword and city. You see which market is healthy and which needs local content, budget or Google Business listings.
Where it matters most: professional services, healthcare, real estate, retail with physical stores, and any business competing in the Maps pack. Where it matters least: purely informational content and global SaaS. There is a cheap way to find out which case you are in further down.
Country is not city: the gl=co trap
A good share of tools claiming to offer "local tracking" really just swap the country parameter in the Google query. It is fast, cheap to implement, and useful for separating Colombia from Mexico. But within a country it returns an average result that in practice looks a lot like the capital's, because that is where most of the search volume comes from.
The alternative is simulating the search from a specific city. It costs more to operate — each city is a separate API call — which is why many tools sell it as an add-on. It is also the only way to answer the question that actually matters: what does my client see when they search this from their city?
| Criterion | Country-level | City-level |
|---|---|---|
| What it measures | National average, skewed to the capital | What a user in that city sees |
| Useful for local intent | Barely | Yes |
| Detects differences between locations | No | Yes |
| Quota consumption | 1 unit per keyword | 1 unit per keyword and city |
| In Rankiamos | Included | Included in 19 countries (52 cities), no price change |
How much quota it really uses
This is the part almost nobody explains properly, so here it is straight.
Every keyword × city combination is an independent measurement against the Google API, and therefore consumes one unit of your keyword limit. Tracking "seo agency" across 5 cities is 5 units, not 1. Any tool telling you otherwise is either charging you elsewhere or not actually measuring each city.
Two things worth being clear about:
- 1. The price does not scale with cities. Adding locations never triggers an add-on or pushes you to a higher plan. You spend quota, not budget.
- 2. Each device counts separately. Mobile and desktop consume 1 unit each (“Both” = 2) on Pro; Free measures each keyword on a single device. That is not a differentiator: most tools, SEO Edge included, also bill the device separately.
The typical agency scenario, with numbers
Since 1.1.3, every city and every device consumes its own unit of quota — mobile and desktop no longer share one. 10 keywords across 5 cities on a single device is 10 × 5 = 50 units, exactly the Pro plan's ceiling; the same case on both devices becomes 100 and no longer fits in a single project.
What is still genuinely different from SEO Edge is the price: Pro costs $14.99/month ($119.99/year) flat for up to 50 units, whatever mix of cities, search engines or devices you use within that quota. SEO Edge, on the other hand, bills every keyword × city × device combination as a new line item, so its price climbs with every dimension you add.
One honest caveat: if a single project needs more than 50 units, there is currently no way to expand the Pro quota: you have to split keywords across projects or trim cities. We would rather you know that before subscribing, not after.
If you want the full breakdown against other tools, it is in the SEO Edge comparison with a calculator included, and each plan's limits are on pricing.
How to choose which cities to track
The most common mistake is adding every city where the client operates. Quota evaporates, the dashboard becomes unreadable, and nobody looks at 80% of the rows. An order that works:
1. The 3 that bring in the most revenue
Sort by the client's actual revenue or leads, not by city size. Bogotá seems obvious until you discover 40% of sales come from Barranquilla.
2. Where there is real competition
Tracking a city where you are the only provider teaches you nothing: you will always be first. The data has value where there is a fight.
3. The one that triggers a decision
Before adding a fourth city, ask what you would do differently based on the result. No answer means do not spend the unit.
4. The two-city test
Not sure you need multi-city? Track one representative keyword in two cities for a month. If positions move together, you do not need it and you save the quota. If they diverge, you have your answer.
How Rankiamos does it
19 countries — 17 in LATAM, plus Spain and the US — and 52 cities to choose from.
📍 City-level, not a country proxy
Each city is measured with the Google Ads location code for that city, via DataForSEO. Direct captures, not estimates.
📱 Each device, its own measurement
On Pro you can measure the same keyword on mobile and desktop (the “Both” option); each device consumes its own unit of quota, just like each city.
🗺️ Map pack position
On Pro, beyond organic, you see your position in the map block of Google’s results, per city or country — which on local queries often matters more than the blue result.
🏷️ Tags by city and client
Tag keywords with the city or the client and filter by those tags, so a project with 5 cities stays readable on an iPhone screen.
Free plan: 2 projects, 5 keywords, weekly checks · iPhone · iPad · Mac · Apple Watch · App in Spanish only