Why a good site still fails, and which warning signs data can catch

A site that scored well can still fail for reasons the scoring never looked at: customers can see it but cannot easily get into it, the busy hours fall when it is closed, a rival opened six months later, the street was dug up in the first season, or the people nearby are not the people who buy what it sells. Some of these show up in location data early enough to act on; others only show up in the till. Knowing which is which tells you what to check before you sign and what to watch after you open.

This piece is for owners, franchise managers and expansion teams looking back at a disappointing opening, or trying to avoid one. It goes through the common causes one by one, says honestly what data can and cannot see, works through an example, and ends with a watch an AI employee can keep on a new site.

The usual causes

Access, not visibility

A unit on a fast road is seen by thousands of drivers who cannot stop. A corner on a one-way street may be impossible to reach from the direction most customers come. Road speeds help here: Traffic Flow shows how fast the roads beside a site move against their free-flow speed, and a road that flows freely at 60 km/h past your door is a road people drive along, not to. What it cannot show is a missing turn lane or a kerb with no parking; those need a visit.

The wrong hours

An area can be busy, just not when you are open. Offices fill a district from nine to five and empty it by seven; a bar strip is quiet until dark. Mobility Footfall gives the busiest hour and the busiest and quietest days around a place, as people-hours of presence. If the peak sits outside your trading hours, the site's headline footfall overstates what you can sell to.

A rival who came later

The study described the street on one day. A competitor that opens nearby in month four changes the arithmetic the study relied on. This is the most catchable cause of all, because new places appear in xMap's places database as they are found.

Disruption in the first season

Roadworks or a closure outside the door in the first months, before regulars form, can cost more than the same works a year later. Traffic incidents shows closures and roadworks around a place as they appear.

The wrong audience

Plenty of people live nearby, but they are students when you sell family meals, or retired when you sell gym memberships. Where population figures exist, Demographics gives the age groups around a site. In countries without them, including the United Kingdom and the United States, this check falls back on judgement.

Execution

Sometimes the site is fine and the operation is not. Ratings are the clearest outside sign: a new branch rated well below the others in the same chain, against well-rated neighbours, points inward rather than at the location.

What data can and cannot catch

What each cause looks like from outside, and what only you can check.

An example

A made-up example. A bakery chain opens a sixth branch in Melbourne on a street that scored well: plenty of residents, few rivals, good footfall. After four months it sells far less than the others. Looking back:

• The footfall around the site peaks at 12:00 on weekdays, from offices; the bakery's best sellers are morning pastries, and it opens at 07:00, when the area is still quiet.
• Two new bakeries opened within 800 m in month three, one rated 4.8 within weeks.
• Roadworks closed the side street with the loading bay for six weeks in the first month.
• The branch's own rating is 4.4, in line with the chain, so service is probably not the cause.

None of these was in the original study, and three of them were visible in data at the time. A change in opening hours toward lunch, and a quick look at what the new rivals offer, are the obvious first moves.

Mistakes when diagnosing

• Blaming the location first. Check the branch's rating against the chain and its neighbours before deciding the site was wrong.
• Reading people-hours as customers. Footfall adds up presence; an office district can have huge daytime figures and no evening trade.
• Looking only at the day the study was done. Compare the street as it was then with the street as it is now: new rivals, closures, works.
• Expecting fresh footfall. Each day of footfall arrives about 10 days late, so last week's dip will not show yet.

A watch an AI employee can keep

Most of the catchable causes can be watched from opening day, so the next disappointing branch is explained in weeks, not quarters.

1. When something changes: watch the new branch and up to four others for a new place of your type opening within 1 km, checked every day. A second trigger of the same kind, set to road closures, watches the streets around the site.
2. xMap LLM: told to say in one line how much each change matters to the branch, given its distance and rating.
3. Team inbox or Email: only when something changed.
4. Separately, a monthly flow with Mobility Footfall around each branch for the busiest hour and day, and Nearby places for the branch's own rating against rivals within 500 m, sent as a Make a report PDF.