Giter VIP home page Giter VIP logo

traveller-ansights-analysis's Introduction

Discovering Instagrammable Attractions in Switzerland: From Data Collection to Data Visualization

Developed end-to-end data pipelines to address decreasing online conversion rates for Swiss online travel agencies

"Travel is complicated to book", Expedia(2015). Booking a trip is a hassle for many travellers as the booking process involves various products, making the customer journey even more complicated than in other industries. To capture traveller preferences through digital touch-points and help online travel agencies to improve conversion across travellers’ digital journeys, a data lake and a data-driven traveller insight dashboard are created aims to help OTAs optimize digital experience and marketing strategy. In this project, the Design Thinking methodology is adopted to define the use case, persona and critical touch-points to establish a data collection strategy. A series of Data engineering includes planning data architecture, constructing data pipelines and fetching data using open APIs deployed for building a traveller database. Following that, data manipulation and data analysis were conducted to generate traveller insights, and finally, a dashboard is to address decreasing online conversion rates for Swiss online travel agencies.

Project Requirements

  • To build an end-to-end data pipeline and make it available for business analysis in a data warehouse environment
  • Use at least 2 APIs to obatian static and non static data
  • Present a prototype use data visualiation tools

Background

According to Expedia(2015), travellers visit an average of 38 websites before booking a trip. The entire trip booking process involves several steps and can be stressful for travellers, this is because there are too many products and options that occur during the search phase, and that puts travellers in the paradox of choice and make the customer journey even more complicated than in other industries.

Problem

Websites that provide travel-related products have a 98% abandonment rate among all businesses. Statista (2022)

Despite online travel agencies still playing an important role in the customer journey, OTAs are suffering from a low conversion rate as Statista reveals that travel-related websites received the highest shopping cart abandonment rate among other consumer goods sectors. This is because are too many things involved in the booking process, and this will affect the traveller's decision. As consumers spend more time searching for trips, they will take longer to decide whether to book. This means that the OTAs will have less impact on consumers' buying decisions.

Solution

When mapping out a traveller's digital journey, we found social media has gained importance in the traveller's decision-making process (Smith & Anderson,2018). Travel feeds on Instagram have become one of the powerful indicators that trigger travellers’ decisions to plan where they should visit on upcoming trips (Han & Chen, 2021). To understand how social media and online search impact travellers' buying decisions and to help Swiss OTAs improve online booking experience and effective marketing, we designed a traveller insights dashboard that presents crucial metrics and trends by analyzing social media and search engine data.

Methods

A travel database is built with the following data engeneer techniques:

  1. Define Data Collection strategies: Define the scope of the project and collect data from across traveller digital touchpoints
  2. Design Database and Construct Data Piplines: Present a business data schema Construct ETL and ELT workflow use Apache Airflow
  3. Built Data Architecture: Demonstrates data system with AWS RDS according to business needs
  4. Prototyping: Based on predefine business questions to design visualiation datashboard layout

Data Architecture
Data Architecture


Data Source

Data Source Description
Instagram Graph API 1. Need Facebook & Instagram Business accounts (for an access token).
2. Access information to public account users such as @mention, #hashtag, metadata(no. of likes, comments)
3. Total requests cannot exceed 200 times per hour per user, with a maximum of 50 responses a page (25 responses for one API call).
4. A maximum of 30 unique hashtags in a 7 days timeframe
MySwitzerland API 1. A public API that provides swiss tourist content via MySwitzerland website.
2. Access information name, an identifier that refers to a unique ID of each destination, and corresponding geo-location data.
3. 4114 destinations stored in 412 pages
4. 1 request per second and 10 returns in a burst
5. Set time. sleep(2) to avoid blocking by the system
die Post Download the “PLZ_Verzeichnis” CSV file via the link
It provides Postleitzahl, Ortbez18, and Kanton, and information of each city

Result: Proposed Prototype

I constructed data pipelines to address the business problem that presents critical metrics with our proposed traveller insight dashboard. The IG Hashtag Statistic graph gives information about hashtag usage in each Instagram post. We can answer the following questions:


Business Questions 1: How can we improve the tourist travel experience in Switzerland by utilizing social media and search engine data?

  1. What are the popular destinations in Switzerland that people search for travelling?
  2. Which city do people are interested in the most in Switzerland?

Business Questions 2: “How do individuals and businesses utilize social media hashtags to boost their online presence and performance?”

  1. What are the popular hashtags (top 5) used in Switzerland?
  2. What are the popular hashtags (top 5) based on Swiss canton, city, or destination names?

IG Hashtag Statistic
IG Hashtag Statistic


Posts with the hashtag Zurich receives the most likes, which has 169’235 likes and 6’170 comments; in the same period, Valais and Luzern are the second and third popular place, with 144’957 and 120’267 likes respectively. Following that, the doughnut chart shows Zurich, Wallis and Uri are the most popular cantons, and Zurich, Zermatt, and Interlaken are the most admired cities. Therefore, we can conclude that Zurich is the most popular and most visited place in Switzerland.

A full dashboard can be seen in Tableau Public

Limitation

Initially, we wanted to get location tags to create a location-hunting app. However, due to privacy issues, we couldn't access location tags because Instagram deprecated the local tag endpoint from its API. Therefore, we decided to use #inLoveWithSwitzerland, a hashtag used by MySwitzerland, and 26 Kantons' names in English as the basis to fetch relevant hashtags in each Instagram post. Although we couldn't capture all the locations in Switzerland, the collected information was sufficient to create a minimal viable prototype for this project.

Conclusion

The entire process of constructing a database, from data collection to visualizing data, helps us to generate traveller insights and to answer our business questions. The proposed dashboard focuses on value creation for business providers such as online travel agents and marketing firms, so they can utilize matrics from this dashboard to provide travellers with more personalized services and products.

Project Reflection

This project taught me many hard skills in data engineering, which equipped me with fundamental skills and knowledge to interact with database tools and allowed me to combine design thinking with product development. Besides, having the opportunity to use popular software such as AWS and Airflow to realize our business idea into a monetizable product was an invaluable experience I had from this group work.

Code Files

Reference

traveller-ansights-analysis's People

Contributors

chitrug avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.