What is customer data? Definition, types, collection methods and analysis best practices

We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 – 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!


What is customer data?

The modern business world is driven by data and some of the most valuable is customer data. Finding the best way to satisfy customers requires studying their needs and that means analyzing the data trail they leave behind. 

Customer data comes from a variety of sources. Some of the most concrete information comes from completed sales; successful transactions contain a wealth of decision-making data that spells out what went right during the sales process and offers a solid guidepost for planning future sales presentations. 

The actual transaction details, though, are just the beginning. Businesses also routinely capture information from advertisers and sales channels that track the customers as they explore their options and move to a decision. These details from ad channels, websites and in-store venues offer useful waypoints that can help revise and focus the entire sales narrative. 

Finding the best way to gather this data, clean up any irregularities, organize it for consistency and then analyze it is the challenge for modern enterprises. This requires a data scientist’s knowledge of statistics along with a marketer’s instincts for the customer and a visionary’s impression of what the future could bring. 

“All of this [data] is converging and the [customer data platform] is the central hub to marketing teams and more,” said Ryan Fleisch, director of product marketing for Profile & Activation at Adobe. “It’s so important that more teams are using and flowing data into a central place where everybody’s reading off the same playbook.”

Key attributes 

There are many different forms of customer data, and they come from a variety of sources. Customer data specialists often categorize these by making several key distinctions:

  • Does the data come from an individual or a group? Some information is connected with one particular customer, perhaps from his or her browsing or purchase history. Other data is gathered from some or all of the customers and represents averages or totals. 
  • Is the data anonymous or identifiable? Some customer data is tied directly to a person’s name, address or other personally identifiable information. Others have no direct link, either because the name was stripped out deliberately or it was never available in the first place because it came from an anonymous source like an open website. Personally identifiable information requires more care. 
  • Is the data raw or aggregated? Some customer data is collected directly from websites, store fronts, advertisers or other foundational sources. It often represents simple, basic events like a click on an ad or a submission of a form. Other data has been aggregated and analyzed by reporting software to summarize the behavior of one or more customers over time. 
  • How much structure does the data have? Data scientists often make a distinction between structured and unstructured information. When the data is placed in fields with specific data types that are checked and enforced, then it is easier to analyze. This structure might include fields like the time, date, ZIP code or other details. Unstructured data is often found in human written text that might be submitted by the customer or gathered in the notes from customer service representatives. 
  • Did the customer voluntarily give up the information? Some information is gathered directly from the customer voluntarily. This may be a form that’s filled out to request information or a discussion with a customer service representative. Other information like the browsing pattern at the storefront or the history of advertisement views may be gathered without asking the customer. 
  • Does the data represent all customers or just some? Some data tables, like the list of completed orders, are comprehensive. They include everyone. Other tables only include customers who opt-in by initiating a conversation or filling out a form. Other tables from ad trackers may miss some users because the matching process is error-prone. 
  • Are there privacy concerns about the data? When the customer data includes personally identifiable information like a name or purchase history, many customers don’t want this information to be disclosed publicly. Guarding their privacy is important. Other data that doesn’t have personal information or has been aggregated already is less sensitive because it’s often not possible to connect the details with any particular person. Still, care should be taken because in some cases it’s possible to infer someone’s name from anonymized data. 

Types of customer data

Data about the customers comes from a wide variety of sources and the list of possible resources is getting longer and longer. Some of the traditional sources are:

  • Customer requests: Did the customer fill out a form, ask for a price quote or request information? 
  • Customer purchases: When did the customer buy in the past? How big was the order? Was it connected to any promotions or sales?
  • Web advertising: Who was able to see particular advertisements? How often were they shown? Did the customer click on any particular message? 
  • Direct email advertising: Was the customer sent any particular email messages with marketing? Did they open the messages? Did they respond soon afterwards? 
  • Phone contacts: Did the sales force contact the customer by phone? Did the customer pick up? Did the customer request any information? 
  • Display advertising: Was the customer reading or watching any content where display advertising was shown? 
  • Storefront traffic: Did the customer visit any physical stores? Did the customer make any purchases or make any requests? 
  • Financial information: Which payment method did the customer use? Is the same method used with other transactions? 
  • Daytiming: Is the customer active at particular times of day or days in the week? Is the customer more open to buying or listening to marketing messages at particular times? 

Customer data collection: Key methods

Much of the work for enterprise data managers is collecting the data from a variety of sources and then finding the best way to integrate it into reports, charts and tables that can guide future decisions. Some of the data comes directly to the enterprise and some comes from third parties or government agencies. 

The challenge is to start to integrate it so the enterprise can understand what’s happening to all customers and users. Careful collection and analysis leads to a more complete picture. 

“I think most companies have a natural reaction to attach themselves to a really bad example, when someone writes in with some really bad consequences that happened for them,” noted Chris Martinez, co-CEO of Idiomatic, which specializes in using AI to understand customer data. “You surface that one case and you say, ‘Hey we have to fix this because look how bad of an experience one person had.’ But oftentimes you’re missing the slow-burn issues — the things that are not as shocking, but happen to a lot more people. By having a data-driven kind of rigorous approach instead of an anecdotal approach to this, you can actually solve problems for more people.”

Some of the best data collection tools and methods are:

  • Direct contact: Did the customer ask any questions about products or services? Did the customer write directly through email or a web form? 
  • Requests: Is the customer looking for a product, service or feature? Is this something that the business doesn’t offer yet?
  • Complaints: After sales, a follow-up can generate important details about failures that can guide future transactions. 
  • Ad channels: These third parties can report how well certain marketing messages are being communicated and received. Are they generating any sales leads? 
  • Physical stores: How is the foot traffic? Are certain items selling better? Is there any connection between marketing and sales? 
  • Web traffic: When customers come to your website, which path do they take? What items do they see? Are they following any hints? 
  • Telephone contacts: When are they sent text messages or phone calls? How and when are they responding? 

One of the most popular data collection approaches is to integrate much of this information with a tool known as a customer data platform (CDP) or a data management system (DMP). There are dozens of companies that are making tools that fall into one or both of these categories and they’re being widely adopted. 

The tools are popular because they are optimized for integrating data from all of the possible sources. They often have predeveloped pathways for importing data from the most common ad platforms and store tracking software. Some are also connected with some of the best customer service platforms. 

“Companies are already sitting on [data] they basically just are not using and we really see this as a lost opportunity,” said Kevin Yang, co-CEO of Idiomatic. “Just to give you a concrete example, customer service companies spend an incredible amount of money just having customer service agents to answer these tickets. But they don’t use the opportunity to actually analyze this to make their experience better. If they were able to do this, customers would be really much happier to share information with them, because they’ll be a lot more responsive and the companies that are best at doing this.”

Top 10 best practices for customer data integration and analysis in 2022

Here are 10 best practices for data collection, integration and analysis:

  1. Guard customers’ privacy. Consumers are nervous about how their data may be misused. Even if there are no abuses, the potential can be scary. Guarding their privacy is essential. 
  2. Choose a good tool like a CDP or a DMP. The marketplace is already delivering a number of good options for tracking consumers and their data. There are fewer needs to build custom platforms. Leveraging the power of commercial options can produce much better results at a lower cost. 
  3. Activate as many pathways for gathering information from partners as possible. Work with your partners to collect data. Negotiate access up front and use it frequently. Insist that ad companies provide tracking data so marketing campaigns can be evaluated. 
  4. Build a culture of data collection in house. Don’t just sit passively and wait for insights to fall from the sky. Nurture data collection internally. Watch for good sources. Improve your current data collection practices. Ensure that team members are contributing. Emphasize that the future is about doing the best job possible collecting data. 
  5. Educate sales teams on the best practices for using the data. Many traditional sales teams are focused on connecting with potential customers. Help them understand the power that data can bring to their communications. The sales teams will be the ultimate users of the data and so they should understand the process and contribute to each step. 
  6. Explore the best data science algorithms. It is easier than ever to use good statistical methods and data science practices to analyze the data. Without them, the data is just a pile of numbers. 
  7. Look at artificial intelligence for guidance. The best CDPs and DMPs are integrating good AI algorithms into them to help with classification and prediction. These algorithms can help teams understand the customer base and anticipate how they will behave in the future. 
  8. Watch regulations for changes. Some governments are rewriting the laws and regulations that control how we store and use data. This is not a static, settled part of the law. Watch for changes and be ready to shift as new regulations make some old practices impossible or illegal. 
  9. Search for new data sources internally and from third parties. The data available is going to shift. Ask your team to look for ways to squeeze more data from internal tracking sources. At the same time, watch for third-party data vendors who might be able to tap open source, governmental databases or other sources to add more insight. 
  10. Build long-term partnerships with good, trustworthy sources. Improving your data is a constant process. Don’t expect to make a few quick purchases and be done. Work to create long-term relationships with the data vendors, advertising companies and other sources because this is the only way to keep the customer data clean and fresh.

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn more about membership.

Leave a Comment