Why Your Map Pack Ranking Changes Street by Street in London

Ask a London business owner where they rank for their main service and you will get a position. Number two, maybe. They searched it from their desk this morning. What they have measured is the view from one chair.

This is the most useful idea in local SEO, and it comes straight from the way Google describes its own system. London makes it vivid, because the city is so dense and so varied that the pack can change between two stops on the same Tube line. Here is what Google says, and what I take from it.

What are the three factors Google uses for local ranking?

Google says local results are mainly based on relevance, distance and prominence, and that together they find the best match for a search. Relevance is how well a profile matches what someone is searching for. Distance is how far the business is from the person searching. Prominence is how well known the business is, informed by things such as links and reviews.

Google also says there is no way to request or pay for a better local ranking. That sentence deserves to sit on every local SEO proposal, because it cuts through a lot of selling. Nobody can buy you the pack. They can only help you improve the inputs.

Why does my ranking change depending on where I search from?

Because distance is measured from the searcher, so every location produces a different set of three results. Google uses the searcher's location, or what it knows about it. Move the searcher and you move the distance for every competitor at once.

So the question 'what is my ranking?' has no answer without a second clause: ranking where, for whom, from which street? A rank tracker that checks one point gives you one observation of a surface that has thousands. I explained the same idea for hiring a consultant on the consultant page, and it is the reason I lead with a grid scan rather than a position report.

What is a grid scan and what does it show?

A grid scan samples the Map Pack at a lattice of fixed points across an area and records which businesses appear at each one. The output is a map of coloured markers: green where you hold a top-three position, amber where you are close, red where you are lost, grey where you do not appear at all.

The summary number that matters is share of local voice: the percentage of sampled points where you appear in the top three. It answers a different question from average rank. Average rank tells you how high you place where you place at all. Share of local voice tells you how much of the area you actually cover.

The Guard Up Pest Control case study shows why the distinction is not academic. On one keyword, share of local voice moved from 6.1% to 65.3% while average rank barely shifted. The average hid almost the whole result. The grid is third-party tracker data, labelled as such on that page, not a Google report.

What does this look like in a dense city like London?

In dense areas the pack is decided over very short distances, because many businesses in the same category sit within a few streets of each other. In London that compresses the territory you can realistically own. I would expect a strong profile to hold a tight cluster and thin out quickly beyond it, and I would expect the edge of that cluster to follow transport and street layout more than a neat circle.

• Commuter areas versus residential areas. A search from a City office on a weekday morning and one from a residential street on a Saturday come from different people in different places. Which of those you want matters more than most businesses realise.
• Rivers, rail lines and major roads. A straight-line distance of a few hundred metres can be a very long journey. Google's distance measure and a customer's real travel time are not the same thing, so check where your pins stop.
• Shared postcodes. London postcodes bundle many streets. A profile placed at the postcode centroid rather than its real door can sit in a different part of the grid from where its customers are.

Can I rank across all of London from one profile?

Almost never, and the answer is structural rather than a matter of effort. Distance is built into the ranking, so a single profile with a single address cannot be the nearest result across a sprawling city. Searching 'rank number one in London' as a goal assumes a single pack that does not exist.

Wider coverage has three honest routes. Earn enough prominence that Google shows you further from home than proximity alone would allow. Build location pages that give relevance for specific areas. Or, where you genuinely operate from more than one place, run a separate profile for each. I deal with the last one in my multi-location post. What does not work is faking addresses, which breaks Google's guidelines and which I explain in the serviced office post.

What does a service-area business do differently?

A service-area business hides its address, so ranking leans more heavily on prominence and relevance because there is no fixed point to be near. Google says a business that travels to customers should hide its address and define a service area, and that the area should not stretch beyond roughly two hours of driving.

For a London plumber, locksmith or removals firm that is the right setup, and the grid will often look patchier than a shopfront's. Do not read that as failure. It is the cost of operating without a storefront. The levers you pull are reviews, a service page per area you genuinely serve, and links from local sources.

How do I read a grid scan properly?

Read the edges first, then the centre, then the gaps. The centre almost always looks good, because that is where your pin is. The information is at the boundary, where green turns amber and then red, and in any pocket of red or grey inside an area you thought you owned.

• Check the keyword. A grid is the answer to one search. A scan for 'dentist' and a scan for 'emergency dentist' can look completely different, and the second is often the one that pays.
• Check the radius and spacing. A wide, sparse grid shows reach, and a tight, dense one shows the real street-level pattern. Neither is more correct. They answer different questions.
• Treat grey markers carefully. A point where you do not appear in the top results is not necessarily a point where you rank badly. It may simply be beyond what the tool samples.
• Compare like with like. Repeat the scan with the same keyword, grid and settings, at a similar time, or the comparison tells you about the settings.
• Remember the tool is not Google. Grid tools sample the pack in their own way and are a third-party estimate, which is why I label them as such wherever I show one.

Then ask the commercial question. Does the green cover where your customers actually live and work? A beautiful grid over a park, a river or an industrial estate is worth less than a patchy one over the streets that book jobs.

What should I do with this information?

1. Stop reporting a single rank. Ask for the grid, and the share of local voice, for the keyword that pays the bills.
2. Decide your real territory. Where do your best customers actually come from? Aim the profile and content at that, not at 'London'.
3. Fix relevance before chasing prominence. The primary category decides the competitor set you are ranked against. I explain why in the primary category post.
4. Keep a review flow running. Prominence is the factor you can still move once relevance is right. Recency matters more than volume.
5. Rescan on a schedule. The pattern drifts as competitors move, so one scan is a baseline, not a result.

The reason I like this topic is that it is honest. It tells a business exactly what is controllable, what is not, and why a promise of 'number one in London' should make you walk away.