A Dutch vacancy dataset should at minimum contain the job title, employer name, location, contract type, publication date, and a unique identifier for each listing. For the dataset to be genuinely useful in recruitment analytics, labor market research, or job board applications, it also needs structured salary data, occupational classification codes, and legal compliance fields. The sections below unpack each of these requirements in detail, from the most essential fields to validation practices that keep your Dutch vacancy dataset reliable over time.
Which data fields make a Dutch vacancy dataset actually usable?
A Dutch vacancy dataset becomes usable when it contains a consistent set of core fields that allow filtering, deduplication, and analysis without manual cleanup. At minimum, every record should include a unique vacancy ID, job title, employer name, publication date, expiration date, contract type, work hours, and a direct URL to the original listing. Without these fields, downstream applications cannot reliably process the data.
Beyond the bare minimum, the following fields significantly increase usability for job boards, HR systems, and labor market researchers:
- Job title (raw and normalized): The raw title as posted by the employer, plus a standardized version for grouping similar roles
- Employer name and employer ID: Allows tracking vacancy volume per organization over time
- Contract type: Permanent, temporary, freelance, internship, or side job
- Work hours: Full-time, part-time, or a specific hours-per-week range
- Publication and expiration dates: Essential for freshness filtering and trend analysis
- Source URL and scrape timestamp: Supports traceability and deduplication across sources
- Job description text: Needed for keyword extraction, skills parsing, and natural language processing
A dataset missing any of these core fields will require significant preprocessing before it can feed into a job matching engine, a market intelligence dashboard, or an automated reporting workflow.
What location data should Dutch vacancies include?
Dutch vacancy location data should include at least the city name, province, and postal code region. For datasets used in geographic analysis or commute-based filtering, adding GPS coordinates and municipality codes aligned with the Dutch CBS (Centraal Bureau voor de Statistiek) classification makes the data significantly more powerful.
The Netherlands uses a four-digit postal code system (postcode) that maps reliably to neighborhoods and municipalities. Including the first two digits at minimum allows regional aggregation, while the full four-digit code enables precise local filtering. For remote or hybrid roles, a dedicated field indicating work location flexibility is increasingly important in 2026, as many Dutch employers now specify remote-first or hybrid arrangements explicitly in their postings.
Useful location fields to include:
- City (gemeente)
- Province (provincie)
- Postal code (postcode, 4-digit)
- CBS municipality code (gemeentecode)
- Latitude and longitude (for map-based applications)
- Remote or hybrid flag (boolean or categorical)
How should salary information be structured in a Dutch vacancy dataset?
Salary information in a Dutch vacancy dataset should be stored as a structured range with a minimum value, maximum value, currency (EUR), and pay period (monthly, annual, or hourly). Storing salary as a raw text string extracted from a job description makes filtering and comparison nearly impossible without additional parsing.
Dutch employers frequently express salaries as monthly gross amounts, often written as ranges such as “€3.000 tot €4.500 per maand.” A well-structured dataset normalizes these into numeric fields and adds a standardized pay period so that monthly and annual figures can be compared directly. Where salary is not disclosed, a null value with a “not disclosed” flag preserves data integrity better than leaving the field empty or filling it with a placeholder.
Recommended salary fields:
- salary_min: Numeric minimum value
- salary_max: Numeric maximum value
- salary_currency: Always EUR for Dutch listings
- salary_period: Monthly, annual, or hourly
- salary_disclosed: Boolean flag indicating whether salary was published
- salary_raw: Original text as published, for reference
What classification standards apply to Dutch vacancy data?
Dutch vacancy data is most interoperable when classified using the ISCO-08 (International Standard Classification of Occupations) or its Dutch equivalent, the ROA-BRC (Research Centre for Education and the Labour Market Beroepenclassificatie). For sector-level analysis, the SBI (Standaard Bedrijfsindeling) code from the CBS identifies the industry of the hiring organization.
These classification standards matter because they allow datasets from different sources to be merged and compared without manual mapping. A vacancy tagged only with a free-text job title cannot be reliably aggregated with records from another source unless both are mapped to a shared taxonomy. For Dutch labor market research, CBS and UWV (the Dutch Employee Insurance Agency) both publish vacancy statistics using these classifications, so aligning your dataset with the same standards makes it directly comparable to official government data.
Key classification fields to consider:
- ISCO-08 code: Four-digit occupational classification
- SBI code: Industry sector of the employer
- Education level required: Mapped to the Dutch NLQF framework (MBO, HBO, WO levels)
- Experience level: Entry, mid, senior, or management
What legal and compliance fields must a Dutch vacancy dataset include?
A Dutch vacancy dataset must include fields that support GDPR compliance, particularly when the data contains any personal information about applicants or contact persons. At the dataset level, this means recording the data source, the collection method, the legal basis for processing, and the retention period for each record.
Dutch and EU data protection law requires that personal data be collected and stored with a clear legal basis. For vacancy data that includes recruiter names, contact email addresses, or hiring manager details, the dataset must document where this information came from and how long it will be retained. Even for datasets that appear purely organizational, it is good practice to flag whether any field could be considered personal data under GDPR definitions.
Compliance fields to include:
- Data source: URL or name of the originating platform
- Collection timestamp: When the record was scraped or received
- Legal basis for processing: Legitimate interest, consent, or public task
- Contains personal data: Boolean flag
- Retention date: When the record should be deleted or anonymized
- Terms of use compliance: Whether the source platform permits data reuse
How do you validate the quality of a Dutch vacancy dataset?
Validating a Dutch vacancy dataset involves checking for completeness, consistency, deduplication, and freshness across all records. A high-quality dataset has no duplicate vacancy IDs, all mandatory fields populated, location data that resolves to real Dutch municipalities, and salary fields that fall within plausible ranges for the Dutch labor market.
Practical validation steps include:
- Completeness check: Calculate the fill rate for each field. Fields like job title, location, and publication date should be close to 100% populated.
- Deduplication: Use a combination of employer name, job title, location, and publication date to identify and remove near-duplicate records from multiple sources.
- Format validation: Verify that postal codes match the Dutch four-digit format, dates follow ISO 8601, and salary values are numeric.
- Freshness audit: Flag records where the expiration date has passed or where no update has been received within a defined window.
- Outlier detection: Identify salary values that fall outside realistic ranges for the Dutch market or job titles that appear garbled or machine-generated.
- Source cross-check: Compare a sample of records against the original source URLs to confirm the data was captured accurately.
Running these checks automatically as part of a data pipeline, rather than manually on a one-off basis, is the only way to maintain consistent quality in a dataset that grows or refreshes continuously.
How Openindex helps with Dutch vacancy data collection
Collecting, structuring, and maintaining a Dutch vacancy dataset at scale is a significant technical challenge. We specialize in exactly this kind of work. At Openindex, we build and operate crawling and data extraction pipelines that gather job listing data from across the Netherlands, normalize it into structured formats, and deliver it as clean, ready-to-use feeds or API endpoints.
Here is what we offer for organizations working with Dutch job market data:
- Crawling as a Service: We handle the full crawling process, from source discovery to data delivery, so your team focuses on using the data rather than collecting it
- Custom data schemas: We structure vacancy data to match your specific field requirements, including ISCO codes, CBS classifications, and GDPR compliance fields
- Data as a Service: Receive structured vacancy feeds on a scheduled basis without managing any infrastructure yourself
- Deduplication and validation: Our pipelines include built-in quality checks that keep your dataset clean and consistent over time
- Legal and ethical collection: We operate in compliance with GDPR and respect the terms of use of the platforms we index
Whether you need a one-time dataset for labor market research or a continuously updated feed for a job board or HR platform, we can deliver it. Contact us to discuss your vacancy data requirements.
Frequently Asked Questions
What is the minimum viable set of fields to start working with a Dutch vacancy dataset?
At a minimum, you need a unique vacancy ID, job title, employer name, publication date, contract type, and location (city + postal code). This gives you enough structure to filter, deduplicate, and run basic analysis without heavy preprocessing.
How do I handle vacancies where salary information is not disclosed?
Use a boolean ‘salary_disclosed’ flag set to false and store a null value in the salary fields rather than a placeholder like ‘0’ or ‘N/A’. This preserves data integrity and prevents salary fields from skewing any aggregations or comparisons in your analysis.
Which classification standard should I prioritize if I can only implement one — ISCO-08 or SBI?
Choose based on your use case: ISCO-08 is better for occupational analysis and job matching, while SBI codes are more useful for sector-level or employer-focused research. If your dataset will be compared against CBS or UWV government statistics, aligning with both from the start saves significant rework later.
How often should a Dutch vacancy dataset be refreshed to stay reliable?
For active job board or HR platform use, daily refreshes are the practical standard, since many Dutch vacancies expire or are filled within days. For labor market research or trend analysis, weekly updates are typically sufficient, provided your pipeline flags expired listings and removes stale records automatically.