PORTFOLIO
Digital Transformation, Strategy, Organisation Change Management
What are we trying to solve:
Digital has a variety of definitions and many organisation did not have a clear and well-defined definition of digital, before going or accelerating digital. This would be an issue as departments would be running in "different" directions. Many field based employees are also skeptical on the value of digital and are resisting change to adopt digital into their practice.
How are we solving it:
Conducted surveys and interviews to get an understanding of what digital means, at various levels from business head to marketers to sales representative. Work on small and tangible projects to show employees what digital is and how it is not an opposing force, but something that can empower the field based employees to be much more effective in driving outcomes
What is the outcome:
Take a look at the video, the outcome we have in effecting change within the organisation.
Creating a Data Science Culture - Strategy, Change Management
What are we trying to solve:
The organization has run into a "commitment fatigue" with the multiple failures in their data science effort. For example, previous advanced data analytics projects such as marketing mix model do not really give much insights to support promotion budget allocation for marketers. Data driven decision making is an important strategic direction of the company but it's people are not seeing the value of data and think that those advanced analysis is not suitable in their industry.
How are we solving it:
Interview marketers and sales person to get a sensing on what are their pain points and in their opinion, what kind of data or analysis can help them in their business. We work on small projects, tangibly showing the value of data analytics, beyond merely just a BI dashboard. We also implemented company wide programs to promote data literacy and hence, data democratisation
What is the outcome:
Take a look at the video, it is a resounding success and the company decide to scale what was done globally and mandating every market to hire their own data science team in the commercial unit.
F&B: Customer Segmentation and Optimization
What are we trying to solve:
An European restaurant offering 5- and 8-course set lunch/dinner would like to know how customer is choosing the items within each of the course category and use this information to optimise the menu and also, to turn their business back to growth.
How are we solving it:
The POS sales data includes: 1) Items ordered by each individual, 2) Time sitting in and checking out, 3) Number of people per table
​
A simple visualisation (BI) tells us that 8 course is more popular during dinner time and weekend, where customers have more time to sit down and enjoy the companion of their family and friends. Certain items are popular for 5 course and 8 course while some are all time favourites
​
We also ran an Apriori (Market Basket) model to and observe that there are 2 distinct group of ordering patterns - A more flavourful segment of customer and the other who prefers healthier combination. This would help them to understand how customers are interacting the menu and how they could further customise and change some items in their menu
Here, we created 2 other indicators, seat turnover and table turnover rate. This would tell us 1) How packed the restaurant is and is the restaurant efficient and maximising it's capacity (or seats) for the customers. Obviously, there are so much room for improvement we observe here. We also correlate the table(seat) turnover rate with time of day and proportion of 5- and 8- course set lunch/dinner.
What is the outcome:
The average turnover rate has improved by 33% and the proportion of loyal customer (who visit the restaurant at least once a quarter) increased by 41%.
Creative Pricing
What are we trying to solve:
An European restaurant offering 5- and 8-course set lunch/dinner would like to know how customer is choosing the items within each of the course category and use this information to optimise the menu and also, to turn their business back to growth.
How are we solving it:
The POS sales data includes: 1) Items ordered by each individual, 2) Time sitting in and checking out, 3) Number of people per table
​
A simple visualisation (BI) tells us that 8 course is more popular during dinner time and weekend, where customers have more time to sit down and enjoy the companion of their family and friends. Certain items are popular for 5 course and 8 course while some are all time favourites
​
We also ran an Apriori (Market Basket) model to and observe that there are 2 distinct group of ordering patterns - A more flavourful segment of customer and the other who prefers healthier combination. This would help them to understand how customers are interacting the menu and how they could further customise and change some items in their menu
Here, we created 2 other indicators, seat turnover and table turnover rate. This would tell us 1) How packed the restaurant is and is the restaurant efficient and maximising it's capacity (or seats) for the customers. Obviously, there are so much room for improvement we observe here. We also correlate the table(seat) turnover rate with time of day and proportion of 5- and 8- course set lunch/dinner.
What is the outcome:
The average turnover rate has improved by 33% and the proportion of loyal customer (who visit the restaurant at least once a quarter) increased by 41%.