Experience with Adobe Analytics
I’m pretty sure Neel gave a sublime overview of the first phase of the competition. If you haven’t checked it out yet then here’s the link to Phase 1. https://bit.ly/gdsc-vjti-neel
The Next Phase
The competition from here on was relatively much more guided because now we had our hands on the documentation of various metrics that we weren’t previously aware of causing us to go astray and shoot in the dark. Most of the time we don't really think about how important it is to have relevant and clean data, we tend to directly jump into the analysis phase. We made the same mistake and surely lost a significant amount of time but nevertheless we were able to make it up in our further tasks.
Disney had provided us with business questions that we had to primarily focus on in this challenge.
Business Questions from Disney
- For those whose first purchase is during the period, what behaviours are most indicative of subsequent (or multiple subsequent) purchases? How do we encourage this?
- Looking at our cart, how do we drive the average order value up? Is that answer different on mobile vs. desktop?
- What key differences exist between a purchase flow on mobile vs. desktop, and how should we treat those platforms differently/the same?
- When a cart is abandoned, and a Guest comes back and purchases it; how do they come back, and from what Marketing Channels are most successful? How do we help drive recovering these carts?
Neel has already discussed the first two business questions, hence in this blog, I’ll be dealing with the next two questions.
Slides from Our Presentation
Here I’m focusing on how the purchase flow of the customer differs from website to app. This was achieved with the help of a Control Flow Diagram followed by a Fallout Diagram and noticing the patterns of where the most visitors are navigating to. Eg. In the case of an E-Commerce website, the flow would be like this:-
Homepage -> Product List -> Product Detail -> Cart -> Checkout Page
Even though the flow of purchasing an item remains alike, I noticed that a customer is 10x times more likely to purchase a product from Disney’s website rather than the App.
This slide further focuses on how to drive the average cart value up. As the title suggests, we bring forth another solution of suggesting the Disney App for download to a friend to avail a discount. According to the survey (the survey is still open if you want to check the questions), I conducted, 62.2% of users will recommend the app to a friend if it benefits them.
Also, the bar graph clearly shows how the maximum number of visits are referred while it's only a small chunk which directly visits falling under the miscellaneous category. (Y-Axis shows the no. of visits through referrer)
A small modification in the Disney website like we have shown here would make a major difference to drive the average cart value up.
The amount of revenue one might lose due to something like cart abandonment is massive. Hence here we are shedding some light on our fourth business goal, where we are providing a solution to curb the problem of cart abandonment.
In the cart abandonment graph, you can see a plethora of users who abandon the cart after adding an item to it, for whatsoever reason. (Y-Axis represents the occurrences) and this seemed to skyrocket in the year 2020.
We have crafted a remedy with the current dataset, where we analyze the occurrences of the user who has abandoned the cart along with the use of marketing channels such as email and notifications. Reinforcing an analysis of a user returning on receiving an email as a reminder (Y-Axis represents the visits through email for web) or in the case of an app, one can check the return through notifications. (Y-Axis represents the number of unique visitors for the app)
The timeline is a recommendation in case our 3R approach (Rundown, Remedies and Reinforcement) is implemented by Disney. Here the bubbles represent the average revenue month-wise (2019–2020). You can see clearly how the month of May seems to generate a very high revenue so it's better to launch any updated feature at this time as it leaves scope for improvisation because following this we have the month of December which produces the highest revenue.
Hence you can also see how we have taken seasonality under consideration and provided a development plan for the same.
Towards the End
Another solution amongst the surplus I would like to discuss in brief would be the Disney Reward Program where we provide rewards in case the user visits for the first time, or to avail cash points on certain occasions. There are so many minor flaws that the company can benefit from if fixed. And we believe that even the smallest change can cause the greatest impact.
Finally, we got our proposal reviewed by our mentor, Professor Suratkar who was very supportive and gave us the final green signal and with that, we submitted our project at midnight.
Overall it was an exhilarating experience! We gave a brief proposal on how to tackle the business problems along with providing solutions that make conventional sales soar! With all the efforts put in, and without attempting to blow our own trumpet, we genuinely believe that we did a decent job. Out of 500 teams, only 15–20 were allowed to advance further. We were butting heads with students from reputed universities like IIT, UC Davis, Brigham Young University and many more. We did have expectations but in the end, they were shattered.
It was a tough pill to swallow but as they say, trying will risk failure, but not trying will ensure it.
This was an experience that I and Neel are ever so grateful for. We learnt a lot about how customers and their actions on sites and mobile apps play such a vital role in promoting sales and identifying the brand’s flaws in marketing. Usually, as software engineers, our job is to primarily build and maintain such applications for the client, but now that we visited the commercial side of a brand, we got a deep understanding of how researching and tracking the users’ navigation throughout the session can make a humongous difference in promoting and increasing sales. Such research is beneficial to the development as well as the marketing team which collectively benefits the brand. It was a fun-filled ride.
Team Members
- Neel Dandiwala, you can connect with him via LinkedIn.
- Sahethi DG, you can connect with her via LinkedIn.
