Dover uses machine learning to define personas through analysis of current titles, previous titles, and keywords associated with the work under the selected persona. Dover scans LinkedIn profiles to search for titles, in both the "Headline" section and the "Experience" section.

Dover has a list of pre-defined personas that each have a list of titles patterns within its definition. Some personas also use machine learning to scan the candidate’s full profile to determine if the candidate is a match.

For example, Dover’s “Frontend Engineer” persona has 20+ title patterns within its definition (e.g. “frontend engineer”, “UI/UX developer,” “software engineer, frontend”) and also scans profiles with generic title patterns like “software engineer” to see if candidates are focused on the frontend.

Dover personas also use “dealbreaker” logic to ensure accuracy. For example, this logic would help to block titles like “Software Engineer Sourcer” from an engineering search.

You can target or exclude a certain geographical areas from your search. Please note that if you target New York (US) and United States, Dover will select the larger-sized location. Please note that limiting the location significantly impacts the candidate pool size — flexibility here can help increase the number of matching candidates

Seniority level looks at a candidate’s seniority level indicated by the title pattern.

Specifically:

If a company size bucket is selected, Dover will check both current and previous companies to determine if a company's size criteria passes (excluding internships). To account for changes in company size since a candidate worked at a previous company, Dover will check the size of the company during the time when the

candidate worked there.

Current position only

Toggle this on if you'd like your company size requirements to apply to their current position only.

Add exclusions:

If a company size bucket is selected as an exclusion, Dover will check the current company size of the candidate’s most recent company to determine if the candidate should be excluded.

Dover has a database of rankings for 20k+ companies to assist in grouping the prestige of the companies a candidate has worked for into 5 groups. These rankings are influenced by the following variables: general reputation, strength of investor reputation, strength of schools employees attended, and strength of previous companies that current employees worked at previously. These rankings are often differentiated based on the type of role you are targeting. For example, McKinsey has a high score for operations and finance roles, but has a low score for engineering and design roles. Apple has a high score for marketing and design roles, but has a low score for sales roles.

You can target specific companies, or specific lists of companies, for outreach. When company names or lists are input, candidates who work at those companies will still pass the search regardless of whether or not they pass the other variables in the “Companies” section (company size, company prestige, industries).

Dover will look at a candidate’s past 3 companies when checking for target companies. For example, if Google is listed as a target company, and a candidate has worked at 3 or more companies since their experience at Google, then the candidate will not pass on the target company portion of the calibration.

Companies in Dover’s database have industry labels associated with them. Selecting industries will tell Dover which industries to target. If an industry is selected, a candidate will pass this criteria if they have more than 1 year of experience working at companies in the specified industry. Selecting multiple industries will pass candidates that have more than 1 year of experience working in any of the specified industries.

Dover has a database of rankings for around 3,000 colleges and universities, both US and international, to assist with filtering by prestige of the schools candidates have attended. These rankings are influenced by the following variables: general reputation, U.S. News rankings reports, schools with strong Computer Science programs, schools that place students into well-reputed companies.

You can target specific schools, or specific lists of schools, for outreach. When school names or lists are input, candidates who attended these schools will be given a higher score overall, and will be more likely to pass the search even if they don’t match on a few other “nice to have” criteria.

Field of Study

You can filter based on what candidates studied while attending university.

Adding keywords will enable Dover to scan a candidate’s profile to search for mentions of the keywords specified.

Keywords often have "aliases" and "children" keywords associated with them. For example, the keyword “frontend” has the two aliases — “front end” and “front-end” — so that all variations will be picked up. The “frontend” keyword also has many children keywords (e.g. "user experience," "user interface," "web development"), so that all candidates mentioning keywords that relate to frontend development will pass the search.

This filter allows you to filter by gender or racial diversity. Dover defines non-males as female or non-binary identifying candidates. Dover defines underrepresented minorities as candidates who identify as Black/African American, Hispanic/Latino, or American Indian.

Dover uses a combination of the following factors to determine gender and racial groupings:

The strictness scale is used to determine the percentage of “nice to have” variables that need to be present for a candidate to match the search. If you increase the strictness scale, you are decreasing the candidate pool to remove candidates that have the fewest “nice to have” variables.