A perfect match: financial services and customer data platforms

Rob Fuller, Matthew de Ganon on November 11th 2020

This is a guest submission by our partner Accenture. Matthew de Ganon is the Northeast Financial Services Lead at Accenture. Prior to Accenture, Matthew spent four years at Capital One leading product management for Digital Payments and Customer Understanding. Rob Fuller is the Customer Data Orchestration Practice Lead for Accenture Interactive, focusing on how brands collect, manage, and integrate customer data. In this post, they share how financial services companies can drive growth by rediscovering how to put their customers first.


Over the last few years and especially today, companies have had to rethink how they do business. For banks, this means rediscovering their original purpose of putting customers first.

Having worked with dozens of top financial institutions over the past six years and delivering digital solutions across industries for over twenty, both Rob and I have seen that customers and companies alike want the same thing: relevant experiences that build trust and increase customer loyalty.

A recent Accenture report found that 5% of banks’ revenue is at risk as millions of customers are persuaded by the transparent and tailored offerings of fintech entrants. Twenty million customers in the UK alone have opened accounts with neobanks, putting pressure on traditional financial service companies to defend their market share by innovating how they engage with their customers.

Goldman Sachs and Citigroup have established funds to invest in or acquire fintech startups. Others, like Bank of America, have embraced going digital with robust mobile banking services. A few, like JP Morgan Chase, are investing in data to build machine learning practices that deliver hyper-personalized experiences and improve security for their customers.

In this post, we share the infrastructure for good data and how our customers use that data to launch machine learning programs that earn customer's trust.

Laying the foundation for customer understanding

Zoom with margin

Customers have high expectations — they are becoming accustomed to technology understanding who they are, what they are doing, and how they want to engage.

To meet these high expectations, companies rely on rich data. Traditionally, it required a fair bit of data wrangling by engineering and data teams to cobble together systems and clean the data. Fortunately, a new breed of solutions, Customer Data Platforms (CDPs), have come to market. It provides organizations with a unified platform to capture and synthesize data across all customer touchpoints. A CDP plays a crucial role in reliable data capture, identity resolution, and data activation. It can resolve data quality blockers previously preventing the data integration and analytics functions from driving customer-centric and contextualized engagement.

For example, a global media customer wanted unified customer analytics to drive digital traffic, reader loyalty, and revenue across its portfolio of 36 brands. The company had built data pipelines internally, but each pipeline served individual brands resulting in dozens of unique data architectures and siloed data sets.

The walled-off tech stacks limited its ability for executives to measure enterprise-wide trends and performance. Instead of untangling its current in-house infrastructure, the company chose a CDP to speed up the unification process and scale with the growing needs and requests across its multiple brands and teams.

Investing in a CDP not only eliminated manual work to build and maintain data, but it also opened an opportunity for them to go beyond their current capabilities—unlocking audience segmentation to launch more targeted experiences.

The CDP implementation saved the company over $2m in engineering resources and hundreds of thousands of dollars in infrastructure costs.

Key capabilities of a CDP

Zoom with margin

While there is a standard set of functionality that qualifies a platform as a CDP, not all CDP solutions are created equal Some CDPs like Segment, for instance, prioritize collecting clean, consistent, compliant, and consented data to deliver reliable experiences.

When evaluating a CDP, some key functionality to look for:

  • Securely collects user engagements across the customer journey.

  • Consistently formats customer data to be used in downstream applications.

  • Manages user consent and privacy requirements.

  • Creates an accessible view of the customer across all business systems.

  • Activates data in preferred channels to create compelling, individualized user experiences.

Implementing a system that has a robust set of tooling for data governance ensures the data in your tools is accurate, consistent, and complies with internal privacy and security policies. By enforcing common data standards at the point of collection, a CDP like Segment guarantees you are feeding consistent data to your machine learning models for accurate personalization at scale. 

Delivering customer-centric experiences at scale

Zoom with margin

While a CDP can bring lots of functionality on its own, it’s critical to ensure other systems integrate with your CDP. Other systems can maximize a CDP’s functionality of real-time collection, identity resolution, and activation to create incredible customer experiences. 

For example, let’s say a customer tries and fails to pay a bill online. Within five minutes of that event he calls into the Interactive Voice Response (IVR). The system should receive that data and be aware of that online failure and adjust accordingly. Instead of, “How can I help you today?,” the greeting is, “Hi Isiah. I see you tried to pay your bill online. May I help you complete that task?” 

Isaiah then receives a follow-up message encouraging him to set up automatic payments. Isiah is happy because he no longer has to take the time to explain why he is calling. The company is happy as they can now deflect volume and provide a better experience through a self-service channel. Each of these data points allows the company to build a better understanding of its customer.

Customers want a personalized experience to match their context. Sending the wrong message at the wrong time, too frequently or not enough, can diminish trust in your brand. In an industry like financial services, trust is paramount. 

Launching a machine learning practice

The right data and systems coupled with a strong machine learning practice can deliver the right messages to your customers in the format and system they expect. Accenture’s research suggests that machine learning drives competitive advantage and will add US$1.153 billion in value to the financial services industry by 2035.

Clean and consistent data is critical for powering accurate machine learning models. Using a CDP as the key source for models, companies can define relevant features that drive the most accurate predictions, like how, when, and where a customer wants to interact with your business. 

For example, a new financial services entrant was able to increase onboarding activation from enrollment to direct deposit by 140% by leveraging Segment’s capabilities to ingest and activate data. Segment collected and sent hundreds of different user actions along the customer journey to its machine learning application. The application trained a lookalike model to forecast user intent and build cohorts based on propensity to engage in key milestones along the customer journey. Those cohorts were then sent back to Segment to activate in its web experimentation and email platforms. So now, when a user reaches a qualifying step, its marketing automation platform can send them an SMS, email, or in-app notification depending on propensity score to activate in the coming week.

With rich customer data powering machine learning models, this company can now personalize the message, channel, and even identify the best time to communicate for each stage in the funnel.

While clean and consistent data is key for machine learning, implementing a successful machine learning initiative requires more. Before kicking off your ML project, take time to:

  • Understand who your user is, what they need, and what delights them.

  • Identify ownership across key internal stakeholders.

  • Define the metrics and business goals.

  • Determine which areas of the user journey you will experiment with first.

  • Ensure agility from insight to action. How often will your team refine, stop, or build new experiences based on insights uncovered?

By first understanding your customer and then defining the metrics and experiments you want to run, you are ensuring you are optimizing the experience for your target persona while impacting the most important metrics for your business.

Looking to the future: Building trust to accelerate growth

Pre-pandemic, having the ability to deliver consistent experiences was differentiating enough. Now, with consumer behavior changing and financial pressure growing, consistency is considered the bare minimum.

Banking and financial services have the opportunity to drive a new narrative by building customer empathy into everything it does, or it risks customers moving their business to those that have adapted to their needs.

Rolling out a CDP such as Segment provides businesses with the flexibility and capability to support real-time personalized experiences that can build the trust that drives purchase intent. Doing so requires choosing the right CDP, applying the right governance, and the willingness to adapt or transform operational models to capitalize on the data and insights surfaced.

Are you starting a digital transformation project? To get started unifying your data, request a demo to speak to a Segment expert.

The State of Personalization

Our annual look at how attitudes, preferences, and experiences with personalization have evolved over the past year.

Become a data expert.

Get the latest articles on all things data, product, and growth delivered straight to your inbox.