Customer Data Platforms

Use People Data to Transform the Future of Marketing Engagement

Martin Kihn, Christopher B. O'Hara

• Éditeur: Wiley
• Année: 2020
• ISBN:
  9781119790129
• Édition: 1
• Sujet: Negocios y empresa
• Sous-sujet: Relaciones con los clientes

Master the hottest technology around to drive marketing success

Marketers are faced with astarkand challenging dilemma: customers demand deep personalization, but they are increasingly leery of offering the type of personal data required to make it happen. As a solution to this problem, Customer Data Platforms have come to the fore, offering companiesaway to capture, unify, activate, and analyze customer data. CDPs are the hottestmarketingtechnologyaroundtoday, but arethey worthyof the hype? Customer Data Platforms takes a deep dive into everything CDPso you can learn how to steer your firm toward the future of personalization.

Over the years, many of ushave built byzantine "stacks" of various marketing and advertising technologyin an attemptto deliver the fabled "right person, right message, right time" experience.This can lead tosiloed systems, disconnected processes, and legacy technical debt.CDPs offer a way tosimplify the stack and delivera balanced and engaging customer experience. Customer Data Platforms breaks down the fundamentals, including how to:

• Understand the problems of managing customer data
• Understand what CDPs are and what they do (and don't do)
• Organize and harmonize customer data for use in marketing
• Build a safe, compliant first-party data asset that your brand can use as fuel
• Create a data-driven culture that puts customers at the center of everything you do
• Understand how to use AI and machine learning to drive the future of personalization
• Orchestrate modern customer journeys that react to customers in real-time
• Power analytics with customer data to get closer to true attribution

Inthisbook, you'll discover how to build 1: 1 engagement that scales at the speed of today's customers.

CHAPTER 1
The Customer Data Conundrum

Nobody but a farmer wakes up one morning and decides to build a silo – yet that's exactly what has happened, naturally, over the past decades. Starting with the best intentions, marketers and other divisions acquired applications and built customer data stores, then built patches and organized data lakes and marts, signed up with exciting start-ups and … ended up with over a dozen (on average) separate databases storing data, often about the same customer. Meanwhile, organizations have built up around these silos, making the problem much worse. In this chapter, we discuss the current state of (disconnected) marketing and why it needs to be solved.

DATA SILOS

What's keeping marketers from achieving the “right customer, right message, right time” nirvana they've been chasing for decades? Put simply, it's the nature of customer data itself – it's siloed in different databases, stored in different formats, and used by different parts of an organization in different ways. It's constantly growing, pervasively available, and getting accessible in real time, but continues to defy brands' efforts to unify it and make it easily actionable.

Salesforce's sixth State of Marketing report surveys over 7,000 senior marketers across a wide swath of industries to find out what their top priorities and challenges are from year to year (the survey is conducted by a third party, and respondents are not aware that Salesforce is the sponsor of the research). In the most recent edition, two of the top five challenges were “unifying customer data sources” and “sharing a unified view of customer data across business units” (Figure 1.1). In this section, we will talk about why data unification continues to defy marketers' efforts – and discuss the challenges and opportunities for sharing customer data across your organization.

FIGURE 1.1 Marketers' top challenges.

Source: Courtesy of Salesforce State of Marketing 2020.

Does this experience sound familiar? You start to carefully research new cars online, you “build” a few models on the website, research pricing and financing – even go to several dealerships and take test drives. After a few weeks or months narrowing down your choices, you pick the right vehicle and drive it home. Then for the next two years you get consistently blasted with emails, social media ads, and mobile ads for the car you already bought. What a tremendous waste of money! Despite all of the technological innovation driving marketing and advertising over the last 20 years, it seems like every brand who smells even the smallest whiff of purchase intent immediately starts a barrage of full-funnel marketing meant to overwhelm consumers into buying something.

The problem is not a lack of data. In fact, we have seen that marketers report using a median of 8 distinct data sources in 2019, and will expand that to 12 next year, a 50% increase. Think about the amount of data generated by a consumer in the car shopping process: online cookies generated through website visits that reveal the type and model of the car; web form data with name, phone number, and email address gathered from test drive request and finance forms; a user's mobile advertising ID from in-app experiences; and lead form data input at the dealership.

Put together, these data would reveal intent across the entire lifecycle of a car purchase – with attributes ranging from vehicle type, price range, color, location, financing type (buy or lease), and even past purchase history. With this type of data, a smart automotive marketer could move consumers from consideration to purchase in stages, close the deal – and continue the campaign post-sale with service deals and offers.

However, this rarely happens today. The problem is not the amount of data being generated, but where that data is stored and who is using it. The data from the website is not connected to the customer relationship management (CRM) system at the dealership. The “take a test drive” form never makes its way to the web marketing team to improve their targeting. The purchase data from the dealership and manufacturer's warranty data stay in silos and never get used to update the marketing team to prevent the company from trying to sell you a car you have owned for six months.

Over the years, we've built up a number of data silos, where critical customer information never gets the opportunity to enrich each other and lead to insights. How much money could a marketer save by simply turning off marketing for products a customer has already purchased? By using the carefully tuned buying propensity models from known customer data and applying them to unknown prospects coming to the website? By connecting what's happening on their e-commerce site to their email marketing? It seems obvious, but the problem of disparate, siloed customer data has only expanded and hardened over time.

Defining the problem space is simple: too many types of customer data, stored in different systems. But to start to solve it, we must look at the problem in two distinct ways, defined by fundamentally different data types: “known” and “unknown” customer data (Figure 1.2).

FIGURE 1.2 Data silos, organized by marketing function, across known and unknown data types.

KNOWN DATA

For simplicity, “known” data is any type of customer data that is personally identifiable, called “PII” or “personally identifiable information.” This is where you fill out a web form, purchase something from a website, give the cashier your information at a retail store, subscribe to a site like the New York Times. This is the virtual gold of marketing – real information about real people that has been given with consent to a company you trust. It is increasingly rare thanks to privacy legislation (more on that in Chapter 5), and very expensive to obtain. In 2019, Mary Meeker's Internet Trends Report called out the increasingly high cost of customer acquisition as traditional brands start to compete with AI-driven direct-to-consumer startups, such that a customer's lifetime value has started to become less on average than the cost of acquiring them.

Depending on the industry, acquiring a new customer can cost anywhere from 5 to 25 times more than retaining current ones – and the cost will continue to rise as marketers attempt to stay afloat in a noisy digital marketplace. Two years ago, Comcast paid over $1,200 for every net-new wireless subscriber. The cost to purchase email can range anywhere from $200 to $600 per thousand addresses (CPM, or cost per mille), depending on the accuracy and granularity of the list. Real “people data” is expensive, and most companies have put a lot of energy into organizing and optimizing it, starting with its use in traditional CRM systems by sales, email systems by marketing, and service applications by call center employees.

The problem, however, is that even when the brand has a customer record, it lives in those silos (sales, service, commerce, marketing) and rarely gets unified in such a way that can power a customer journey. A typical company may use one CRM application to power its sales operation, another to keep track of customer service details, and yet another system to store data used for marketing. In addition, they maintain a data warehouse or data lake to create a “golden record” of customer data, and use it for things like propensity scoring and lifetime value (LTV) modeling. But a customer record may be replicated four times across those four systems – and the company may have multiple instances of data across different regions, brands, or operating companies.

Imagine how great it would be if Joe Smith in the service system (Joe returned the shoes he bought) could be unified with the Joe Smith inside the marketing system (don't email Joe about these shoes)? Or the Joe inside the data warehouse (Joe has a high propensity to buy) could be connected to the sales system (call Joe now)! This seems like a fairly straightforward problem to solve – why not just create a single system with one record? As above, most “stacks” are like old rambling houses, featuring additions built over many years – they're almost impossible to renovate.

CUSTOMER RELATIONSHIP MANAGEMENT (CRM)

The CRM system is the operating system for customer data that the sales organization plugs into, and is in many ways the organization's true “source of truth,” going well beyond sales records. The service system in the call center is where tickets are opened and logged, and call center reps own a relationship with customers and prospects that may start on Twitter, go to the phone, and end up on email or text message. That system is the operational heart of the call center, and the “source of truth” for how customers move through specific parts of the funnel. Marketing and commerce systems are their own closed ecosystems, complete with different ways of identifying customers, storing their data, and analyzing it. Customer data is the lifeblood pumping through those systems and powering their operations and, over time, those systems have evolved to leverage customer data in the service of different outcomes: revenue (sales), customer satisfaction (service), direct purchase (commerce), and engagement (marketing).

Breaking down those silos requires companies to resolve cus...