Closing the Customer Data Gap
POSTED : June 26, 2015

Most of us talk a good game about customer-centricity and using data (big or otherwise) to deliver personalized experiences that give customers what they need before they even ask. But actually achieving this customer-centric nirvana is much easier said than done.

Often the biggest gap is in understanding who the customers really are, what their needs are, and where they will be reachable, both now and in the future. The thing is, customer data can’t do anything without a process around it. And we’re often stuck with a lake of questionable data, dated customer research, and few actionable insights.

In a recent survey of CMOs conducted by EY and Forbes Insights, 87% said their strategic vision for building trust includes perfecting the customer experience and that the latest data and analytics technologies could help build credibility and long-term relationships with customers. But as Woody Driggs, Principal and Global Lead of the EY Customer Practice, said, “CMOs are specifically challenged in how to leverage advanced analytics to sort through the information to gain deeper insight.”

Organizational silos, complexity & lack of trust

The problem is not that we don’t have enough data; it’s that it’s not connected. It lives in too many places, with too many owners, and supports too many disparate goals and objectives. CRM data is owned by Sales, while marketing automation is owned by IT and managed by Marketing, and outbound performance is locked up in an Excel sheet emailed to no one by the ad network. In the end, there is no common language or unified strategy for how customer data is used.

The sheer number and complexity of these digital platforms and channels creates the illusion that we have a “complete” picture of customers and their behavior. As a result, marketers are spending more time and resources understanding the various touchpoints and tactics, instead of understanding their customers.

In many cases there is also a gap in trust. Anyone who’s ever had even a tiny issue with the quality of their data may believe all customer data is inherently flawed.

I recently spoke with a mixed group from a large team of marketers, with roles ranging from data practitioners to sales execs to the CMO herself. Their assessment of the organization’s customer data varied from hair-on-fire terrible to not so bad, depending on how close they were to the data itself. Hands-on data practitioners were not too worried, while those who relied on the data to inform decisions were much more dubious.

This stuff is hard, especially if it is flying in the face of entrenched practices that have very deep tentacles into an organization.

Three types of data give marketers a more complete picture of customer interactions

Closing the customer data gap requires a unified strategy for your organization’s customer data, including how it’s collected, maintained, and used. Once you break down all the walled gardens, we see three core components of customer insights data. Together they can paint a more coherent picture of how customers are interacting with your organization:

  1. Known user data: who they are
  2. Observed behavior data: what they do
  3. Voice of customer (VoC) data: how they feel

Let’s break each down and discuss how they contribute.

Known user data

What is a “known” user? Well, it’s someone you know, meaning you have enough information about them to be reasonably certain who they are. This could be as little as a name and email—or as much as an entire biography of attributes. The “data” in this case is really more like metadata than spreadsheet data, meaning it describes relevant attributes about a person, such as whether they are a customer or not.

Many lead generation marketers are completely focused on knowing people. Their whole job is to “convert” unknown users to known users. Then sales can again convert them to customers. Hooray, hive-fives all around!

Generally this type of user data is recorded, augmented, and maintained in a customer relationship management (CRM) platform like Salesforce, which acts as a record of truth for customer data, hence the “M” in CRM.

How do you get to know users? Normally, ask them to tell you about themselves via a form, whether all at once or progressively. A big trend now is to use third-party platforms like Demandbase and LeadLander to pull in additional company or demographic information.

Marketing automation platforms (MAPs), such as Marketo and Eloqua, also offer CRM-type capabilities and can act as another customer database.

And this is where things get fuzzy between the data types.

Observed behavior data

Observed behavior data is a record of an interaction with a touchpoint, digital or otherwise. Essentially, we use this data to “observe” what users are doing with our stuff. This is where web analytics platforms like Google Analytics and Adobe Analytics come into play. Their whole purpose is to record user interactions with digital touchpoints. MAPs can record user interactions as well, though specifically to profile known users and rate how engaged they are.

This type of data is what is marketers are usually referring to when they talk about “analytics,” by the way. As I always say, ask 10 people what they mean by “analytics” and you’ll get 8 different answers.

So, now we can see what customers and other users are doing. But what if you want to know how your customers feel? Are they happy with their experience with you and your services? How would you know (or predict) whether they were likely to continue as customers or drip out of your leaky bucket?

Voice of customer (or consumer) data

What’s interesting about VoC is that marketers so often overlook it when thinking about their customer data stack. We like to describe this type of data as the spackle that holds other data types together. You know a customer; you can even observe them touching your stuff. But do you really know how they feel about you and the carefully crafted experience you have built for them? BAM! VoC.

We have all taken surveys, and they are mostly not a fun experience. Until Google achieves the singularity or we learn to literally read minds, we are left asking succinct and meaningful questions of our customers to better understand them and what they’re thinking.

(Note that these three data types are not meant to be all-inclusive. For example, media and other outreach performance data are not included.)

Crossing customer data streams for actionable insights

Ok, great…we got all the data. Whoopee-do. Now, what do we do with it?

Data from any one of these categories is useful as far as it goes, but there’s far greater potential for meaningful and actionable customer insights when you start to cross the streams.

Known user data + observed behavior data = the ability to build activity profiling about your known users that can be used to understand the impact of touchpoints and drive personalization in future interactions.

Known user data + VoC data = the ability to know how specific and aggregated customers feel about you and your touchpoints.

Observed behavior data + VoC data = the ability to gauge how satisfied users are with specific or broad aspects of the user experience.

All three = the ability to build a fairly complete picture of a customer’s identity, preferences, opinions, feelings, and overall needs.

With a well-founded and “complete” set of customer data, you can start to make real decisions about what to prioritize.

  • See some issues with customer opinions? Go fix it.
  • People really clicky on that new whitepaper? Go ask them why and do more of that.
  • Need to gather more known contacts to raise interest in a new product? Look at a similar cohort among your current customers and learn what their preferences are to attract new customers.
  • Need to build personas but not sure you can get all the qualitative interviews done? Poll your customers to create cluster definitions.

Suddenly, you can make all those data-driven decisions you always thought you could.

Tags: , , ,

Blog Archive