Catching the unicorn: Integrating customer data into a single view
- Wednesday, June 12th, 2019
- Share this article:
Ahead of Teavaro’s upcoming webinar The Six Challenges of Personalisation on 26 June, Nico Pizzolato looks at the challenge of creating the single customer view.
The idea of personalising a service or a product in response to a customer’s need is as old as commerce itself, but in the past generation the internet, cloud computing and machine learning have given businesses an unprecedented opportunity to increase its reach and scope. When, in the noughties, Jill Dyché and Evan Levy tried to explain in a book (Customer Data Integration: A Single Version of the Truth, 2006) what customer data integration was like, this phrase had been already criticised for being a repackaging of previous notions such as Service-Oriented Architecture (SOA), Extraction, Transformation and Loading (ETL) or plain old Business Intelligence. But the criticism missed the point that the frontier of data analytics keeps shifting.
As the volume of customer interactions has surged over the past ten years, new processes and technologies have become available to retain and link customer data so that companies can engender trust in both prospects and customers and make better marketing decisions. However, at every turn, such technologies and processes, together with the jargon associated with them, have had a short lifespan or seemed insufficient to keep pace with the plethora of new data. Reaching a single customer view (SCV) has so far remained an elusive goal – chasing the proverbial unicorn – but at every turn the instruments available have become more sophisticated.
A SCV’s definition usually includes words like ‘holistic’, ‘accurate’ or ‘consistent’ to indicate that any successful effort to join the dots between various channels have to bridge the offline and online divide, break down data siloes, and improve data quality. For instance, in retail, this could mean bringing together e-commerce transactions, in-store purchases, contact information, browsing behaviour on the company website, call centre interaction and email engagement. In this case a SCV would be the foundation of campaigns that do not target prospects who are unlikely to buy or have just bought the product you offer – such personalisation failures cost credibility – while instead cross-selling or upselling to customers who might be near the end of a subscription, have recently interacted with the social media or haven’t made any purchase recently. The more integrated the view of the customer, the higher the return on marketing investment. But the multiple points of entry of customer data pose real challenges to existing ways and systems with which businesses have worked so far.
Customer Relationship Management (CRM) software stores data from customers’ direct interactions and keep tracks of how far along they have moved into the sales funnel, but is limited to transactions and communications with the company, usually not capturing information from other multifarious sources, such as social media, the website or the company app. CRMs stitch together information across some channels, but not across all of them. They also lack the machine learning power to unify data that might belong to the same customer but presents some discrepancies, for instance a misspelled name or address. As a system used by the whole company, it is not particularly oriented to marketers need; for instance, it does not provide further segmentation insights for personalisation.
Data warehouses, which ingest data for every area of the business, similarly lack capacity for identity resolution and, while useful for spotting patterns and trends, do not provide insights that marketers can action directly. Both CRMs and data warehouses are company-wide tools that are not specifically focused on marketing improvements.
Businesses also often work with Data Management Platforms (DMPs), which are tools more oriented towards marketers needs and do provide segments that can be used for personalisation. However, a limitation of DMPs is that they are disproportionately reliant on third-party data and/or anonymous first-party data tags, such as cookies, device IDs or IP addresses. Since the storage and usage of personally identifiable information (PII) is strictly regulated, DMPs must anonymise any data that they aggregate in profiles. While DMPs can provide anonymised audience segments for specific campaigns, they do not provide the single customer view across all data points that some marketing actions call for.
Marketing Cloud Platforms fall short in this respect as well. The likes of Salesforce, Adobe or Oracle Marketing Clouds can be effective in delivering a message to an identified audience on specific channels, but their segmentation ability often relies on demographics characteristics (i.e. female, >40, UK-based) rather than on the last up-to-date registered behaviour. Also, they often constrain marketers into using a single vendor.
Customer Data Platforms (CDPs) have entered this jungle of buzzwords and acronyms by offering a precision tool for marketers that can sit with these others and enhance them. Initially hard to differentiate from other platforms, since they were first introduced a few years ago CDPs have emerged as offering the closest approximation to a true Single Customer View that the industry has so far achieved. This is because it is a tool with one single purpose, precisely the SCV, and aimed at one single team, marketers. In comparison to the platforms above, CDPs ingest data in real-time, so are always up-to-date; they work primarily with first-party and PII data; they capture offline, online, and multi-channel data; and they use machine learning power to join the dots between customer data in different shapes (avoiding data duplication) and to churn out segments that can be activated in personalisation campaigns.
CDPs often complement, rather than substitute, other tools. Their precise configuration in any company, unsurprisingly, varies according to the sector, the type of customer and the business goals. It is your decision what data you want to prioritise and which datasets you want to match. That’s why Teavaro has started offering live webinars to support companies in exploring how they can get the most out of their data through a CDP to make the customer journey as smooth and exciting as possible. Join in for our next free webinar (details below) if you’d like to hear more and ask questions to our SCV specialists.
If you are interested in joining Teavaro’s The Six Challenges of Personalisation webinar on Wednesday 26 June at 1300 BST / 1400 CET, please register here.