E-commerce business is getting more and more competitive every day. It is crucial to understand what customers want. In addition to data, which is already being collected from customers’ online behaviour, we decided to introduce a new way to track users’ online shopping experience to analyse indirect insights – by tracing their eye data. This way companies can make deeper marketing research on their web design, accuracy of ‘recommended’ items, product range as well as brand selection.
What it does
User installs Eyes extension to chrome browser. As soon as he or she opens a website, which is connected to Eyes, the app starts to track a user’s eyes motion: where their look and for how long. This information can be correlated to clicking and buying activity on the website. For example, if a user was looking on a particular item, but did not end up clicking on it to see more details, this product can pop up as a suggestion or special deal later specifically for this user. Moreover, if there is a repetitive tendency in overall behaviour of majority of users (e.g. ignoring certain products or parts of the shop), it could be a hint for redesigning the website. Collected data is accessible for business owner as a heat map on top of the web page of the shop.
How we built it
We used WebGazer.js library to infer the eye-gaze locations of web customers on a page in real time. This data is stored in MongoDB. While the users is on the web site, we also take a screenshot of where they are looking with selenium driver. Next eye data gets converted into a heap map using D3.js library (the darker the red colour is, the more time a user looked at that area). This map is put as a layer on top of the screenshot and renders on a Eyes dashboard page of the business owner.
Challenges we ran into
The main challenge was to connect the front and the back-end sides of the app. Also figuring out how to convert eye data into heat map was tricky.
Accomplishments that we're proud of
We managed to use and implement a combination of technologies that none of us has ever used before. Although there are some bits left to fine-tune, we are very proud of how much we have learnt and made work in only 24 hours.
What we learned
Sleep is important
What's next for Eyes chrome extension
In the future Eyes functionality can be extended by implementing machine learning by Amazon recognition. This way not only eyes motion tracking will be available, but also customers emotions. This way business can do the sentiment analysis.