Placer.ai

From manual surveys to location-based foot traffic estimates

For years, a retail chain that wanted to know whether a new location was worth it relied on surveys, manual counts and a lot of intuition. Placer.ai, founded in April 2015 by Noam Ben-Zvi, Zohar Bar-Yehuda, Ofir Lemel and Oded Fossfeld, produces this figure from a different direction: it buys anonymous location signals from millions of phones through various apps, and runs statistical models on them that fill in the gaps and expand a sample into an estimate.

What the data reveals: frequency, dwell time and competitor effects

The outputs are things that did not exist before: the number of visitors to a branch by hour and day, dwell time, the geographic radius customers come from, which other businesses they visit on the same day, and how traffic changed after a competitor opened. The customers are retail chains, commercial real estate, local authorities and investors.

A 2018 platform, and a funding path to $1.5 billion

According to the official company page, Placer.ai launched as a commercial analytics platform in 2018, three years after it was founded as a company in April 2015. That gap is typical of data companies, which need time to collect enough signal volume before the product becomes useful. The funding path: $100 million in a Series C in January 2022 at a valuation of about $1 billion, and $75 million in a Series D in July 2024 at a valuation of about $1.5 billion. In February 2024 the company officially announced that it had passed $100 million in annual recurring revenue (ARR). "Reaching $100 million ARR is a significant milestone for Placer.ai," said CEO Noam Ben-Zvi at the time. "Our growth reflects the critical role our platform plays in helping businesses adapt in a changing market landscape." In January 2025 the company laid off about 150 employees as part of a move to profitability.

The privacy question in anonymous location data

The question that follows this field is privacy. The data does not identify a specific person, but the raw material is the movements of real people, who are usually not aware that an app they installed sells their location. Regulation of location data keeps changing in the United States and Europe, and this is a real business risk for every company in the category.

The statistical challenge: sample bias and correcting it

The main statistical challenge is bias. A sample of users of certain apps does not necessarily represent the whole population, and different geographic areas, age groups and income levels are represented in it to different degrees. Correcting the bias, by calibrating the sample against known ground-truth data such as census data or reported sales, is the heart of the work, and also where the companies in the field differ from one another.

The real value: comparison over time, not absolute measurement

The clearest value of the data is actually in comparison over time: even if the absolute number is not accurate, the change between quarters or the difference between two branches is much more reliable, and that is usually the question decision makers care about. This is how an estimated figure becomes a useful decision tool, as long as it is used for comparison and not as an absolute measurement.