Creating a 360-degree customer view in a COVID-19 world

Airlines have always dreamt about having a centralized view of their customers to understand interactions across all touchpoints and be able to deliver rich, value-creating experiences. 

However, delivering this unified customer experience with personalized recommendations, content or propositions across many devices and channels requires a true understanding of who the customer is and how and when to reach them.

Today, there is the perfect storm which creates a real urgency for a 360-degree customer view: COVID-19, the primacy of privacy, the post-GDPR landscape, multiple device and channel usage and the explosion of messaging technologies.

This perfect storm forces airlines to finally address the issue of identity resolution. To know if you have a complete view of your customer, ask yourself these questions:

• Can your airline reach your target customers across all devices, browsers and touchpoints?
• Does your airline know what channels a customer used before they bought from you?
• Can you attribute the purchase made by your customer to the right channels?
• Is the experience your airline customer receives seamless and consistent, no matter what platform or device they use to connect with your airline? 
• Does your airline know who the customer is if they call through to the call centre? 
• Instantly recognising passengers whenever they engage on any digital touchpoint or channel is hard. So, where should airlines start? 

Creating your 360 View: The challenges 

The first step in the journey is to create a single-customer-view. Creating a single customer views starts with data. This data is gathered by tracking customer interactions, understanding customer history and preferences and predicting customer behaviour.

Examples include Passenger Servicing Systems (PSS), eCommerce systems, mobile devices, browser cookies, Customer Relationship Management (CRM) systems, loyalty systems, digital marketing systems and, indirectly, via external systems such as social media platforms. 

The first problem with this sort of data is that it is ‘trapped’ in these disparate systems that weren’t designed to share it with anything else. 

The second problem is that combining customer data from a disparate set of data sources is difficult enough, but what is lacking is a common, unique key that links all records that belongs to an entity. 

Another problem is the implementation of California’s Consumer Privacy Act and other emerging regulations across the U.S., European Union, Brazil, and India, have created both new expectations about data privacy and also a conundrum.

In this new environment, airlines must give complete clarity in terms of how data is collected—and consumers are in control of their information and how it’s used and shared. 

Finally, on top of these data privacy concerns, Google’s announcement that it plans to start blocking third-party cookies has created real problems for digital marketing – as this was the basis of many digital marketing technologies. 

Satisfying each of these problems poses new challenges for airlines – and this can be solved with identity resolution.

The importance of customer identity resolution

The goal of identity resolution is to provide a single view of each customer across multiple channels – at the same time as mapping to the new world of privacy law and customer expectations. 

To understand the importance of identity resolution, let’s look at a common airline passenger use case.

Think of a passenger who books flights or hotels on an airline’s e-commerce platform: they can use different combinations of personal information, using a passport as an ID when flying internationally but also using a driver’s licence or national ID when flying domestically. A passport or national ID is as close to a unique identifier a person can have. 

In most airlines, these two flights will not be considered as being purchased by the same person. 

Let us look at a few other scenarios: 

1. An airline customer visits their website from his desktop, checks their business flight departure time, and later starts browsing holiday destination pages on the site. 
2. The same airline customer visits their website anonymously on his mobile without logging in. They check out the cost of checking in a bag and drops off the site.
3. The same airline customer visits the app to check-in for their business flight tomorrow. 

In Scenario #1, as the user has not logged in, the airline cannot easily identify the user unless the browser cookie history is still intact. They only know an anonymous user’s visit and his activity on their website through a mobile device. 

In Scenario #2, if the customer hasn’t cleared his browser’s cookie history, the airline will be able to check if the user had previously visited their website earlier using his cookie ID. If the customer had previously logged in from the same cookie ID, the airline will be able to know who the user is exactly. If the customer had used another browser to browse the bank’s website, the airline will not be able to guess the identity of the user. 

In Scenario #3, the airline can immediately identify the passenger. 

In the above scenarios, true identity resolution is obtained if the airline is able to analyse and arrive at a conclusion that the user in all 3 scenarios is the same. 

These scenarios explain why identity resolution is required, as it provides an approach to linking entities from disparate data sources together. Identity resolution also links transactions within one system together.

Why is identity resolution difficult? 

Conceptually, identity resolution is very simple to understand. However, it is notoriously difficult to implement due to a myriad of factors.

For example, many of the airlines we work with at OpenJaw transport over 50 million passengers a year. When you collect this data over several years, it may resolve to over 100 million unique customer profiles.

When you also consider the multitude of source systems that this data can originate from, beyond core systems like the PSS systems, volumes are very large. The velocity at which this data is generated is also a challenge, as well as data quality issues (missing values, inconsistent data input) and schema variations across systems in the travel ecosystem. 

Another reason why identity resolution is so difficult to implement is that consumers using multiple channels before they make a purchase, and using multiple devices interchangeably meaning it has gotten exponentially more difficult to understand the complete customer journey. 

Identity resolution helps airlines connect the dots across all online and offline touchpoints that each individual customer uses.

Without such capability, airlines risk wasting customer experience, risk of privacy and compliance issues and was and particularly over or under communicate with customers at key times – such as right now in the middle of a pandemic when customer are looking to know even if it is safe to travel. 

The question for airlines is how confident that their customer ID profiles are both complete and accurate.

The building blocks of ID resolution 

The goal of identity resolution is to deliver an improved customer experience with a high level of personalisation and continuity of voice across all touchpoints and channel, whilst meeting privacy and compliance obligations.

Or, more simply, to know enough about your customers so that they feel like you know them, understand them and want to help them. 

Here at OpenJaw, we believe there are four building blocks to a strong identity resolution strategy:

• Recognition: Use data not only available internally but also looking at behaviour across all channels
• Accuracy: The importance of identifying customers across all of their devices and browsers, consistently and at scale, for a better customer experience. 
• Persistence: Being able to follow the customer as they change along their customer journey with updated data actions both online and offline. 
• Privacy: Data must be scrubbed of personally identifiable information before being used to comply with the European General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA). This step also involves ongoing management of opt-out status. 

Identity resolution capabilities

Customer identity is about more than just linking different data. It’s a way to fill in the blanks around your customers. It’s not the end game in and of itself, but an important tool in the arsenal as an airline seeks a better understanding of its customers.

There are many technical challenges associated with identity resolution. 

Identity resolution capabilities typically involves identifying customers across devices, controlling communications frequency and sequencing, or building a single customer profile using all available customer data points. To this means a very involved process to combine large data sets, multiple devices and the right privacy infrastructure. 

Identity resolution capabilities can mean using deterministic and/ or probabilistic matching to resolve customer profiles. What is the difference? 

• Deterministic Matching: In this method, the first-party data of customers is analysed and matched to customer records using identifying variables such as username, email, phone number, etc. 
• Probabilistic Matching: This method is used when first-party data is limited and uses identifiers such as device type, browser type, IP address, OS, etc. Multiple customer identities are matched using the identifiers through a statistical estimate. 

Our experience working with large airline datasets at OpenJaw showed that probabilistic matching greatly outperforms deterministic methods.

Extensive experiments conducted on publicly available data and private airline customer data that compare the OpenJaw Probabilistic Matching identity resolution algorithm performance to traditional, deterministic matching confirmed this. 

Recommendations for airlines 

Creating A 360-Degree Customer View in a COVID-19 world always starts with identity resolution. Building a single customer view of your data with identity resolution means linking customers records across data sources in a probabilistic manner.

Linking customer information across data sources leads to more comprehensive customer profiles, allowing greater personalized customer experiences. 

OpenJaw recommend the following best practice for airlines looking to creating a 360-Degree Customer View:

• Leverage broad range of identifiers: There is a need to look beyond using just email addresses and authenticated data as the basis of identifiers. Probabilistic Matching gives the best results.
• Focus on early wins: The earliest benefits of identity resolution usually revolve around reduced waste. Better customer profiles improve targeting and uses attribution to determine the communication mix.
• Understand the full benefits of identity resolution: Once early wins have been achieved, don’t stop there. Beyond targeting, take advantage of the benefits across the customer journey and all touchpoints.
• Enlist stakeholder support: Because of the scope of a well-executed identity resolution programme address privacy and legal issues, all interested parties (top management, IT, customer service, marketing, commercial) must have an understanding of all components of the programme.
• Select partners wisely: Since an identity resolution program involves customer data and privacy ramifications, selecting the right partner is critical.

Start your customer data initiatives with OpenJaw Identity Resolution

The starting point of any 360 Degree customer view, airlines need to perform Identity Resolution on your source customer data to eradicate inconsistencies and duplication. 

OpenJaw Identity Resolution links customer records across data sources in a probabilistic manner to create a single customer view.

Linking customer information across data sources in this way leads to a larger number of comprehensive, accurate customer profiles drive a solid foundation for better personalisation and compliance. 

The performance of Blueprint Identity Resolution has been benchmarked against traditional ‘exact matching’ methods used to integrate customer records. These experiments show that even the simplest configuration of Blueprint Identity Resolution can improve the customer record match rates from 49% to 70%+. 

OpenJaw Identity Resolution is a sophisticated algorithm that uses a combination of probabilistic (‘fuzzy’) matching algorithms, graph theory and machine learning to resolve identities across very large (100M+ records) customer data sets. 

The OpenJaw Identity Resolution algorithm uses an intuitive GUI to allow configuration, such as choice of matching thresholds for each individual field, as well as the choice of matching rules that decide if two records belong to the same customer.

The algorithm uses fuzzy matching to quantify similarities between fields and graph algorithms to link records together to create customer profiles via a unique ID, called the OpenJaw ID. 

OpenJaw’s Identity Resolution is highly scalable, leveraging the power of AWS Elastic Map Reduce (EMR), along with Apache Spark.