Contact us and see what NetOwl can do for you!
How Identity Resolution Helps Consolidate CRM Systems after M&As
Mergers and Acquisitions Cause Big Headaches for Information Systems
Mergers and acquisitions create significant opportunities but can also pose complex integration challenges for the entities involved, arising from different cultures, business processes, and IT systems. Similarly, often one of the most desirable benefits of an M&A is the expansion of one’s customer base, but producing a comprehensively consolidated Customer Relationship Management (CRM) system requires merging multiple customer databases.
Traditionally, CRMs are used to maintain a unified view of customers across all their contact points with a company, whether via phone, email, web site visit, etc. They are meant to construct a complete picture of a customer over multiple engagements and ensure that all past experience with a customer can be brought to bear on a current engagement.
Merging Customer Databases Is a Challenge for Any IT Manager
The issue of how to integrate different customer databases, which may have very different underlying data models, is a sizable challenge.
As a first step, each record in one database must be compared with each record in the other, which entails comparing all appropriate fields for each record such as:
- Date of birth
- Mailing address
Records for the same entity can differ in multiple ways. In the case of person records, names can differ in terms of:
- Different order of the elements (e.g., John Smith vs. Smith, John)
- Missing components like middle names/initials being present or absent (e.g., Jane Murphy vs. Jane Southwick Murphy)
- Nicknames vs. full forms (e.g., Paco vs. Francisco)
- Orthographic variations (e.g., Sean vs. Shaun)
- Ethnicity-specific variations (e.g., Harun al-Rashid vs. Harun Rashid, where ‘al-‘ is a common element in Arab names and it is frequently omitted in transliteration)
- Typos and misspellings
- Decomposition such as when one database assigns the entire name to one field, while the other splits it up into two or more fields (e.g., Full Name vs. First Name + Middle Initial + Last Name vs. First Name Middle Initial + Last Name)
Other relevant fields like addresses can exhibit different variations like abbreviations (e.g., Street vs. St., North vs. N., Suite vs. Ste.) and missing components (e.g., postal code, state).
If the customer records are businesses, they too come with their own peculiarities:
- Full forms vs. acronyms: Hewlett Packard HP
- Full forms vs. abbreviation: Washington Mutual vs. WaMu
- Presence vs. absence of corporate designators: LLC, , etc.
How Identity Resolution Helps Merge Multiple Databases
Identity Resolution is an AI-based technology that solves the challenges of merging databases described above. It uses real-world data and Machine Learning to develop matching models for each data type to be matched. For example, for personal names, Identity Resolution can match names that differ due to nicknames, misspellings, word order, or missing components. Identity Resolution can also detect a wide range of linguistic ethnicities and support ethnicity-specific phenomena such as the aforementioned optional ‘al’ in Arabic names.
Identity Resolution assigns a similarity score to each relevant field in a record. After each field has been scored, the matching algorithm combines the scores for the individual fields into an overall score. The higher this overall score, the more likely the two records are to refer to the same individual or organization.
Identity Resolution allows for the tuning of the matching to support different use cases and business rules. It is not unusual for some fields to be more important than others when comparing records. For example, some use cases may require a strong match on the date of birth, and Identity Resolution can accommodate that through tunable field weights. Business rules can thus be created that determine what combination of fields should be matched and how important each field is to the overall score.
Another way in which Identity Resolution can be configurable is through the setting of a threshold for what constitutes a valid match. For some high-precision use cases, a high threshold would be the way to go whereas for high-recall use cases, a lower threshold may be desirable.
In sum, corporate mergers and acquisitions frequently create CRM data integration difficulties that could take a very long time to resolve and detract from an important benefit of an M&A. Identity Resolution is an advanced AI technology that supports effective reconciliation of CRM data.