Analysing Airline Reviews

Introduction

The what

We’ll guide you through:
✅ Interacting with your graph
✅ Extracting key insights
✅ Understanding the data structure within your graph

The who

Whether you're a:

• Marketing analyst exploring travel patterns and passenger preferences to better create promotional strategies, or

• Operations expert looking to optimise logistics and supply chains for a more competitive offerings,

Conode empowers you with precise, actionable insights—straight from your data. 🚀

The data

This dataset contains British Airways passenger reviews from 2013-2023, covering:

✈️ Routes traveled
💺 Class of travel
👥 Group vs. Solo travel
🎯 Primary purpose of travel

The original data can be sourced from Kaggle here.

The how

In this workflow, you’ll:

1️⃣ Explore a pre-built graph

2️⃣ Run sample queries to extract insights

3️⃣ Dive into the graph structure and learn how to interact with it

The why

🤵🏻 To understand passenger satisfaction and identify pain points to improve overall customer experience.
🛫 To optimise and inform flight routes, scheduling and resource allocation.
😁 To identify customer preferences, set up targeted marketing campaigns, and improve customer retention.
🛎️ To improve in-flight services, amenities, and offer competitive product offerings.

Why the need for a connected data approach with Conode?

Traditional methods like simple reporting or basic data analysis in spreadsheets struggle to answer these questions because they require:

• Connecting data from disparate sources: The questions require linking information from reviews, passenger data, flight data, airport data, and rating data.
• Analyzing unstructured text: Understanding the nuances of customer feedback requires analyzing the text of the reviews.
• Identifying complex relationships: The questions often involve identifying correlations and patterns between multiple variables.
• Speed and agility: Business users need to be able to get answers quickly, without relying on lengthy data analysis projects.

Try with us!

We’ve included this dataset as an example graph in Conode, so you can jump straight in and follow along as we query the graph, produce charts and cross-highlight to analyze the reviews. If you are a new user, please follow the simple directions to create an account first and then feel free to utilize Conode to your content!

Hop into Conode

Approach

Loading the graph

1. Login to Conode. Sign in or create a new account.
2. Click on 'Example Graphs' and select 'Airline Reviews'. A 'Home' view will be the first thing you see, this is the schema of the data.
3. Navigate to 'Ask Graph' on the top right corner. This is where we will ask our graph questions.

Ask Your Graph

1. Although the title is quite telling, let’s say you still want find out what the dataset contains. Use the prompt question: What is this graph about?

2. Next let's get some statistics with this typed in: How many reivews do we have? How many passengers recommned their flight and how many do not?

3. And why would passengers not recommend their British Airway flight? Let's get some answers using: What are the main reasons for passengers not recommending their flight?

4. Let's take a look at a specific review: Show me a review from a passenger who complained about Customer Service.

Generate Charts 📊

Time for some visuals!

1. You would think that passengers in first class would be the happiest and give the highest ratings, so let's test this hypothesis with: Does the class of the traveller influence his/her flight rating?
2. Let's view the results in a chart. If the agent does not automatically generate one, you can always prompt it: Put this in a single chart view.

Question the Graph Alongside the Data 👀

You can go in whichever direction your analysis aims to. For this workflow, we're interested in lounge-related reviews.

1. Let's ask: Can you get me 20 reviews of passengers flying out from all London airports which mention lounge? Put them all in a view.

2. Using the right-side agents drawers to continue our conversation, we ask to: Show me a map of all destinations passengers were travelling to.

Interact Directly with Data to Uncover Patterns 🖍️

An advantage of using Conode is the ability to have a single unified view of our data with the graph structure. In this scenario, it’s a perfect opportunity to see datetime, destinations and section ratings information for this specific 20 reviews.

By selecting and cross-highlighting, we can immediately tell that all reviews from the 20 passengers flying out of London airports and mentioned lounge:

• ⭐ All but 1 were made before May 2017.

• ⭐ None of them sat in premium economy.

Summary

In this session we have:

• Queried our graph
• Generated charts
• Investigated a subset of reviews based on our interest
• Selected and cross-highlighted our data across multiple views

The analysis we’ve done here can be expanded upon to answer deeper questions like "what are the emerging trends in customer complains over the past year, segmented by traveller type (business, leisue, family)", and cover areas such as product development, marketing & sales department, operations and customer service department.