Cybersecurity Job Listings from Dice.com

Challenge pressure

Job postings carry the facts a hiring market runs on (pay, seniority, skills and security clearance) as free text that every employer writes differently. The client needed Dice.com's cybersecurity listings as structured rows.

The situation

On Dice.com, salary arrives in three shapes: annual figures with currency text and decimals, hourly rates, and bare numbers with no unit at all. Location text mixes the work mode with the city, and clearance requirements are free-form text. The salary field needs parsing, not reading.

What we built

The scraper collected the search results into one CSV: 500 listings, 23 fields each, with every value kept as the site returned it, including salary text, location text, relative posting dates and free-form clearance requirements. For the dashboard on this page, Van Data Team then parsed those fields: hourly rates are annualised at 2,080 hours, the work mode is split from the city and region, clearance text is reduced to its highest level, and a salary that cannot be read is excluded rather than imputed.

What changed

The listings became rows that can be counted. They show a remote-first market: 468 of 500 listings (94%) name Remote, though only 162 are remote-only. 334 listings yield a usable salary, with a median of $145,714 a year, and a security clearance carries no measurable premium: $144,910 for the 92 listings that name one against $147,513 for the 242 that do not.

Constraints and decisions

The file is analysed as delivered, flaws included: rows 57 and 58 are exact duplicates and are flagged rather than hidden, and the 166 listings without a usable salary stay out of the pay figures. The requirement text also holds a surprise: after Cyber Security itself (438 listings), Management, Collaboration and Communication are each named more often than Cloud Computing (196) or Python (123).

Technology stack