Money Talk — In your own language

Published in

Rupal Bhatt

5 min readMar 8, 2018

Toronto is a multilingual city. Approximately 200 different languages are spoken here, 45% of the residents speak a mother tongue other than French or English, according to the 2011 census.

Seven languages have more than 50,000 speakers in the city: Cantonese, Italian, Spanish, Tagalog, Tamil, Mandarin, and Portuguese. 55 others have between 1,000 to 50,000 speakers.

It is estimated that almost 170,000 new immigrants will come to Toronto in 3 years (according to The Toronto Star of November 17, 2017). All these people will need jobs, school, housing, and banks.

All the major banks are offering services in various languages besides English. It is easier to build trust and relationships with clients when communication is in their mother tongue.

The goal of my project is to find out how banks can attract these newcomers as their clients.

Main Tools used are:

Here are the major parts of this project.

Gathering the languages offered by different branches of a bank all over the city
Gathering the language data from census 2016 for each postal code in the city
Comparing and analyzing over-usage and under-usage of language services offered by the bank

I have used Python and BeautifulSoup to scrape data from the website of the bank.

Their website gives the list of branches. There are 41 pages and each page has 10 branches listed on it. The program loops through these pages and then through the list of branches to get to the page of each individual branch.When you click on the title of one of them the page for that particular branch opens. This page has the information about the languages offered in that particular branch.

Here is the code of BeautifulSoup to scrape the webpage of a bank and get information about the all the branches

There are total 41 pages to scrape through to get information about 410 branches in the city.

This piece of code will give the list of websites for all the branches.

Once we have that list it is easy to go to each website and scrape the information about languages offered at each branch.

All the information about languages and branch address is stored in a data frame. With help of this data frame, it is easy to use Google Map

Here is one of the Geo Overviews of the branches that comes up when you search for branches in Toronto, ON. As you notice it includes several branches away from GTA area

All the information is organized as per the requirement and gathered in one data frame as you can see below.

Next step is to find out the information from Stats Canada website. This website has extensive information about Canadians, including important information about the languages they communicate in. It is available on this website http://www12.statcan.gc.ca/census-recensement/index-eng.cfm?HPA=1

Once you decide all the data you want from this website you get a .csv file for each postal code (first 3 characters) containing relevant data. The information is stored in a single data frame.

I chose to use mySql windows function to get the percentage; this can be easily done using Python too.

Here is an example of this data frame created at the end.

Using pylot of Matplotlib, I have created some examples to figure out how the language service of the bank compares to the languages spoken in an area.

Postal Code: M4L