Here you will see the results of foot traffic correlation analyses we performed for major brands. We compiled foot traffic and revenue data starting from the fourth quarter 2018 through the third or fourth quarter of 2021 depending on availability of revenue data. Dates were adjusted according to the brand’s fiscal year.
Can mobility data predict quarterly revenue?
Does human mobility data predict quarterly revenue?
It’s earnings season for publicly traded companies in a tumultuous trading period. Pandemic restrictions, the supply chain, and worker shortages are impacting revenue, and financial analysts need timely external data to evaluate portfolios effectively.
What predictions can be made using human mobility?
Unacast specializes in understanding how people interact with the world. We offer aggregated foot traffic insights on a global scale and for any point of interest. Foot traffic insights tied to individual retail locations is particularly impactful for financial analysts in assessing consumer behavior across retail portfolios.
Drawing customers into brick-and-mortar retail locations is part of a longer journey to landing a sale, and we examined how foot traffic data correlates to quarterly revenue between fourth quarter 2018 up to third quarter or fourth quarter 2021 (depending on availability) for brands like Best Buy, AMC Entertainment, and Marriott International.
With Unacast data, analysts can avoid earnings disappointments far in advance of 10-Qs hitting the news. Our data offers daily, monthly, and quarterly foot traffic for national venues with only a four-day lag.
What’s behind our analysis?
The growth of e-commerce in the last two decades complicates the relationship between feet in
the door and total revenue, especially after the pandemic changed B2C relationships perhaps permanently. Even brands like IKEA with its evolving experience-first showrooms pivoted to online shopping and contact-free delivery service.
Can we make the assumption that increased foot traffic means increased revenue?
Spoiler alert: our analysis says yes, as long as we establish a ground truth for what our data can and cannot predict and why. We ran correlation analyses between foot traffic and revenue across multiple industries and noticed some industries showed very high correlations while others did not. We needed to establish why some brands showed limited connections to our foot traffic data, and we determined the following:
1. Brands where in-person transactions over e-commerce are pivotal to revenue show the strongest relationship with our foot traffic data, and these are the brands presented in this package.
- For example, while brands like Marriott International and Hyatt Hotels offer online booking, the transaction itself generally translates to a physical presence at each brand’s locations which can be accounted for by Unacast’s data. (The meta-verse isn’t at a stage where we can enjoy a luxury hotel room from the comfort of our own virtual reality set. Yet!)
- The physical presence component of the Hotel industry, along with industries such
as Self Storage, are therefore very strong candidates for revenue prediction using foot traffic data. Industries where e-commerce plays a huge part in revenue do not show strong correlations, though foot traffic data can still be relevant for portfolio analysis.
2. Our foot traffic counts rely on accurate point-of-interest (POI) construction for a venue, and as discussed in our ground truth blog, the complexity of a venue’s shape directly interferes with the accuracy of visit allocation.
- There are brands largely operating out of locations which can be hard to map with total accuracy, such as multi-story shopping complexes, and this diminishes the correlation between revenue and traffic. We are therefore not presenting those brands here.
3. Our revenue predictions are dependent on our Venue Data Package representing a majority of a brand’s total United States venues.
- The fewer venues in our package, the weaker our correlation analysis. Brands with large location disparities were also excluded from this presentation.
4. Our sample size is small because our data only goes back to Q4 2018.
- The pandemic created a situation where the variance between revenue and foot traffic is more aligned. In theory, it’s important to disclose that this may enhance the correlation we see during this time period, however, the total influence of the pandemic is still unknown.
With these ground truths disclosed, we present with confidence reliable correlations between foot traffic and quarterly revenue in these industries:
In the following sections, you will see the results of correlation analyses we performed for major brands. We compiled foot traffic and revenue data starting from the fourth quarter 2018 through the third or fourth quarter of 2021 depending on availability of revenue data. Dates were adjusted according to the brand’s fiscal year.
What earnings hits and misses have been reflected in Unacast’s foot traffic data?
AMC Entertainment is the Reddit crew’s newest meme stock. The pandemic decimated AMC’s profits, but the theater chain made modest gains in their third quarter 2020. The image on the right shows changes in revenue and foot traffic that align per quarter, where these dips and gradual increases are reflected in visitation counts up through the third quarter of 2021.
Our analysis of multiple theater chains bears the same fruit. The R-Squared values listed in the image on the left indicate the strength of our linear trend model as calculated using Tableau on a 0 to 1.0 scale, and the P-value describes the significance of the model. We want P-values to fall below 0.05. Each circle on the graph indicates a financial quarter for the period assessed.
AMC Entertainment Holdings and Cinemark Holdings show a strongly correlated relationship between revenue and traffic.
Why is the model a worse fit for IMAX Corporation?
While the relationship is still strong, we determined IMAX operates in association with other theaters, making the traffic to revenue relationship less clear-cut compared to AMC and Cinemark. Additionally, IMAX’s presence in shopping malls also diminishes the correlation.
Another industry closely tied with in-person consumer visitation is hotels. Our examination of five major hotel brands follows similar patterns to movie theaters, with linear relationship trends between 80 to 96%.
We are also strong for the Self Storage and the Casual Restaurants industries. Variation in R-Squared values for these models can be accounted for by minor differences in each brand’s total U.S. venues compared to our Venue Data Package.
Dave & Buster’s Entertainment experienced similar swings in revenue and traffic to AMC Entertainment. The bar chart to the right demonstrates revenue to traffic variations for all assessed quarters.
For the Car and Truck Rental industry, you can view how a major disparity in Venue Data Package locations impacts the model fit. The correlation for Avis Budget Group shows a 0.45 correlation with a significant P-value. The company’s 10-K reports over 4,000 North American locations between their Avis and Budget brands, whereas our Venue Data Package accounts for only half of this venue number. We include the weaker correlation figure here to demonstrate how our data still retains a significant relationship with revenue even with differing location counts.
The Computers and Consumer Electronics industry relies extensively on e-commerce, which does obscure our ability to use traffic as a revenue predictor. In the image above, we show how foot traffic data can still share a relationship with in-person revenue. According to Best Buy’s SEC filings, online sales account for nearly 45% of their total revenue. Our model fit shows a 0.49 correlation, which generally matches the remaining revenue total. Similar to Avis Budget Group, our foot traffic data reliably accounts for the revenue share derived directly from in-person transactions.
Foot traffic clarifies a company’s financial outlook even when major portions of revenue come
from online sales. Unacast data is also an essential barometer of business performance on the ground, especially for those businesses considering pivots to e-commerce. This is important for analysts evaluating company strategy over the long-term and predicting future stock fluctuations.
Conclusions
Data is essential for evaluating a company’s financial performance, but can foot traffic data be used for revenue forecasting? Our analysis confirms yes. The pandemic pivoted consumers to online shopping, but many industries rely on traffic for a majority of their revenue, and that’s where we found the strongest traffic to revenue correlations.
- Foot traffic is a predictor of financial performance for industries where in-person transactions inform a majority of a company’s total revenue. We found significantly high correlations between traffic and revenue in the following industries: hotels, self storage, movie theaters, car and truck rental, and casual restaurants.
- Correlations between foot traffic and revenue aren’t always limited to industries centered around in-person transactions. We found significant predictive relationships for industries like Computers and Consumer Electronics where e-commerce sales are dominant but brick-and-mortar stores are still in play.
- Foot traffic is a revenue forecaster even when our data set includes only a portion of total venues for a brand. Our Venue Data Package does not include all of Avis Budget Group’s North American locations, but our correlation was significant in accounting for revenue since 2018.
Schedule a meeting to learn more about how foot traffic can help your business.
