D&I / belonging measurement template

D&I and belonging survey form — measure composition and inclusion without compliance gaps

A D&I survey form built for the regulatory reality of 2026 — GDPR Article 9 and LGPD Article 11 compliant handling of sensitive demographic data, re-identification thresholds for intersectional analysis, voluntary self-identification with "prefer not to say," and integration with Lattice DEIB, Culture Amp Diversity Inclusion Index, Workday Belonging.

Free — included in every plan

D&I and belonging survey form — measure composition and inclusion without compliance gaps

Live preview — try the fields below.

No fields to preview.

Who this template is for

D&I surveys are the most regulatory-sensitive HR template in the market because they collect demographic data that falls into sensitive-data categories under every major privacy framework: EU GDPR Article 9 (special category data — race, ethnicity, religion, sexual orientation, health, political opinions all require explicit consent and specific lawful basis), Brazilian LGPD Article 11 (dado pessoal sensível, with the same categories plus enhanced protections), California CPRA (sensitive personal information), US EEOC (regulated employment data with specific category requirements), UK GDPR (post-Brexit framework that closely mirrors EU GDPR). The compliance overhead is real and the design choices matter. Two distinct measurement dimensions need different structures. First, demographic composition — what the workforce demographic profile actually is, captured through voluntary self-identification questions on race, ethnicity, gender, age, sexual orientation, disability status, veteran status (US-specific), socio-economic background (UK-specific), national origin, and other categories relevant to the org's DE&I framework. These questions must be voluntary with a "prefer not to say" option, with explicit consent capture, with data-handling transparency communicated to employees before they take the survey, and with re-identification protection at the analysis stage. Second, inclusion and belonging perception — how the workforce experiences the company across those demographic dimensions, measured through validated frameworks like the Culture Amp Diversity Inclusion Index, Lattice DEIB scorecard, Microsoft Viva Glint inclusion module, or custom belonging questions based on Amy Edmondson's psychological-safety research and broader belonging-in-the-workplace literature. The intersectional analysis — how Black women experience the company differently from Black men or white women, how LGBTQ+ engineers experience the company differently from LGBTQ+ sales professionals or straight engineers — is where the most-actionable signal lives, but it's also where the re-identification risk is most acute because intersectional cuts produce small N's. "The one Black female engineer" can't be anonymous if the survey reports a separate "Black female engineer" row in the results. Minimum-N thresholds for reporting (typically 5-10 per intersection, sometimes 15+ for highly-identifiable demographic categories) are essential protective patterns. The 2023-2026 legal landscape adds new complexity: US SCOTUS Students for Fair Admissions v. Harvard (2023) creates uncertainty for some DEI practices that explicitly target specific demographic groups; the EU Pay Transparency Directive (2023) requires demographic pay-gap reporting that depends on accurate demographic-composition data; Brazil continues to strengthen DEI compliance under Lei 13.146 (Estatuto da Pessoa com Deficiência) and the broader PCD framework; and the backlash against "DEI" branding has shifted some companies to "Belonging" or "Inclusive Culture" framing while keeping the underlying measurement work intact. This template gives you the structure that addresses all these realities — compliant sensitive-data handling, voluntary self-identification with re-identification protection, validated inclusion/belonging measurement, intersectional analysis with appropriate thresholds, and integration with the HR-tech stack.

From survey deployment to action plan with intersectional analysis in 6-8 weeks

HR or the DE&I team deploys the survey via email or in-app prompt to the eligible employee population, with clear communication about three things. First, anonymity guarantees — responses stored separately from identifying data, minimum-N thresholds for demographic cut reporting (typically 5-10 per category), demographic-segmentation breakdowns only above the re-identification floor, and the explicit data-handling architecture (sensitive data stored encrypted, HR access only to aggregated results, no manager access to individual responses or to small demographic-cut breakdowns). Second, voluntary self-identification — every demographic question has a "prefer not to say" option, employees can skip the demographic section entirely and still complete the inclusion/belonging questions, and the survey explicitly states that demographic data is not used for individual decisions (hiring, performance, compensation). Third, the action commitment — what the team will do with the results, communicated before the survey deploys so employees can decide whether to invest the time. The survey runs for a defined window (typically 3 weeks) with response-rate visibility for HR but no individual identification. Reminders fire to non-responders without identifying who responded vs. who didn't. On close, the responses aggregate to organization-level reporting (composition + inclusion summary), to per-demographic-cut reporting where N permits (gender, race-ethnicity, age, tenure, location, role-type, sexual orientation, disability status where the survey collected and where N allows), and to intersectional analysis at the org-wide level (intersectional cuts at team-level often violate the re-identification floor and should not be reported). The HR or DE&I team analyzes the results, identifies the 3-5 themes that warrant action (typically: specific intersectional groups reporting lower belonging than the org-wide average, specific career-equity gaps surfacing, specific psychological-safety patterns), and develops an action plan with named owners and timelines. The action plan is communicated back to the workforce within 6-8 weeks of survey close — the closing-the-loop step matters more for DE&I surveys than any other HR survey because the perception of "performative DEI without follow-through" is corrosive to trust in a way that other survey types don't trigger.

What's included

Each section below maps to one of the three structural pillars: demographic composition with voluntary self-identification and "prefer not to say," inclusion/belonging measurement with validated framework questions, and re-identification protection at the analysis stage with minimum-N thresholds for demographic cuts. The compliance posture (consent, data-handling transparency, sensitive-data treatment) is built into the form structure rather than bolted on as an afterthought.

Built for the org sizes where D&I measurement is operationally meaningful

  • Mid-size B2B (50-500 employees, growing DE&I maturity)

    The primary use case. Mid-size orgs have enough employees that demographic composition starts to produce reliable signal (below 50 employees, the N's are too small for meaningful demographic analysis without re-identification risk) and small enough that the DE&I program is still being built rather than mature. The HR-tech stack typically uses Culture Amp Diversity Inclusion Index, Lattice DEIB, 15Five DEIB module, BambooHR with custom DE&I survey, or standalone DE&I specialists (Affirmity, Diversio, HiBob DEIB). The measurement cadence is typically annual or biennial — DE&I surveys are heavier than engagement surveys (longer, more sensitive, require more action follow-through) and more frequent cadence produces survey-fatigue without proportional value. For Brazilian B2B mid-size (RD Station, Pipefy, Conta Azul, Bling, Omie, Movidesk, Octadesk, ContaWise scale; Hotmart, Eduzz, Kiwify in scale-up tier; Stone before scale, Loft, QuintoAndar pre-IPO scale), the DE&I framework is typically built around the Brazilian demographic reality: race classification using IBGE categories (preta, parda, branca, amarela, indígena), PCD (pessoa com deficiência) tracking against the Lei 8.213/91 quota requirement, gender identity including specific Brazilian context (mulheres trans, travestis as recognized categories), and socio-economic background tracking (which is more actionable in Brazil than the US given Brazilian inequality patterns). The HR-tech stack for Brazilian DE&I is typically Mereo with DE&I survey support, Solides DE&I module, Convenia DE&I, or the global tool deployed with Brazilian localization. For Spanish mid-size B2B (Holded, Quipu, TravelPerk, Factorial HR itself, Cabify, Glovo in scale, Wallapop, Devo, Capchase), Factorial HR has a strong DE&I module for the Spanish-market reality which includes specific tracking around discapacidad (disability) under the Ley General de Derechos de las Personas con Discapacidad and broader Spanish equality-of-treatment framework.

  • Enterprise (500+ employees, sophisticated DE&I program with CDO)

    Different operational scale. Enterprise DE&I programs run on Workday Belonging and Diversity, Qualtrics EmployeeXM with DE&I module, Microsoft Viva Glint inclusion module, Culture Amp Enterprise DEIB, Lattice Enterprise DEIB, with dedicated Chief Diversity Officer (CDO) leadership and structured intersectional analysis. The pay-equity analysis is typically embedded in the DE&I program at enterprise scale — combining demographic-composition data with compensation data to surface pay gaps that need remediation. The 2023 EU Pay Transparency Directive requires companies with 100+ employees in EU member states to publish gender pay-gap data starting in 2027 (with phased implementation), with similar requirements expanding to ethnicity and other demographic dimensions in some member states. Enterprise DE&I programs typically pair the survey with broader workforce-analytics: representation across role levels (the typical pattern: demographic diversity at entry-level but narrowing demographically at senior levels — the broken-rung problem), promotion velocity by demographic, attrition by demographic, hiring-pipeline composition by demographic. For Brazilian enterprise (Vale, Petrobras, Itaú, Bradesco, Banco do Brasil, Magazine Luiza, JBS, Ambev, Embraer, TOTVS, BTG Pactual, Stone, MercadoLivre, Hotmart, XP Inc, Pagseguro, Natura&Co, Globo), the DE&I program typically embeds the PCD quota tracking (Lei 8.213/91 requires Brazilian companies with 100+ employees to hit 2-5% PCD workforce depending on company size) as a core operational metric alongside the broader DE&I measurement. For Spanish enterprise (Telefónica, Iberdrola, Santander, BBVA, Inditex, Repsol, Naturgy, Mapfre, El Corte Inglés, Movistar), the discapacidad quota under the Spanish Ley General de Derechos de las Personas con Discapacidad and the gender pay-gap reporting under the EU Pay Transparency Directive together drive the DE&I measurement requirement at enterprise scale.

  • Startups (often weak DE&I discipline, homogeneity baked in early)

    Different dynamics than mid-size and enterprise. Startups (under 50 employees) often have weak DE&I discipline because the org is small enough that founder intuition replaces formal measurement, the demographic homogeneity of founding teams systematically transfers to early hires through referral networks, and the formal DE&I survey feels disproportionate to the operating mode. The reality is that startups have the highest leverage on DE&I outcomes because the early demographic patterns scale persistently — a startup that's 90% male in the first 20 employees will struggle to reach gender parity at 200 employees, while a startup that prioritizes balanced demographics in the first 20 employees can scale that balance through growth. The DE&I survey at startup scale is typically a lighter-weight pulse (5-10 demographic questions, 5-10 inclusion questions) run annually, with the small-N reality acknowledged in the analysis (often the survey produces directional signal rather than statistically-significant breakdowns). For Brazilian startups (Y Combinator graduates, Cubo Itaú accelerator startups, ACE Startups portfolio, the early-stage Brazilian SaaS ecosystem), the PCD reality starts at a small scale because Brazilian companies with 100+ employees hit the quota requirement; startups planning to grow past 100 should start the PCD inclusion work earlier rather than retroactively. For Spanish startups (Y Combinator graduates from Spain, South Summit and 4YFN ecosystem), the discapacidad quota under Spanish law and the gender-balance expectations of Spanish business culture create similar earlier-investment incentives.

  • Public sector (legally-mandated tracking, civil-service framework)

    Different operational constraints. Public-sector DE&I measurement is typically governed by specific legal requirements that vary by jurisdiction. US federal: EEOC's Form 100 (EEO-1) reporting, plus specific agency-level requirements; state and local governments have their own analogues. EU member states: varying requirements layered on top of EU Pay Transparency Directive, with strongest implementations in Nordic countries. Brazilian federal: SIGEPE captures civil-servant demographics with specific reporting requirements; the PCD quota under Lei 8.213/91 applies; the Estatuto da Igualdade Racial (Lei 12.288/2010) creates specific obligations for federal entities. Spanish public sector: the EBEP framework plus specific autonomous-community requirements; the Ley General de Derechos de las Personas con Discapacidad sets PCD quotas. Public-sector DE&I surveys often have stricter privacy and union-coordination requirements than private sector — unionized public-sector environments may have the union's approval required for survey design or for use of survey results. The public records consideration also differs from private sector — some public-sector survey results may be subject to disclosure under freedom-of-information laws depending on aggregation level and jurisdiction.

  • Specific industry contexts (tech, finance, healthcare with industry-specific demographic challenges)

    Different demographic patterns and different DE&I priorities by industry. Tech industry: gender imbalance in technical roles (engineering, infrastructure, security), racial-ethnic imbalance especially at senior levels, age discrimination patterns (the "young company" culture often skewing against older candidates), neurodivergent representation increasingly tracked. Finance: similar gender and racial patterns at senior levels, with additional regulatory pressure from EU Pay Transparency Directive and from US OCC/Federal Reserve diversity reporting requirements. Healthcare: gender imbalance in specific roles (nursing female-dominated, surgery male-dominated), racial-ethnic patterns in patient-provider matching that affect quality-of-care, language-fluency considerations for serving diverse patient populations. Each industry has specific DE&I framework elements that generic D&I surveys don't capture well — the survey should accommodate industry-specific dimensions while maintaining the core measurement structure. For Brazilian industries specifically, the financial sector under Bacen has begun requiring gender pay-gap disclosure; the tech sector competing for the Brazilian talent pool with US-based companies faces specific compensation and visa-pathway pressures that affect DE&I patterns; the healthcare sector under ANS regulation has specific patient-population diversity considerations.

  • Companies running annual D&I report or ESG disclosure

    Specific use-case. Companies that publish annual D&I reports or ESG disclosures use the DE&I survey as the primary data source. The S&P 500 and FTSE 100 companies almost universally publish annual D&I metrics; the practice is spreading to mid-size B2B companies as part of investor relations and recruiting-employer-brand strategy. The DE&I survey for these companies must produce the specific metrics required by the disclosure framework — typically gender representation, racial-ethnic representation, leadership-level breakdown, pay-equity analysis, attrition by demographic. The disclosure compliance reality drives survey design — the questions need to map to the disclosure framework's specific metrics, which often means EEOC categories in US-deployments, EU Sustainability Reporting Directive (ESRD) categories in EU-deployments, and emerging frameworks in other jurisdictions. For Brazilian companies publishing ESG reports (B3-listed companies, especially those in the ISE Sustainability Index), the demographic disclosure typically includes PCD percentage, gender at leadership levels, racial composition using IBGE categories, and increasingly LGBTQ+ representation. For Spanish companies in the IBEX 35 or publishing under EU CSRD (Corporate Sustainability Reporting Directive), the DE&I disclosure is mandatory for in-scope companies starting in 2024-2025 with phased implementation.

Configure the form to your demographic framework, inclusion-measurement choice, and disclosure obligations

Start with the regulatory compliance layer because it determines the rest of the design. For EU GDPR Article 9 sensitive data, explicit consent is required before any sensitive demographic question, with consent stored as audit trail, with right-to-withdraw clearly communicated, with data-handling architecture transparent. For Brazilian LGPD Article 11, the same explicit-consent pattern applies with the additional consideration that race classification uses IBGE categories (preta, parda, branca, amarela, indígena — these are the official Brazilian categories used in census and government data; deviating from them creates analytical incomparability). For California CPRA sensitive personal information, similar consent pattern. For US EEOC compliance, the categories should map to EEO-1 categories at minimum (Hispanic or Latino; White; Black or African American; Native Hawaiian or Pacific Islander; Asian; American Indian or Alaska Native; Two or more races) for reportable demographic data. Add "prefer not to say" as a non-optional answer choice for every demographic question — voluntary self-identification is required under every major privacy framework. Choose the inclusion-measurement framework. Culture Amp Diversity Inclusion Index is the most-used validated framework for B2B mid-market. Lattice DEIB scorecard is the standard for Lattice-customer orgs. Microsoft Viva Glint inclusion module integrates with Microsoft 365 stack. Custom belonging questions based on Amy Edmondson's psychological-safety research work for orgs that want to design their own framework. Set the re-identification thresholds. Minimum N for demographic cut reporting is typically 5-10 (some orgs use 15+ for highly-identifiable demographic categories). Below the threshold, the demographic cut isn't reported separately — the data either aggregates to a higher level or the cut is omitted from the report. Intersectional analysis (e.g., race × gender × tenure) should typically be reported only at organization-wide level because team-level intersectional cuts almost always violate the threshold. Integrate with the HR-tech stack: Workday Belonging and Diversity, Qualtrics EmployeeXM DE&I, Microsoft Viva Glint inclusion, Culture Amp Diversity Inclusion Index, Lattice DEIB, BambooHR DE&I, 15Five DEIB for international; Mereo DE&I, Solides DE&I, Convenia DE&I for Brazilian-market; Factorial HR DEI, Personio Spain, Sage HR Spain, Bizneo, Cezanne HR for Spanish-market. For Brazilian deployments, LGPD-compliant handling of sensitive data with consent capture, separate processing channel, and data-minimization principles built in. Translate into the languages of your workforce.

Diversity, equity, and inclusion survey FAQ

Layered approach. First, explicit consent for sensitive demographic data — the survey opens with consent capture that explains what's being collected, why, how it will be used, who will see it, and how the employee can withdraw consent later. The consent capture is stored separately as audit trail. Second, data-minimization — only collect demographic data you'll actually use. The temptation is to collect every demographic dimension for completeness, but each additional dimension creates additional sensitive-data handling responsibility and re-identification risk. Most orgs end up with race/ethnicity, gender, age, sexual orientation, disability status, plus 1-2 org-specific dimensions (veteran status in US, socio-economic background in UK, PCD specifics in Brazil). Third, separate-processing — sensitive demographic data should be stored and processed separately from non-sensitive HR data, with stricter access controls and shorter retention periods. The data architecture should clearly separate the demographic data store from the engagement-survey data store from the performance-review data store. Fourth, third-party processor considerations — if you're using a vendor (Culture Amp, Lattice, Workday), the data-processing agreement (DPA) must cover the sensitive-data handling specifically, with sub-processor flow-down for any further sharing. For Brazilian deployments specifically, the LGPD Art. 11 sensitive-data category triggers additional ANPD (Autoridade Nacional de Proteção de Dados) scrutiny on data handling. For EU deployments, the GDPR Article 9 sensitive-data category may require Data Protection Impact Assessment (DPIA) depending on the data-processing scale and risk.
Three structural protections. First, minimum-N thresholds for demographic cut reporting — typically 5 as the absolute minimum, 10 as the standard threshold, 15+ for highly-identifiable categories (specific intersections, small demographic groups in homogeneous orgs). Below the threshold, the demographic cut isn't reported separately. The cut either aggregates up to a parent category (e.g., specific Hispanic-Latino sub-groups roll up to Hispanic-Latino aggregate if the sub-groups don't hit threshold) or is omitted from the report entirely. Second, intersectional analysis at organization-wide level only — team-level or department-level intersectional cuts almost always violate the threshold (a 50-person team with one Black female engineer can't have anonymous results for the Black-female-engineer intersection at team level), so intersectional analysis should be reported only when N is large enough. Third, response-pattern aggregation — beyond raw count thresholds, the response patterns within small cuts can re-identify even when N is technically above threshold. The example: when 6 LGBTQ+ employees in a small org all give the same response to a specific inclusion question, the pattern can be reverse-engineered to identify their individual responses through correlation with other survey signals. The mitigation is reporting demographic cuts only on aggregate metrics (e.g., "LGBTQ+ employees report belonging at X% versus org-wide Y%") rather than per-question response distributions for small cuts. The protections compound — minimum-N + intersectional restriction + aggregate-only reporting for small cuts together produce robust re-identification protection.
Yes, with "prefer not to say" as a required answer option for every demographic question. Voluntary self-identification is required under every major privacy framework (GDPR, LGPD, CCPA, CPRA), is best-practice under US EEOC guidance, and is the right pattern under any DE&I framework. The reason: forced demographic disclosure produces gaming (employees don't trust the system and provide inaccurate data, which biases the analysis) and creates legal exposure under sensitive-data frameworks. The pattern that works: each demographic question offers the standard category choices, "prefer not to say," and (for some categories) "other — please specify" with text capture. The survey lets employees skip the entire demographic section if they choose, with their inclusion/belonging responses still captured. The voluntary nature should be explicitly communicated before the demographic section starts: "the following questions are voluntary, you can skip any question, and your participation in this section is not used for individual decisions about hiring, performance, or compensation." The voluntary participation rate matters too — if a significant portion of employees skip the demographic section (typically 20-40% will choose "prefer not to say" or skip entirely in a healthy program), the demographic-cut analysis needs to acknowledge the response-rate gap and not over-interpret the available data. For Brazilian and Spanish-speaking workforces specifically, the voluntary self-identification communication should be in the employee's native language with clear explanation of the categories — direct translation of US-centric category names without cultural adaptation produces lower-quality demographic data.
Choose a validated framework rather than designing custom inclusion questions from scratch. Culture Amp Diversity Inclusion Index is the most-used validated framework for B2B mid-market — 10-15 questions measuring belonging, voice, authenticity, career equity, and broader inclusion dimensions, with extensive benchmark data. Lattice DEIB scorecard is the standard for Lattice-customer orgs with similar dimensional coverage. Microsoft Viva Glint inclusion module integrates with Microsoft 365 stack and uses psychological-safety-grounded questions. Custom belonging questions based on Amy Edmondson's psychological-safety research (Harvard Business School, 1999-present) work for orgs that want to design their own framework with academic grounding. The choice matters because bad inclusion-measurement framework produces useless data — vague questions like "do you feel included?" produce equally vague responses, while specific behavioral-pattern questions ("in the last month, my opinion was solicited in a team meeting") produce actionable data. The mainstream pattern: use a validated framework (Culture Amp, Lattice, Microsoft, or equivalent) rather than designing custom from scratch, with optional 2-3 org-specific questions added to the standard framework for context. The benchmark comparison (how your org's inclusion scores compare to industry benchmarks) is valuable when using validated frameworks because the framework providers maintain comparable benchmark data.
The closing-the-loop step matters more for D&I surveys than any other HR survey because the perception of "performative DEI without follow-through" is corrosive to trust in a way that other survey types don't trigger. The pattern: within 6-8 weeks of survey close (slightly longer than engagement-survey close-the-loop because D&I analysis is more complex), communicate four things to the workforce. First, the participation reality — what percentage of the workforce responded, what percentage of respondents completed the demographic section, what the analysis can and can't say given the response patterns (don't overclaim from data the survey can't support). Second, the top-level findings — composition summary (where the workforce is across the demographic dimensions surveyed, where leadership composition differs from overall composition, where the broken-rung pattern shows up) and inclusion summary (which demographic groups report higher vs. lower belonging, where the intersectional patterns surface). Third, the action plan with named owners and timelines — what specific actions the team is taking in response to the findings, by when, who's accountable. Fourth, the measurement plan — how the next survey will track whether actions worked, and the cadence of the next survey. The four-part communication is more comprehensive than the standard engagement-survey close-the-loop because D&I survey trust requires more reassurance about authentic intent. The biggest failure mode: running a D&I survey, communicating findings, but not committing to specific actions — this is the performative pattern that erodes trust permanently. The second-biggest failure mode: committing to specific actions but not delivering them — every D&I survey after that one starts with the question "did anything change since the last survey?" and a no-progress answer destroys participation rates. For Brazilian and Spanish-speaking workforces specifically, the closing-the-loop communication should be in native language with cultural context.

Ready to build forms that work for you?

Create your first form in minutes. Your submissions will thank you.

Be first in lineLimited early accessSet up in 2 minutes