As you know, to be GDPR compliant, we don't use a database at Datagma.
We deliver all data to you in real-time.
If you submit a LinkedIn URL or domain name as an entry point, there can be no confusion, and we will be sure to have the correct data.
However, if you submit a company name, several companies may have the same name, and this may cause confusion.
Similarly, if you submit a person's name along with a company name, there may be several similar names (such as several employees at IBM named “John”).
To avoid giving you incorrect data, we have developed a scoring system that allows us to determine the reliability of data.
Here's how they work
The closer the score is to 1, the more certain we are of our data.
These scores are not considered when you use LinkedIn URLs or domain names as input, as these permit us to find the right data with certainty. In these cases, the score will always be 1. The scoring is valid for names, company names, and email addresses.
The debug parameter (which can be true or false but defaults to false) is the parameter that allows you to filter out results with a bad score. If you only want to have data that we are sure is reliable, then no change to this parameter is necessary.
However, if you want to take more risks but get more results, then you can set this parameter to false and enter the details of the different scores.
Scores related to personal enrichment
personConfidenceScore: This score answers the question, is this the right person?
If you submit a LinkedIn URL, there is no question, and the score will always be 1, but if you submit a name + company name, we will run a score to help you determine if we have selected the right person.
This scoring results from calculating various elements, including your target's name and current and past experiences.
companyConfidence: This score answers the question of whether the person has been able to change company from the company submitted in input.
We compare the name of the company given in input with the name of the company we get in output.
📘This function is less advanced than our Job Change Detection endpoint, which responds with greater certainty to the problem of job changes
Example: If you submit Raymond Rutjes with the company name Devarium, we will detect that the person now works at Algolia.
Therefore, we'll indicate a low companyConfidence score to indicate a likely shift.
However, we will still enrich the information on Algolia if you ask to enrich the company information in the same call. (By submitting companyFull or companyPremium parameters)
If you submit Raymond Rutjes with the company name Algolia, the companyConfidence will be 1.
Scores related to corporate enrichment
premiumConfidenceScore : This is the reference score for all the information related to the Compay Premium object
If you submit a LinkedIn domain name or company URL, then the score will be 1.
If you submit a company name, then the score may be lower, as the input company name may differ from the output name.
Example: You submit Palantir, but the exact name of the company is Palantir Technologies
• fullConfidenceScore : This is the reference score for all the information related to the Compay Full object.
This is similar to the previous score but for the "full" information i.e. fundraising, website traffic, etc.
Scores related to the companyEmployees parameter
This paragraph is only relevant if you want to use the companyEmployees function, which allows you to identify employees at specific Job Title in companies.
EmployeeJobScore: Compares the input of the job title(s) indicated in input with those obtained in output.
Example: If you submit Sales OR head of OR marketing
head of sales = 1
vp of marketing = 1
Account manager = 0
employeeCompanyScore: Allows you to compare the company's input indicated in input with those obtained in output.
Example: If you submit Palantir, but the company's exact name is Palantir Technologies, so the score will not be 1, and could be filtered.