Local Rank Tracking: A City Is Not a Location

A city is an area; a tracker samples a point. When you set a tracker to “Manchester” it picks some coordinate — often the city centre — and asks from there. For a query with local intent, results depend on proximity, so that answer describes searchers standing near that coordinate and nobody else.

If your service area is a city, one sample is not a measurement of it. It’s a measurement of one spot in it.

Why proximity dominates

Local result sets are assembled with distance from the searcher as a heavy input. That means the ordering genuinely changes as the searcher moves, and it changes over surprisingly short distances in dense areas. A business ranked first from the city centre can be absent from the pack three miles out, where three closer competitors now sit above it.

None of that is volatility. It’s not noise, it doesn’t average out, and checking more often won’t help. It’s a real property of the query: there is no single correct position, only a position-per-place.

This is the main way local tracking differs from national tracking. Nationally, variation between sample points is mostly nuisance. Locally, the variation is the data.

What a grid gives you

The standard answer is to sample a grid of coordinates across your service area instead of one point, then look at the shape of the result. Tools that offer this usually show it as a heatmap.

What that shape tells you:

• Where the boundary is. The line at which you drop out of the pack is a concrete fact about your competitive position, and it moves when the competitive picture changes.
• Whether you’re strong or just close. A business that ranks well only within a tight radius of its address is ranking on proximity. One that holds positions further out is ranking on something more durable.
• Which direction is weak. Weakness is rarely symmetrical. It’s usually a direction, and the direction usually has a named competitor in it.

Grid density is a trade-off with cost: every point is another SERP fetch. A coarse grid over a wide area beats a dense grid over a small one, because the interesting information is at the edges.

Settings that quietly matter

Address versus coordinate. Some tools take a place name and geocode it, which can land somewhere unhelpful. If the tool accepts coordinates, use them and write them down.

Pack versus organic. Local pack position and organic position are separate series for the same keyword and can move in opposite directions. Track both, and never blend them into one number — the pack has three visible slots, so its scale is nothing like the organic one, for the reasons in position is an ordinal, not a quantity.

Language. In multilingual areas the interface language changes the result set independently of location.

Device. Local intent skews mobile, and the mobile layout gives the pack more of the screen. If you only track desktop for a local business you’re tracking the minority case — mobile and desktop are two rank sets.

Reading movement in local data

Two failure modes are common enough to name.

Comparing across sample points. If last month’s report used the city centre and this month’s used a suburb, the whole comparison is invalid. Local sample points must be pinned harder than national ones, precisely because they matter more.

Treating a boundary shift as a ranking change. If your grid shows the same positions everywhere except two edge cells that flipped, nothing happened. Edge cells sit near a boundary and will flip on their own. Look for movement in the interior before concluding anything.

For anything else that looks like a drop, the general procedure applies — how to diagnose a ranking drop.

What to actually do

1. Pin exact coordinates, not city names, and record them beside the report.
2. Sample a grid if the tool supports it, prioritising area covered over density.
3. Keep pack and organic as separate series for every local keyword.
4. Track mobile as the primary device for local intent.
5. Report the boundary, not a single position — “we hold the pack within about four miles” is a statement a business owner can act on. “We’re number two” is not, because it omits where.