Written by Qin Li
Short for “a 360-degree customer view”, Customer 360 is a pillar of most data initiatives. With the explosive growth of data in cloud computing and CRM systems, organizations of all sizes must establish a Customer 360 strategy to deliver a compelling customer experience. Their goals usually include one or more the following:
- Customer onboarding: to orchestrate front-office to back-office data processes while eradicating the meaningless ‘busywork’ throughout the customer lifecycle, enhancing the customer experience and reducing costs
- White space discovery: to determine new product opportunities based on existing customer relationships; new target companies based on our penetration insights; consistency across the customer journey, improving transparency
- Analytics driven operations: to provide rapid integration of data across sales, marketing, customer support, and IT teams for new insights and collaboration capabilities, enabling better market segmentation and territory management.
The impact from achieving any one of these goals can be significant — both to the customer experience and the bottom line — making it no surprise that Customer 360 initiatives are so common. But before embarking on this journey, it’s essential to be aware of the hard truths associated with these initiatives, and have a plan in place to overcome them.
Truth 1 – “Easy-button” solutions do not exist
The most common (and most misleading) Customer 360 articles promise “easy button” solutions. The truth is that Customer 360 is hard, and achieving it will require you not only to overcome technical and management challenges but also to conquer organizational inertia. As an example, let’s look at the technology required to deliver an effective customer onboarding experience.
First, you will need Customer Relationship Management (CRM) software (like Salesforce) for organizing your sales, service, or marketing efforts. The features of a CRM system, however, are highly dependent on underlying data such as provisioning, order, and touchpoint data, which is likely outside the CRM system’s control for any large organization.
Then, you will need data management solutions to deliver the accurate, up-to-date data needed for operations and analytics. You will need something agile to overcome the constant changes in your customer base and data sources so that you maintain an accurate understanding of your customer base — individually and in aggregate.
Finally, there will be internal applications designed to meet specific goals of yours under your business context. These applications need to consume (and most of the time, also produce) customer data from many, sometimes unrelated data sources. The applications should be able to evaluate the data for quality problems and resolve the issues introduced by data inconsistency.
Navigating through these complicated solutions (and we haven’t even touched private customer information or GDPR) requires a lot of communication, collaboration, and integration. A key risk is trying to apply a ‘one-size fits all’ solution to cover all of these areas. Instead, companies should adopt a “best-of-breed” approach that provides the flexibility to modify or integrate new solutions as business needs evolve. This approach also allows you to begin capturing value within months, instead of the years required when taking a ‘boil the ocean’ approach.
Truth 2 – The data is more fragmented than you think
Most large, mature companies, despite diligent attempts to consolidate their data, fail to do so. There are regional boundaries, layers of legacy systems, data that came with acquired companies, and more, causing the time and effort required to bring data from multiple sources to be longer than is practical. By the time the data is consolidated, it could be out of date, or the costs incurred might outpace the value that a consolidation would provide.
In a customer onboarding process, for example, you will want to know the relationships between the customer you are onboarding with data already in your system, such as entities:individuals and other “objects” like bills, service tickets, policies, contracts, locations, and businesses. This is important in every industry, where customers expect previous interactions to carry over to new interactions, but it is especially important for industries such as financial services, where significant risk can be introduced by not having this data. You also want to leverage existing data or third party data to expedite your onboarding process so that your sales reps don’t have to manually input information into the system (or worse yet, input data with poor quality). Entity resolution, data synthesis, and relationship discovery are all essential to making sure that your complete “picture” of the customer is up-to-date and reflects all the data you have available. This is the core of the Customer 360 challenge that should not be approached lightly.
This is where Tamr’s machine learning-based approach comes into play. With a small amount of user guidance, Tamr Unify can find the connections between data in any number of sources to quickly provide a single view of data related to each customer. This approach is able to overcome the inevitable, constant changes to underlying data sources, including the onboarding of new sources, changes in the schemas of existing sources, and duplicate customer entries that may not be readily apparent. Unify synthesizes those sources continuously, leveraging best-of-breed components such as Spark for distributed computation, to provide an up-to-date understanding of each customer.
Truth 3 – Big investment alone won’t get you there
Traditional approaches to big data projects are expensive and the business value comes too late, if ever. In a traditional project, you usually start with a centralized data governance system, then move on to implement an MDM hub, and try to integrate MDM with other applications. Because everything is rule-based, it’s slow and costly to maintain. It takes many months to start realizing any tangible business value.
Here at Tamr, we believe in rapid iterations and experimentation. Just like our product, with each iteration, the machine learning model “learns” from the subject matter experts and gets more accurate. Investment on data integration projects can be incremental as well. Based on the learning as the project progresses, companies can understand more about their data problems and spend more efficiently and effectively on data engineering. The time-to-value of your investment can also be much sooner.
Customer 360 is important, but you need trustworthy data
Customer 360 positions you for success by putting you in sync with your customers and their current and future needs. Your sales and customer service folks want to be competent in front of a customer—there’s nothing more frustrating for the customer or the representative than a disconnect between your organization’s data and the real world.
To successfully implement a Customer 360 strategy is difficult, but you shouldn’t settle for anything less. We o help our customers maintain trustworthy data, so that their best people are focused on customers; not on contacts or addresses or business unit silos.