UrbanZoom Research

Measuring COVID-19 Impact on Retail Businesses – A Study of Transport Ridership Data in Singapore (2020)

By June 26, 2020No Comments
Singapore Footfall Data for March 2020 at 6pm

Footfall Data for all postal codes in Singapore at 6pm in March 2020 (Weekday).

Retail Businesses = Footfall Traffic = Transport Data?

The coronavirus pandemic has created havoc in the world’s economy in a matter of months.

Retail businesses in particular, where services are provided in brick and mortar premises, have been hit hard due to the various social distancing measures put in place.

To study this impact in Singapore, we plowed through months of transport ridership data (both MRT and buses) and created a footfall model to gain some insights on the scale of the damage to retail businesses.

MRT ridership has dropped 83% in May 2020 vs May 2019.

Monthly Train Ridership Comparison for May

Real estate is all about “Location, Location, Location”, right?

And with great location, comes footfall traffic and customers…

…or so goes the conventional wisdom, before COVID-19 broke all the rules.

Let’s see how the concept of a great location, can abruptly change in this post-pandemic world.

To find out, we approach this study in 3 parts:

  1. A cursory look through MRT & Bus ridership data (available here from LTA)
  2. Footfall data derived from above datasets and our own algorithm
  3. What this data mean for retail business & property owners?

PSA: If you’re a retail business owner (eg. cafe, restaurant, clinic, hair salon, etc) or landlord and want to get more footfall data specific to your retail location, we maybe able to help (see bottom of this post).

1. Insights from transport ridership data


Let’s start with the train ridership data.

For comparison, we look at changes in total rides during recent months in 2020 vs same months in 2019.

Total Train Ridership (2020 -2019 Comparison)

Even before Circuit Breaker (CB), the total rides in 2020 have dipped compared to 2019’s figures. In the first 2 months of 2020, we see a significant drop in the train rides at -16%.

Things got worse once the CB was enforced in April. Train ridership dropped a whopping -83% in May 2020 with 14.4 million rides compared to 83.5 million in May 2019!

With this overview in mind, which train stations have seen the biggest drop in train ridership in May 2020 as compared to a year ago?

Top 8 Worst Performing MRT Stations

The chart above shows us the MRT stations that were hit worst by the pandemic.

Overall, Bayfront MRT was affected the most, with a -98% drop in passenger volume! Only 15,000 riders visited Bayfront in May 2020, compared to 926,000 in May 2019.

Moreover, most of the worst affected stations are near tourist spots such as Garden’s by the Bay, Changi Airport and Esplanade, just to name a few.

Top 8 Worst Performing Stations with Tourist Spots

(If you want to see how other MRT stations are affected, access the whole list of train stations and their corresponding change in percentage values in our annex below here)

It is not surprising that all train stations saw a decline in train ridership due to the pandemic.

However, what are the 5 most resilient train stations? Below are the top 5 stations with the smallest change in train ridership.

Top 5 Resilient Performing MRT Stations

The least affected stations are in industrial zones found mostly in the west of Singapore such as Gul Circle and Tuas West Road. Even then, their values are more than 50% lower than a year ago.

What about the busiest train stations in 2019 and how are their passenger counts faring amidst this crisis?

May 2019 Busiest MRT Stations

In May 2019, Jurong East MRT was the busiest station with slightly over 2.3 million boarding passengers over the course of a month.

One year later, there are only about half a million passengers boarding form Jurong East MRT station.

What if we look at year-on-year % changes for these busiest stations?

At a glance, the stations that are most affected by the pandemic are mostly closer to shopping districts like Orchard, Bugis and popular office locations like Raffles Place and Chinatown in the Central Business District.

These 4 stations saw close to -85% drop in total passengers.

The heartlands such as Tampines & Yishun are faring better as their drop is hovering only around -70%.


When it comes to bus ridership, the trends seen in the data also shows that bus ridership has reduced significantly due to the Covid-19 pandemic.

Bus Data Singapore Map

The video above shows us how the passenger volume has decreased from March to May 2020. The size of the dots in this chart represents passenger volume at individual bus stops.

Pre-CB (March 2020), we see some big dots in specific areas of Singapore. These are your usual suspects, with bus stations like Tampines, Woodlands & Boon Lay interchanges seeing slightly over a million passengers respectively.

We all know that the party stopped once Circuit Breaker (CB) came into effect in April 2020.

But are there any bus stops that have actually gained in terms of passenger volume from March to May 2020?

Increased Bus Ridership due to Covid-19

Surprisingly, there were a handful of bus stops that have increased passenger volume in May 2020 as compared to March this year. The bus stop with the biggest increase is from the Expo Halls bus stop.

Expo Halls 1/2/3 bus stop passenger volume increased by 340%!

This spike in passenger volume can probably be attributed Singapore Expo being currently being used as a COVID-19 community facility to hold stable Covid-19 patients.

The other bus stops that have registered a huge increase in passenger volume are from industrial area/camps which might also be linked to Covid-19 mitigation efforts.

Let’s switch gear and look at the 10 busiest bus stops in Singapore. Were there any movements in the top 10 stations from March to May? Let’s find out.

Top 10 busiest bus stops in March 2020

Bus Stop

Passenger Volume

1 Boon Lay Int            1,622,960
2 W’Lands Temp Int            1,530,885
3 Tampines Int            1,047,862
4 Yishun Temp Int               890,725
5 Bedok Int               872,759
6 Toa Payoh Int               865,273
7 Jurong East Temp Int               769,194
8 Ang Mo Kio Int               712,253
9 Choa Chu Kang Int               571,630
10 Clementi Stn               566,525

Top 10 busiest bus stops in May 2020

Bus Stop

Passenger Volume

1 Boon Lay Int 640,499
2 W’Lands Temp Int 540,770
3 Tampines Int 391,542
4 Bedok Int 370,418
5 Toa Payoh Int 355,077
6 Yishun Temp Int 304,995
7 Ang Mo Kio Int 275,879
8 Jurong East Temp Int 272,965
9 Hougang Ctrl Int 203,717
10 Choa Chu Kang Int 182,611

Based on our analysis, we observe that the top 10 stations are actually all from massive interchanges such as Boon Lay, Tampines and Toa Payoh.

While the top 3 busiest bus stops have stood their ground. There is some movement among the other bus interchanges, with Hougang entering the list as the 9th busiest station in May.

2. Building an AI to predict Footfall

While this has been a cool transport analysis, how is this data useful to say Tenants, Landlords or retail shop owners?

Our hunch is that given these ridership data, we can build a model to provide inferred footfall data (ie. people’s movement patterns) and eventually an analytics engine that ties in other property-level and population data.

This work is still under R&D phase but our hope is that such data analytics could help both retail business and property owners to make more informed decisions over time.

As a start, here’s an example of the output from our algorithm that predicts footfall information for every known postal codes (over 120k+) in Singapore.

Singapore Footfall Data for March 2020 at 6pm

This map shows the footfall information of the entire of Singapore on a Weekday Evening at 6 pm in March 2020.

While our model covers the entire Singapore, let’s zoom into 3 specific regions: Shopping Districts, CBD Office Areas & Residential Estates.

  1. Shopping District- Orchard

Mention shopping in Singapore and the first place that comes to mind is the Orchard Road shopping belt.

It is a well known fact that Orchard Road can get crowded on the weekends. But how much crowd are we talking about?

ION Orchard Footfall Data March 2020 ION Orchard Footfall Data May 2020

On the left side we have footfall data for March 2020 vs right side for May 2020.

As expected, March has bigger dots which means that larger crowds as compared to May 2020.

In fact, according to our model, there were approximately 6,000 people around Ion Orchard Mall during a March Weekend at 12pm noon. This is about 4 times more than the estimated footfall in May, with 1,600 people seen during the same time period (weekend 12pm noon).

How has the footfall patterns changed for other places in Orchard?

Let’s take a look at this list:

Building March Predicted Volume May Predicted Volume
Ion Orchard    6,000      1,600
Wheelock Place    4,100      1,100
Takashimaya    3,600      1,000
The Paragon    3,200         900
Tang Plaza    4,600      1,400
Shaw House    3,200      1,000
Scotts Square    3,500      1,100

The table above shows the absolute numbers of the predicted footfall traffic based on our model.

The following table shows us the estimated percentage change in the footfall data for these places:

Building Predicted  % Change
Ion Orchard -73%
Wheelock Place -73%
Takashimaya -72%
The Paragon -72%
Tang Plaza -70%
Shaw House -69%
Scotts Square -69%

Based on our model, footfall in orchard dropped by ~70% in May 2020 vs March 2020.

2. Central Business Districts Office Areas – Raffles Place

Raffles Place is bustling on weekdays due to the huge office crowd. But with everyone working from home now, how has the footfall traffic been affected by the pandemic?

Ocean Financial Centre Footfall Data March 2020 Ocean Financial Centre Footfall Data May 2020

From our algorithm’s output, we predicted that in March 2020 weekday morning at 9 am, there are about 5,000 people around the Ocean Financial Centre (within a 150m radius circle). However, we are estimating that the crowd volume might have dropped to around 500 in May 2020.

Are these patterns that have been observed consistent with the other places around Raffles Place?

Let’s dive deeper into the footfall data of the other adjacent buildings.

Building March Predicted Volume May Predicted Volume
OUE Bayfront 6,900 700
One Raffles Place 6,500 700
AIA Tower 3,700 400
Republic Plaza 4,600 500
Samsung Hub 4,300 500
Prudential Tower 4,300 500
Ocean Financial Centre 5,000 600

The values in the table above are footfall traffic outputs from our Machine Learning Algorithm. We observe that May values predicted by our algorithm is on the low side.

But what is the percentage change of footfall volume between March and May 2020?

Building Predicted Approximate % Change
OUE Bayfront -90%
One Raffles Place -89%
AIA Tower -89%
Republic Plaza -89%
Samsung Hub -88%
Prudential Tower -88%
Ocean Financial Centre -88%

Based on our model, footfall around Raffles Place might have been reduced by about 89% in May 2020 vs March 2020.

3. Residential Estates- Tampines

Moving on to our last example, the residential estates. We have seen how the footfall has dropped in both the Shopping Districts and Office areas. How about residential areas?

Let’s explore how has the footfall been affected in the Tampines housing estate.

Tampines Footfall Data March 2020 Tampines Footfall Data May 2020

Based on our algorithm, we observed a drop in the number of people in Tampines MRT at 9 am on a weekday morning.

In March 2020, we estimate that there would have been 7,200 people around the vicinity whereas in May there was close to 2,000, representing a -72% drop, according to our model.

Let’s close in on other adjacent locations in Tampines:

Building March Predicted Volume May Predicted Volume
Tampines MRT 7,300 2,000
Tampines 1 6,700 1,900
Abaacus Plaza 4,800 1,400
Century Square 5,100 1,500
CPF Tampines Building 6,400 1,900
Our Tampines Hub 1,900 600
Hitachi Square 2,300 900

And here’s the corresponding percentage change in footfall:

Building Predicted Approximate % Change
Tampines MRT -73%
Tampines 1 -72%
Abaacus Plaza -71%
Century Square -71%
CPF Tampines Building -70%
Our Tampines Hub -68%
Hitachi Square -61%

Heartlands like Tampines saw a somewhat smaller impact in footfall compared to Orchard and CBD areas.

3. What does this footfall data mean for you?

(Continue reading if you own a cafe, restaurant, clinic or any retail shops!)

As mentioned earlier, this work is still in R&D phase. While the precise footfall values still need fine-tuning, we think this can already be useful in trend analysis across time.

Furthermore, since our model covers every postal code in Singapore, at specific timings (different hours of the day, weekend vs weekday, etc), this level of granularity can assist in analysis across various locations as well.

Over time, we hope to integrate these data with our ZoomValue AI, as well as other property and population data sources, to create a holistic analytic picture for retail business and real estate owners.

If you’re interested to find out footfall data specific to your retail location, feel free to drop us a WhatsApp message at 8811 1181. We’ll see how we can be of help 😅


% Change of Passenger Data from All MRT Station