Estimated footfall vs. real footfall: why POINOVA uses estimated

When someone is looking for where to open a business, sooner or later the magic word turns up: footfall. How many people walk past the door. And right away comes the reasonable doubt: is that footfall real, or is it a calculation? At POINOVA we work mainly with estimated footfall, and we do it on purpose: it's a considered decision that has to do with your privacy, with cost, and with what you genuinely need to make a good decision. We're going to explain it with all our cards on the table.

Two very different ways of measuring how many people pass by

There are basically two ways to know how many people move through an area, and it's worth not confusing them, because they answer different questions.

Estimated footfall is calculated from what we already know about the area: how many people live nearby, what amenities are close by (stations and stops, schools, universities, hospitals, shopping centres, hotels) and which points of interest draw the most people. With all those ingredients you model an area's footfall potential. It's a portrait of the context: it tells you whether the place has what it takes to bring people in.

Real footfall, by contrast, tries to count specific people who passed a specific point. The most common source is mobile data from telecommunications operators: the anonymised, aggregated trace that phones leave when they connect to the masts. It's a portrait of the count: it aims to tell you how many devices passed, at what times and where they came from.

Why POINOVA chooses estimated (deliberately)

We use estimated footfall because it's the best tool for the decision that really matters. It's a choice we made for three solid reasons.

1. Privacy and GDPR

Mobile data starts from the movements of real people. However anonymised and aggregated it arrives, its origin is sensitive, and the European data protection framework (GDPR) is demanding about how it is collected, processed and shared. We don't buy this data from any operator. Estimated footfall is built on official, open sources (each country's censuses and public statistics, amenities and public transport stops) that don't track anyone. There are no people behind this estimate, just territory. For a European company that wants to sleep soundly, that difference matters.

2. Cost and coverage

Telco data is expensive and usually contracted area by area and period by period. That pushes up the price of the product and, above all, fragments it: you have data where you've paid, and a hole right next to it. Estimation, by leaning on public statistics, covers the territory in a uniform and comparable way. You can look at neighbourhoods in different cities in the same country with the same logic, without depending on whether someone sold that piece of the map. To compare candidates, that even coverage is worth its weight in gold. And there's a direct benefit for you: by not carrying the cost of mobile data, POINOVA can be one of the most affordable tools on the market, giving you the location intelligence you need at a far lower price.

3. To decide on a location, it's usually enough

And here's the honest crux: for the question almost everyone asks (does this area have potential for my business, and more than that other one?), context weighs more than the exact count. Knowing that a spot has a large population around it, the right income level for your product, amenities that draw people in and convenient access takes you a long way. Estimation answers the "does it make sense here?" question without you needing to count heads one by one.

What each one tells you (and what it doesn't)

Being honest means drawing the limits of both clearly.

Estimated footfall tells you: whether an area has footfall potential, how it compares with other candidates and what drives it (transport, educational and healthcare facilities, retail, tourism). It's excellent for prioritising, ruling out and comparing.

Estimated footfall doesn't tell you: exactly how many people walked along your pavement on Tuesday at six in the afternoon, nor how many of them came in. It's a measure of the area's potential, not a count of people. And like any model, it's more accurate in some places than others (more on that later).

Real footfall (telco) tells you: an approximate count of devices passing a point, with curves by hour and by day, and where they came from. It's powerful for measuring a flow that already exists.

Real footfall doesn't tell you: whether those people are your customers, or whether they'll come in. A nuance almost nobody mentions: it counts devices, not people, and not everyone who walks past is a potential customer. Nor does it solve the "should I open here?" on its own: a huge flow can come with rents and competition that eat it all up.

When would real footfall genuinely matter?

We're not going to pretend real footfall is never useful. There are cases where the fine-grained count makes a difference, and it's only fair to acknowledge them:

• Pure passing-trade businesses, where you literally live off whoever crosses in front: a newsstand, a coffee-to-go right by a tube entrance, a convenience store on a very busy thoroughfare. Here the hourly peaks can be decisive.
• Fine-tuning a unit you already have, to adjust opening hours or staffing to the real footfall curves. That's day-to-day operations, not a location decision.
• Measuring the effect of an event or a campaign on the traffic of a specific street over a few days.

Notice the pattern: real footfall shines when you already have the spot and want to fine-tune, or when your model is 100% "people pass, people come in". For the prior question, where to open, context rules.

Let's be honest: where estimation falls short

An estimate is a model, and models have their weak points. We're not going to hide them from you:

• Tourism and non-resident flows are captured less well. A very touristy area draws people who neither live nor work there, and a model based on residence and amenities only sees that in part.
• The micro-differences within a single street (the good side versus the bad, the corner versus mid-block) fall outside the reach of an area-level estimate.
• Very recent changes (roadworks that cut off a route, a station that's just opened) don't show up until the sources are updated.

That's why estimation is an excellent compass for comparing and prioritising, and it's worth closing the loop with a physical visit to the premises: no dataset can replace seeing the street for yourself.

An illustrative example (made-up figures only to show the logic)

Imagine, by way of illustration, that you're comparing two premises for a speciality coffee shop. The figures below are made up and only serve to show how the reasoning works; the real ones would come from official and open sources.

• Premises A, next to a station with a large population and offices around: very high estimated footfall. The context says "a lot of people from the profile you're after pass through and live here".
• Premises B, on a quiet residential street with a similar income level: notably lower estimated footfall, because there are no amenities to draw people in and no large passing flows.

With that you already have a reasonable decision: A starts with an advantage for a business that lives off passing trade. What would a real telco figure add? It would confirm A's hourly curves (useful for opening hours) and little else for the underlying decision. Estimation already told you the important part. What no dataset will tell you is whether A's rent stacks up: that you cross-check with the competition in the area and with your own numbers.

In short, and with no hedging

POINOVA does not buy mobile phone data from telecoms operators. We work mainly with estimated footfall built from population, amenities and points of interest, all of it from official and open sources; and where a public body openly publishes mobility measured with mobile phones, we show it to you separately, with its source and date. We do it for privacy, for cost and coverage, and because to decide where to open it's usually what you genuinely need. We tell you what estimation does well and also where it falls short, because we'd rather have a customer who understands the data than a customer dazzled by a figure that doesn't mean what it seems to.