Graduate recruitment at Dutch law firms occupies a narrow, high-stakes corridor. The cohorts are small, the training investment is significant, and the consequences of a poor hire are felt for years. At the same time, application volumes at firms such as De Brauw Blackstone Westbroek, Houthoff, Stibbe, Loyens & Loeff, and Linklaters Amsterdam can reach hundreds of candidates per intake season. That tension between volume at the top of the funnel and precision at the bottom of it is exactly what makes choosing the right digital assessment platform so consequential.

This guide is written for HR and talent acquisition professionals responsible for that decision. Rather than restating vendor marketing, it provides a criteria-led evaluation framework, a shortlist of relevant platforms, and two recommended assessment workflows for graduate legal hiring in the Netherlands.

Why graduate lawyer selection demands more than a generic assessment stack

Junior lawyer roles require a specific and somewhat unusual combination of competencies: high analytical reasoning, disciplined written communication, professional judgement under uncertainty, and the capacity to absorb complex information quickly. These aren't soft, difficult-to-measure traits. They are constructs with established psychometric literature behind them.

Generic assessments designed for broad graduate populations don't always measure these constructs with sufficient precision. A numerical reasoning test built for finance roles, for example, may not predict performance in legal drafting or case analysis. Equally, a personality questionnaire calibrated on a general working population may produce misleading profiles when applied to law graduates, whose self-selection and educational background already represent a narrow band of the wider norm group.

Add to that the compliance dimension. Any AI-driven platform used to screen or rank applicants is now classified as a high-risk system under the EU AI Act, which entered into force on 1 August 2024 (Datapeople, May 2026). That means documented human oversight, explainability, and governance measures are legal expectations, not optional extras.

Key buying criteria for legal graduate recruitment assessments

Before evaluating any vendor, HR teams should agree internally on what they're actually trying to measure. For junior lawyer roles, that generally means:

• Analytical and verbal reasoning: the ability to draw accurate inferences from complex text and data, often proxied by tests like the Watson Glaser Critical Thinking Appraisal or bespoke verbal/numerical batteries
• Situational judgement: how candidates respond to realistic professional dilemmas (client ethics, time pressure, conflicting instructions)
• Learning agility and structured thinking: less commonly measured at screening stage but valuable for in-depth selection
• Language proficiency: relevant for Dutch and English-medium firms hiring internationally

COTAN and test quality in the Netherlands

In the Netherlands, the quality benchmark for psychometric instruments is the COTAN evaluation system, maintained by the Nederlands Instituut van Psychologen (NIP). COTAN assesses tests against seven criteria: Theoretical Basis, Quality of Test Materials, Quality of the Manual, Normative Data, Reliability, Construct Validity, and Criterion Validity (NIP, May 2022).

A critical detail that vendor sales material often omits: each criterion is rated independently. A test can receive a positive rating on Reliability while receiving a lower rating on Normative Data. This means "COTAN-evaluated" or even "COTAN-positive" claims require scrutiny. Ask vendors to provide the full COTAN report for any instrument they propose, not a headline rating.

Norms recency is particularly important. COTAN requires that norm groups must not be older than 10 years unless the publisher provides empirical evidence that they remain valid (NIP, May 2022). For graduate assessments, this means asking: when were the graduate-level norms last updated, and do they include Dutch law graduates or a comparable professional cohort?

GDPR and the EU AI Act

Under the EU AI Act, recruitment and HR AI systems that screen, rank, score, or filter job applicants are classified as high-risk and require active human oversight so that automated tools do not act as the final decision-maker (Forta regulatory landscape map). This applies whether the system is a scoring algorithm, an AI-driven chatbot that filters eligibility, or a platform that generates ranked shortlists.

For GDPR compliance, the requirements include documented consent at each processing stage, defined retention periods, a data processing agreement (DPA) with the vendor, and a clear record of where personal data is stored and by whom it can be accessed. For firms using AI-driven intake or screening tools, a Data Protection Impact Assessment (DPIA) is advisable.

Vendor shortlist: profiles for Dutch law-firm graduate hiring

The following platforms represent three distinct categories: enterprise psychometrics suites, assessment content marketplaces, and AI-integrated workflow platforms.

Selection Lab

Selection Lab is an AI-enhanced assessment platform that combines a conversational intake tool (SmartChat) with a modular library of assessments covering cognitive ability, situational judgement, soft skills, hard skills, language proficiency, and culture fit. According to the company's published product documentation (May 2026), SmartChat responds within 10 seconds, operates via WhatsApp or webchat, and handles CV checking, eligibility screening, and interview scheduling, with outputs reported directly into the recruiter's ATS.

For graduate law hiring specifically, the platform's ability to deploy a timed SJT alongside a verbal reasoning measure within a single candidate session, without requiring candidates to log into a separate portal, reduces the friction that typically causes drop-offs in legal graduate funnels. According to client KPIs reported in Selection Lab's 2026 product deck, the platform has been associated with 27% fewer candidate drop-offs, 15 minutes saved per applicant in recruiter processing time, and 21% lower early turnover. These are internally reported client metrics and should be discussed in the context of a firm's own baseline.

On compliance, Selection Lab states that personal data is stored in Frankfurt, consent is requested for each processing purpose with defined retention periods, results are shown to the candidate before being shared with the employer (with a second consent request at that point), and local LLMs are used to remove personal information from conversation logs. The company also states alignment with EU AI Act requirements. Firms should request the DPA and DPIA support documentation to verify these claims independently.

Implementation is stated as 2 to 10 weeks, with ongoing check-ins and quarterly reviews included in the support model (Selection Lab Main Deck, May 2026).

SHL

SHL is one of the most established enterprise psychometrics providers globally and offers a broad catalogue of cognitive, personality, and situational judgement assessments. Its Occupational Personality Questionnaire (OPQ) and Verify cognitive battery are widely used in UK and European law firm recruitment, including by firms with Amsterdam offices. SHL provides extensive technical documentation. Buyers should verify whether Netherlands-specific norm groups are available for the instruments proposed, and confirm COTAN evaluation status directly with SHL's Dutch team. ATS integrations are available but integration depth varies by ATS vendor.

Ixly

Ixly is a Netherlands-based assessment provider with Dutch-language tools and familiarity with the Dutch HR and legal recruitment market. Its assessments include cognitive tests and personality instruments, and its documentation is more likely to reference COTAN evaluation directly than international vendors. For firms prioritising Dutch-language candidate experience and local compliance support, Ixly is worth including in a shortlist evaluation. Confirm norms recency and the scope of any COTAN ratings before proceeding.

TestGroup

TestGroup offers an accessible online assessment platform used by a range of Dutch employers, including some professional services firms. It includes personality instruments and cognitive tests with a largely self-service setup model. The platform's documentation of COTAN status and Netherlands-specific norm groups should be verified during the demo process. It is best suited to organisations with lower psychometric complexity requirements or as a supplementary tool rather than a primary legal graduate selection instrument.

TestGorilla

TestGorilla is an assessment content marketplace that provides a large library of skills-based tests, including role-specific assessments and cognitive measures. It's accessible and quick to deploy, which suits high-volume screening. However, the depth of psychometric validation documentation (particularly for Dutch norm groups or COTAN-referenced instruments) is less extensive than enterprise providers. Firms using TestGorilla for graduate law hiring should treat it as a screening-stage tool and supplement it with validated psychometric instruments for in-depth selection.

Questions to ask on a vendor demo

• Which specific assessments do you recommend for junior lawyer roles, and what is the evidence base?
• Can you provide the full COTAN evaluation report (all seven criteria) for each instrument proposed?
• When were the relevant norm groups last updated, and do they include Dutch law or professional services graduates?
• Where is candidate data stored, and who are your subprocessors?
• Can you provide a DPA, a DPIA support pack, and your data retention schedule?
• How does your platform support human oversight requirements under the EU AI Act?
• What ATS integrations do you support natively, and what does the integration setup involve?
• What does your implementation timeline and ongoing support model look like?

Feature comparison at a glance

Two assessment workflows for graduate legal hiring

Track A: screening to shortlist (high volume, low friction)

This track is suitable for the early funnel, where dozens or hundreds of applications require efficient triage without compromising fairness.

1. Automated intake and eligibility screening: deploy a conversational intake (via WhatsApp or webchat if supported) to capture eligibility criteria, confirm language requirements, and handle scheduling. Any AI-driven filtering at this stage must be documented for human oversight purposes under the EU AI Act.
2. Cognitive/verbal reasoning measure: a timed test of analytical or verbal reasoning (15 to 25 minutes). Use a validated instrument with a graduate norm group. Set a minimum threshold for progression rather than a strict rank-order cut-off, to reduce false negatives at the margins.
3. Situational judgement test (SJT): present 10 to 15 professional scenarios relevant to junior lawyer contexts (client interaction, ethical dilemmas, workload prioritisation). Score as a secondary filter or use as a ranking input alongside cognitive scores.
4. Structured interview invitation: top candidates receive automated scheduling via ATS. Borderline candidates (those scoring within a defined range below the cut-off on either measure) should be reviewed by a recruiter before exclusion, with the review documented.

Track B: in-depth selection (smaller cohort, higher signal)

For the final 15 to 30 candidates, a more comprehensive approach is warranted.

1. Expanded SJT or work-sample simulation: a longer scenario-based exercise reflecting realistic legal work (e.g., analysing a short brief, identifying issues, prioritising actions). This can be conducted online or in an assessment centre.
2. Targeted hard-skill or knowledge check (where justified): for roles requiring specific knowledge (e.g., commercial law fundamentals), a structured knowledge test may add incremental validity. Use with caution: ensure norm groups are appropriate.
3. Structured interview using assessment evidence: interviewers receive a summary of each candidate's assessment profile. Questions are drawn from competency frameworks and probe areas where scores were borderline or inconsistent.
4. Documented final decision: record the rationale for each hire and non-hire decision, referencing assessment evidence. This documentation is a requirement under EU AI Act governance frameworks where AI tools contributed to the scoring or ranking. A human decision-maker must be accountable for the final outcome.

For GDPR compliance at both stages, candidates should receive a clear explanation of what is being measured and why before completing any assessment. Retention periods for assessment data should be defined in the DPA and communicated to candidates at the point of consent.

Running a pilot before committing

Pricing structures vary significantly across vendors (per-assessment, per-user, volume-based, or enterprise licence), and it's rarely productive to compare list prices without understanding the scope. A pilot is a more reliable basis for evaluating value.

A well-scoped pilot for a Dutch law firm should include:

• One graduate intake cohort (ideally 50 to 150 candidates)
• Two to three assessment instruments (cognitive + SJT minimum)
• ATS integration configured and tested
• Reporting outputs reviewed against your existing shortlisting criteria
• Candidate completion and drop-off rates tracked
• At least one recruiter review of the analytics dashboard

Ask the vendor to provide test administration guidance, sample candidate-facing reports, and a sample recruiter-facing analytics summary before the pilot begins. Confirm that the hosting environment and data retention policies are documented and consistent with your firm's GDPR obligations.

Selection Lab, for example, states a go-live timeline of 2 to 10 weeks and includes check-ins and quarterly reviews as part of its implementation model (Selection Lab Main Deck, May 2026). That kind of structured onboarding is worth verifying with other vendors on your shortlist.

What good looks like: evaluating outcomes after the pilot

A credible vendor case study for legal graduate hiring includes cohort size, the assessment stage design, candidate completion rates, and at least one downstream outcome metric (quality-of-hire proxy, interview-to-offer conversion, or early retention). Be sceptical of case studies that report only platform usage metrics without any link to hiring quality.

For reference, Selection Lab's client KPIs (internal metrics from their 2026 product deck) include 15 minutes saved per applicant, 27% fewer candidate drop-offs, and 21% lower early turnover. These figures are specific enough to be used as a benchmark when scoping your own pilot outcomes, though they reflect client-reported data and will vary by context.

For Dutch law firms evaluating platforms across the digital assessment landscape, the most productive next step is to compile a short requirements document covering: the assessments you want to use, your ATS and integration requirements, your data protection obligations, your EU AI Act governance posture, and the cohort sizes and timelines for your next graduate intake. Share it with two or three vendors from this shortlist, and use the demo question checklist above to evaluate their responses. The right platform for your firm will be the one that can answer those questions with documentation, not just assurances.