How telecom retailers decide where to put phone shops and kiosks
Phone shops follow people, not footfall alone. How operators and dealers find blocks with many residents and few shops, from city malls to rural market towns.
A mobile operator or phone retailer places a new shop where many people live or gather and few shops already serve them. The working measure is residents per phone shop, counted block by block, then checked against what draws people to the spot: a mall, a bus station, a market, a campus. Shops go where that ratio is high and the spot is easy to reach; kiosks go where crowds pass but nobody stays long.
This guide is for retail and channel teams at operators, at their franchise partners and at large dealers. It covers the three formats telecom retail uses, the measures behind each, a worked example, the traps, and how an AI employee keeps the shortlist fresh without a monthly spreadsheet exercise.
Three formats, three different questions
| Format | What it needs | The question to ask |
|---|---|---|
| Full store | Residents nearby, room for service and repairs | Which blocks have the most people for each phone shop already there? |
| Kiosk or pop-up | People passing at set hours | Which malls, stations and campuses have none of my shops close by? |
| Dealer or partner shop | An existing independent shop to sign | Which independent phone and repair shops sit in the gaps between my stores? |
The format decides the measure: residents for stores, places people gather for kiosks, independent shops for dealers.
Mixing these up is the most common reason a new shop disappoints. A kiosk in a quiet residential block starves, and a full store in a transit hall pays for space that queues never fill.
Measuring people per shop, block by block
Population is read on a grid of hexagons of about a tenth of a square kilometre each, so a city of several million becomes thousands of small cells. For each cell, count the residents within a kilometre and the phone shops in the same circle: cell phone stores and, if you treat them as rivals for the same customers, mobile repair shops. Dividing the first by the second (adding one to the shops, so a cell with none does not divide by zero) ranks every block from most underserved to most crowded.
Two refinements make the ranking sharper. First, keep only cells above a minimum number of residents, so empty industrial land with no shops does not rise to the top. Second, look at the age groups: a block with many residents in their twenties buys more handsets and data plans than one of the same size that is mostly retired.
A worked example in Bangkok
The numbers here are made up to show the reasoning. Say an operator has 40 shops across Bangkok and wants two more. The monthly ranking returns these blocks at the top:
| Block | Residents within 1 km | Phone shops within 1 km | People per shop | Draws nearby |
|---|---|---|---|---|
| North of a new rail interchange | 42,000 | 3 | 10,500 | A station, two schools |
| Riverside housing estate | 36,000 | 2 | 12,000 | A market |
| Suburban ring | 51,000 | 12 | 3,900 | A large mall |
A made-up example, not real figures for Bangkok.
The riverside estate has the most people per shop, but its market trades mornings only and the nearest road is slow. The rail interchange has nearly as many people per shop and a station that brings commuters twice a day, so it gets the full store. The suburban ring is already well served by shops, but its mall has none of the operator's own: that is a kiosk, not a store.
Rural reach: the other half of the map
Outside the big cities the method turns around. Instead of hexagons, start from the places a region already gathers at, its markets and bus stations, then count the people living within 5 km of each and the phone shops in the same circle. Towns with thousands of people and no shop at all are where a small store, a van that visits weekly or a signed local dealer reaches customers nobody else does. The rural reach template does exactly this and lists the twenty towns with the fewest shops first.
Traps to avoid
- Counting only your rivals' branded shops. Independent shops sell prepaid cards and handsets too. Count them, and consider signing the good ones as dealers.
- Trusting a straight-line circle across a river or a motorway. For the final few, draw the area people can actually reach in ten minutes on foot or by car.
- Assuming a busy mall means a good store. Busy for whom, and when? People-hours of presence tell you an area is busy, not that its visitors will buy a phone there.
- Shortlisting once a year. New housing fills in months; a monthly run catches it while the rent is still reasonable.
How an AI employee runs it every month
In xMap Flows the store-placement method becomes an AI employee whose shortlist is checked by a person before it reaches the retail team:
- A Schedule starts it every month, and an Area of interest names the city.
- Demographics brings every hexagon; Sort and limit keeps the 100 most populated.
- Nearby places, set to each incoming place with counts, counts the phone shops within 1 km of each.
- A Calculated field works out people per shop, and a second sort keeps the top 20.
- xMap LLM scores each block from 0 to 100 for your format, with a reason.
- Approval asks the head of retail to check the batch; once approved, an Email and a map go to the team.
That is the next phone shop template. Its siblings cover the other formats: malls with no store of mine nearby for kiosks, independent shops to recruit as dealers, and where rival operators opened shops to see the competition move. Each lands in the Team inbox as a note from its AI employee, and a reply such as "add Chiang Mai" changes it. See running a team of AI employees from one inbox.
Questions
Should mobile repair shops count as competitors?
Count them if your stores repair phones or sell prepaid plans, because customers use them for the same errands. If you sell only contracts, count only cell phone stores.
How close should two of my own shops be?
Close enough that each serves its own block of people and no closer. Ranking blocks by people per shop, with your own shops counted in, shows where a new one would only take customers from an existing one.
Can I place shops in countries without population figures?
Yes. Start from the places people gather, such as malls, stations, markets and campuses, and rank them by how few phone shops are around them. Residents per shop needs population figures, which some countries do not have yet.
How large an area should the count use?
One kilometre suits dense cities, where people walk to a shop. In suburbs and rural regions use three to five kilometres, since people drive or take a bus to reach one.