The location study: how to choose the site for a new shop, pharmacy or clinic

1. The problem: a five-year decision made in an afternoon

A pharmacy chain wants to open its sixth outlet. An estate agent offers a ground-floor unit in a new apartment block on a busy street, at a reasonable rent. The director drives past twice, sees plenty of traffic and signs. A year later, sales are half the chain average: the traffic was through-traffic, the pavement on the pharmacy's side of the street is little used, and local residents already shop at a pharmacy next to the bus stop. A location study exists to ask these questions before the signature, not after it.

Any business that depends on customers walking through the door faces the same problem: shops, cafés, clinics, gyms, bank branches, petrol stations, private kindergartens. So do public institutions deciding where to put a service centre, a kindergarten or a health post. Only the measure of success differs: sales for a company, access and coverage for an institution.

2. What a location study is

A location study estimates how many customers and how much revenue a specific site can generate, compared with the alternatives. It answers four questions:

• Demand: how many people live in, work in or pass through the area customers could come from, and how much they spend on the category.
• Competition: who already serves that demand, where, with what offer and how well.
• Access: how easily customers can reach the site on foot, by car or by public transport, and how visible it is.
• Potential: what level of sales is plausible, judging by similar outlets in similar conditions.

A location study is not the same as sizing a market for a whole city or country, which is covered in our article on market sizing and segmentation. Here the question is local, at the level of a neighbourhood or a street, and a few hundred metres can make all the difference.

3. When you need one, and when you do not

A location study is justified when the investment is hard to recover: long leases, expensive fit-outs, large units. It is especially useful for chains expanding into a new city, where management's instinct no longer works, and for choosing between two or three sites available at the same time. A company entering a country for the first time first needs a market entry study; site selection comes afterwards, once the city and the format have been chosen.

A full study is not needed for a temporary unit, for a small, low-investment format, or when one option is clearly better on every criterion. In those cases a structured site visit with a standard assessment form and a basic count will do.

4. The catchment area: where customers come from

The catchment area is the area from which an outlet draws most of its customers. It is usually split into three rings: the primary zone, which supplies most customers and where the outlet has its highest share; the secondary zone; and the tertiary zone of occasional customers. Their size depends on the format: for a neighbourhood grocery the primary zone may be a few minutes' walk; for a shopping centre or a specialist clinic, tens of minutes by car.

Two rules separate an estimated catchment from a real one. First, it is defined by travel time, not straight-line distance. A river, a railway line or a boulevard with no crossings can cut a neighbourhood in two. Second, for existing chains it is built from real customer data, such as addresses from the loyalty programme, delivery records or a short exit question ("where did you come from today?"). The map of real origins usually looks very different from the theoretical circles.

Once the zone is defined, demand is estimated: residents and households, age structure, jobs located in the area, residential projects under construction. In many Moldovan localities official data at neighbourhood level are limited or out of date, so estimates are supplemented with administrative records, an on-the-ground count of residential buildings and a household survey. Working with existing sources is described on our data analysis and desk research page.

5. Fieldwork: counts, competitors, access

Pedestrian and vehicle counts

Counts follow a fixed protocol: the same points, the same 15-minute intervals, on weekdays and at weekends, in the morning, at lunchtime and in the evening, ideally in two seasons. Pedestrians are counted separately on each side of the street and by direction, and for car-based formats only vehicles that can actually stop are counted. A single count on a sunny Friday says nothing about a rainy Tuesday in February.

Competitor mapping

Every competitor in the area is listed: location, format, size, opening hours, price level and observed service quality. For direct competitors, a structured store check shows range, prices and how busy they are, as described in our article on retail audits.

Accessibility and visibility

A standard form assesses each site: how far away it can be seen, whether there is parking, how close the nearest public transport stop is, whether the entrance is accessible to people with reduced mobility and parents with pushchairs, and which traffic generators are nearby (a market, a school, a clinic, a station).

The catchment survey

When demand cannot be estimated from existing data, a short survey of households in the area, face to face or by telephone, measures where people shop now, how often, what would make them switch and what the neighbourhood lacks. Details of this kind of fieldwork are on our CAPI and CATI page.

6. Analysis: from data to a sales estimate

There are three approaches, increasingly quantitative, and they can be combined.

• A weighted checklist: each site is scored on criteria agreed in advance (demand, competition, access, visibility, cost). It is simple and transparent, but the weights are subjective and the score does not translate into sales.
• The analogue method: the proposed site is compared with the chain's existing outlets in similar settings. Their sales, set against population and competition in their own areas, give a plausible range for the new outlet. It works well when the chain has enough comparable outlets.
• Statistical and gravity models: with enough outlets, a regression estimates how much each factor (population, competition, traffic, floor space) contributes to sales. Gravity models estimate the probability that a resident will choose a given outlet based on its attractiveness and distance. Both need good data and are validated on existing outlets before being used to forecast.

The output of a good study is not a single number but a range of likely sales for each site, with its assumptions written down, and a recommendation: which site, in which format, with which risks. For a chain, the analysis also covers cannibalisation: how much of the new outlet's sales will be taken from the chain's own nearby outlets. Turning these results into an investment decision is part of our decision support work.

7. From location study to decision

The report ends with a direct comparison of the candidate sites, on the same criteria and with the same estimated sales range, so that the decision can be discussed openly by management. For each site it also states the break-even point: the monthly sales needed to cover rent, staff and fit-out depreciation. If the whole estimated range sits above break-even, the site is solid; if break-even falls in the middle of the range, the risk is real and should be negotiated, for example through a lower first-year rent or an exit clause.

After opening, the location study becomes a learning tool: actual sales are compared with the estimate, and the gaps show which factors were underestimated. A chain that does this at every opening improves its model from one outlet to the next.

8. Limits and common mistakes

• Counting on a single day or at a single hour and extrapolating to the whole year.
• Mistaking traffic for custom: through-traffic and hurrying pedestrians are not shoppers.
• Circles on a map instead of travel times, ignoring physical barriers and real habits.
• Ignoring planned changes: a boulevard about to close for roadworks, a shopping centre under construction, a relocated bus stop.
• Unestimated cannibalisation of your own outlets.
• Too few analogue outlets for a statistical model: with three or four outlets, a regression cannot be trusted and a qualitative analogue approach is more honest.
• Outdated demographic data in neighbourhoods that have grown quickly in recent years.

9. Which method for which question

10. What to ask the agency

• How the catchment area is defined: travel times, current customer data, physical barriers.
• Which demographic sources are used and how recent they are.
• The counting protocol: points, days, intervals, seasons.
• The competitor inventory and the criteria used to assess them.
• The standard accessibility and visibility form.
• The sales estimation method and its assumptions, written down.
• An estimate of cannibalisation for nearby outlets of your own.
• A range of likely sales rather than a single figure, and a comparison of the candidate sites.
• Planned urban and residential projects that could change the area.

A location study costs a fraction of a year's rent and can prevent a contract that cannot be undone. If you are assessing one or more sites, our market research team can help you compare them on evidence rather than impressions.

See also: Market research and segmentation Data analysis and desk research Decision support Market sizing and segmentation Retail audits and store checks Market entry and feasibility study

The Perspicuus research team · 2 June 2025