Blog

Entity Resolution for a Single Customer View

“The ability to view the customer through a single lens enables critical measurement, optimization, efficiency, and personalized use cases.”

Jason Niemi, Director Digital Engagements, Kraft Foods Group

Maintaining a single, consistent view of customers across the enterprise is fast becoming a key need for businesses that seek accurate, complete customer information so they can know who they are interacting with better. The successful implementation of a Single Customer View can be challenging where representations of a customer are held in more than one system and customer entity and discrepancies in customer data must be resolved both within and between systems.

In a Harvard Business Review study, more than 400 customer experience executives were surveyed, and achieving a single customer view came out as one of the top challenges.

Let’s take a look at what exactly a Single Customer View is.

A Single Customer View is a holistic, consolidated and consistent representation of an organization’s customer data. Customers expect interactions with an organization to reflect their current profile, history, and preferences. For this reason, a Single Customer View is particularly important when organizations interact with customers through multiple channels.

And when it comes to interactions through multiple channels, we invariably get multiple data sources. The average enterprise uses 65 different systems. Data from all of these internal sources, external third-party brokers or partners, and public data sources need to be integrated to build this Single Customer View.

This means you could potentially have dozens of records across your data sources for each customer. Some records may contain their buying preferences, some may have demographic data, others may have more complete contact data, etc. And in each of these records, the key identifier is a little different. Perhaps the customer used a nickname when filling out your loyalty rewards program card, or they could’ve used their personal email address when signing up on your website. Regardless of the reason, you’re now dealing with a variety of representations for each customer stored across a variety of systems.

Entity resolution is a key data integration problem when merging different sources, because each of these first first-party, second-party, and third-party data sources may be using a different ‘identifier’ to refer to an individual. Let’s take a deeper look at what entity resolution is.

What is Entity Resolution?

Research reveals that 94% of businesses deal with duplicate records of the same entity and perform entity resolution on some level, either manually or using data quality software. Consolidating related records and tying them to a single entity is inherently complex because exact matches are far and few.

Entity resolution, is a core data quality process used to identify records that refer to the same entity within or across data sources. This could be done for deduplication and cleansing purposes, or to enrich and create golden records that absorb entity fragments across your business and create a unified entity profile. The latter applies in this context: building a Single Customer View.

As a business, you could potentially have dozens of records across your data sources for each customer. Some records may contain their buying preferences, some may have demographic data, others may have more complete contact data, etc. And in each of these records, the key identifier is a little different. Perhaps the customer used a nickname when filling out your loyalty rewards program card, or they could’ve used their personal email address when signing up on your website. Regardless of the reason, you’re now dealing with a variety of representations for each customer stored across a variety of systems.

Using Entity Resolution to Build a Single Customer View

Now that we have a better idea of what a Single Customer View is and how it’s primary challenge can be resolved, let’s take a look at the capabilities necessary for creating a single view: Integration, Profiling, Cleansing, and Matching.

Integrate Your Data Sources

The first step is obviously to identify and integrate your data first-party, second-party, and third-party data sources that contain the customer information that you want to bring together. Which format are you receiving data from your partners and brokers in? Which applications or databases do you have in place internally? Will you be connecting to public data sources, and if so, which format will you download those lists in?

Customer data that you’d need to build a true Single Customer View can include:

  • Social media
  • Transactions
  • Sales team interactions
  • Firmographics
  • Customer preferences
  • Web and mobile browsing activities
  • Demographics
  • Sentiments
  • Etc.

Make a list of data sources and create a strategy for how you want to integrate each of those data sources. Now cross-check this list against the tools you want to use for entity resolution and to build your single customer view. Make sure each of those tools can integrate perfectly with your data sources, whether it’s your CRM, social media platforms, accounting apps, or Big Data lakes.

Profile and Discover Your Data

When integrating data from multiple sources at scale, few businesses really understand the underlying data themselves. This data is usually gathered over the course of years, if not decades, and is chock full of issues like typos, irrelevancy, incompleteness, inaccuracy, and lack of standardization. The standardization issue is further magnified when multiple sources come together, because each system may be storing data in a completely different way.

Profiling your data once you’ve integrated the sources helps you get a snapshot view that pinpoints data quality issues. Fixing these issues ultimately helps you get better results when matching different representations of the same customer in the final step of the process. Drilling down and profiling data at the onset allows businesses to quantify the data issues that they would run into somewhere down the road.

Obtain Clean, Accurate Data

Once you’ve identified issues in your data, it’s time to scrub it clean and standardize it to get the best results in the entity resolution step. To clean your data, you can set up business rules that help recognize and fix spelling errors, standardization issues, misfielded data, etc. You should already know which business rules to create if you’ve profiled your data. Keep in mind that this step can be very time-consuming and requires considerable attention to detail. For faster, accurate results, data cleansing software is a good choice.

Industry-leading data cleansing solutions typically offer pre-built cleaning and standardization rules, along with various other features that help efficiently clean data while providing more insight into your business’ data.

Perform Entity Resolution

Standardizing your data before you match it helps minimize false negatives, effectively increasing match rates. It’s time to create your match definitions now. Ideally, your entity resolution software of choice should allow you to visually create match definitions where you define what should or shouldn’t be considered a match. You can think of Match Criteria and Match Definitions in terms of AND/OR SQL statements. The relationship between match definitions would be an AND statement and the relationship between the match criteria would be an OR statement.

Once you’ve defined your match definitions and criteria, it’s time to start matching:

If you’re using state-of-the-art entity resolution software, you will see something akin to the above screen when you run your matches. Individual matches should be paired in groups, identified by a unique group ID. You should also see the match score, allowing you to weed out false positives quickly.

With the matches identified and false positives eliminated, it’s time for the final step in the process: Choosing a master record. After all, our purpose here is to build a Single Customer View that contains the cleanest, most complete record for every customer. Now that you’ve matched all of your data sources that contain customer data, it’s time to bring all of that information together.

From the match groups, you can choose which records and which fields to merge until you have a single, comprehensive record. You do not have to manually go through each record; the process can be automated in DataMatch Enterprise’ Merge and Survivorship step.

Conclusion

Implementing a single customer view is an essential step in meaningfully improving customer experience. By taking a journey-based view, integrating data, profiling and cleaning it, and then matching customer identities through entity resolution software like DataMatch Enterprise, you can more easily bridge the disconnect between customer expectations and customer experience. The result will be better retention, increased loyalty, greater customer satisfaction, improved process efficiencies, and enhanced customer service and support levels.

In this blog, you will find:

Try data matching today

No credit card required

"*" indicates required fields

Hidden
This field is for validation purposes and should be left unchanged.

Want to know more?

Check out DME resources

Merging Data from Multiple Sources – Challenges and Solutions

Oops! We could not locate your form.