According to new research from McKinsey, fast-growing companies drive 40 percent more of their revenue from personalization than their slower-growing counterparts, evidence that the value of getting personalization right – or wrong – is multiplying. Deeming it a business necessity, the study reveals that roughly 70 percent of consumers consider personalization a basic expectation – understood as a business recognizing them as an individual and knowing their interests.
The link between personalization and revenue helps explain the deepening interest many organisations have with securing a customer data platform (CDP), ostensibly for gathering all sources of customer data to better understand their customers. Yet in speaking with many companies about their underlying business challenges, I find that few have a solid grasp of the true problem they’re trying to solve. Many think of a CDP as little more than a synonym for a marketing automation platform or data centralization tool, and they harbor a misguided notion there is very little in the way of differentiation; any deployment will magically deliver a personalized customer experience, and the company will be off and running toward newfound riches.
A Data Quality Issue
In probing potential business use cases that these organizations have for a CDP, what is often the case is that while personalization may indeed be the desired outcome, there is little recognition that what is often holding them back is really a data quality issue. At its core, it’s not about data aggregation or automation per se, but about identity resolution. Without a good idea of who your customers are, you’re not only missing out on revenue opportunities, you’re probably also introducing loss.
How many unnecessary emails are going out because of duplicate records, or because household dynamics are not accounted for? Is a retention strategy off kilter because thousands of customers who make frequent purchases have multiple customer profiles, so they’re perhaps offered promotions they don’t need, or they’re not treated as loyal customers? Marketing spend, marketing strategy and reporting all suffer adverse effects when data quality issues impede identity resolution. If you have 20 million customers, what will a 5 percent inaccuracy rate cost the business with 1 million customers potentially receiving an irrelevant offer, email or other content, or being marketed to in juxtaposition with a customer journey? Average customer spend will offer a guide, as will churn rates and acquisition costs.
Perfected Data & Identity Resolution
Unfortunately, many companies accept such inaccuracies as a cost of doing business. Introducing friction into the customer journey of one of every 10, 50 or 100 customers with an irrelevant message, content or offer is understood to be the price to pay for offering a personalized customer experience for the majority.
The reality is, it doesn’t have to be this way. Perfect data is possible, and advanced identity resolution underpinned by probabilistic matching techniques is what separates customer experience platforms from those that claim the CDP mantle but more often than not fail to do more than simply ‘unify’ data. Probabilistic matching techniques account for the fact that identities are not static, making continual updates as more customer data becomes available to build on and improve a customer record. For example, there are roughly 250,000 marriages in the UK every year, which of course influences householding.
Probabilistic matching approaches ensure the creation of a layered, accurate profile that accounts for and recognizes the customer’s relationship with the brand over time, and how that relationship evolves across the complete customer lifecycle. CDP’s that purely ‘unify’ data and have no advanced resolution capabilities may render a ‘profile’ but fail to create true identities that provide the foundation and customer understanding required for personalization and effective customer engagement.
From a marketing standpoint, a lack of context limits opportunities to create deeply relevant, hyper-personalized experiences that are not only in synch with an existing customer journey, but that also demonstrate a recognition of an ongoing relationship with a customer. For a simple example, consider the dynamics of a household. A simple match of a device to an IP address may tell a brand that John Smith is browsing the website. But perhaps Mr. Smith is browsing products for his wife or child. Or maybe Mrs. Smith has borrowed the device. For a financial institution, having an up-to-date, accurate customer record that includes household dynamics is vital for ensuring Mr. Smith is presented with the right product and messaging. Does he have children about to go to university? Is he newly separated? Downsizing his home? Or are he and his wife expecting a child and interested in building an extension? Another possibility is that Mr. Smith is browsing for products or engaging with the financial institution on behalf of his small business.
A fleet management or car rental business provides another example of the importance of having an accurate, persistently updated customer profile. If the company knows, for instance, if a customer is renting a vehicle for business vs. pleasure, the home page may render an image of a sleek Mercedes pulling up to an office car park. Conversely, it might display a happy family piling into an MPV.
Similar to the McKinsey research, a recent survey Harris Poll conducted with Redpoint shed light on how important it is for a brand to possess – and act on – a detailed, contextual understanding of a customer. In the survey, 69 percent of consumers said that the pandemic has made it even more important for a brand to know their individual needs and preferences, and 65 percent said they now consider personalization a standard expectation – within 5 points of the McKinsey survey.
The persistent matching of keys at data ingestion vs. customer data that is already keyed is the difference between a brand making decisions with the dynamic flexibility that aligns with today’s dynamic customer journeys that consist of multiple physical and digital channels.
Data matching at the pace of the customer adds vital context to interactions that is a cornerstone for providing an omnichannel CX, irrespective of either the volume or variety of engagement touchpoints. Done well, advanced identity resolution reconciles records across all types of data – structured, unstructured, semi-structured, batch, streaming, etc. – and all data sources.
A comprehensive customer understanding is necessary to deliver an omnichannel CX in line with customer expectations. Taking shortcuts at data ingestion by using data that is already keyed will, in the end, short-change the customer. With a personalized customer experience shown to drive revenue, the era of having to tolerate inferior data is over. Customers expect and deserve nothing less.