Segmentation

Segmentation is the process of taking data with a lot of disparate values and creating smaller segments to target more effectively.

Common attributes to segment include: Title, Industry, Annual Revenue, Number of Employees.

Prerequisites:

1. Data Essentials: It is highly recommended that you implement Data essentials to cleanse, standardize and normalize your data prior to segmenting your data.
2. Enrichment (optional): to provide more comprehensive data on records for full segmentation

Segmentation is recommended prior to more complex use cases, such as routing, particularly if routing requires region, industry, or similar segmentations.

Job Function, Job Level, Industry

Openprise segments titles by job function, job level, and industry by using one of the below reference tables in Data > Data Catalog:

Each table contains a keyword column, priority column, and language column. 

Priority

The priority column is a whole number data type used to resolve conflicts if multiple keywords in a title match to multiple keywords in a reference table with varying job functions, job levels, or industries.

Priorities are ranked from 1 to 10; the lower the number the higher the priority. Openprise uses the following best practices when assigning priority values to keywords:

1. Priority 1
   • Specific brand or product names, multi word combos or full names  
   • E.g. Oracle E-Business, Chief Financial
2. Priority 2
   • Complicated multi-word combos, full names with additional characters
   • E.g. communication superintendent, CTO & 
3. Priority 3
   • Regular single words, non-specific full names
   • E.g. biochemical, consultant
4. Priority 4
   • Generic single words
   • E.g. communication, director
5. Priority 10 
   • Last resort words, abbreviations 
   • E.g. CTO, pm

Keyword and Match Method

Openprise uses the infer value task template to match values from titles to values in the keyword column. Titles from the input data source are compared to keywords in the reference data source using input value contains reference value as the match method.  

Why input value contains reference value and not vise versa?

1. The input value is a title you want to look up

2. The reference table holds extra information (job function, etc) that can be inferred if a match is found

3. The look-up value is the keyword column in the reference table you match titles against
4. The reference data source contains keywords commonly found in titles, but it may not contain an entire title as there are infinite possibilities. Therefore, the input data source must contain a value that corresponds to the reference value. Without it, there’s no way to know what to look up.

Selecting reference value contains input value is the equivalent of asking “For every possible value in my reference table, do I have a match in my input data source?”, which is different from segmentation. 

Annual Revenue, Number of Employees  

Openprise segments records by annual revenue and number of employees by using one of the below reference tables in Data > Data Catalog:

Each table contains a maximum column, minimum column, and range name column. 

The ranges specified in these tables are examples. If your input data source contains values that exceed the maximum value, or you want to change the name of the range, you will need to create your own table and use it as a reference table when configuring the infer value task template. 

Resources

Click HERE to register for the Segmentation course in Openprise Academy.