Deliverable D2.5 · Framework validation report

Validated by the sector, refined by its feedback

In September 2026, 102 professionals answered our validation survey and 34 discussed the ESG4GLAM Competency Framework in ten focus groups in Finland, Latvia, Italy and Serbia. Here is what they told us, and what changes next.

Download the report (PDF) Explore the results
91.2%consider the framework a valid reference tool
93 of 102 endorse it 70 without reservation
  • 102survey responses
  • 70+organisations
  • 12countries
  • 10focus groups
  • 34focus group participants
  • 84change requests logged

How we validated

Two methods, one set of criteria

The survey shows how widely the framework is endorsed. The focus groups show why, and what should change. Both tested the 31 competencies, four domains and five levels against the same five criteria.

Part 1 · Breadth

Online validation survey

  • 102 valid responses from more than 70 organisations
  • 12 countries, most from Latvia, Finland, Italy and Serbia
  • Five yes/no criteria plus training priorities and an overall verdict
  • Data collected in September 2026
Part 2 · Depth

Partner-led focus groups

  • 10 groups run by five partners, two each
  • 34 participants (moderators and note-takers excluded)
  • 5 in person and 5 online, in four countries
  • Held in September 2026, with 84 change requests logged
  1. 1Completeness and relevance
  2. 2Clarity for non-specialists
  3. 3Levels L1–L5 and career paths
  4. 4Practical applicability
  5. 5Basis for the 35-hour course

The overall verdict

Is it a valid reference tool for European GLAM?

93 of 102 respondents (91.2%) say yes. This meets the 90% approval target set in the project application. Of these, 70 endorse the framework without reservation and 23 subject to adjustments.

36 34 23 9 90% target
  • 36 · 35.3% “Yes, I fully endorse it” Wave A
  • 34 · 33.3% “Yes, a clear, reliable and comprehensive standard” Wave B
  • 23 · 22.5% “Yes, but it requires adjustments” Wave B
  • 9 · 8.8% Do not endorse it in its current form
The answer options were refined during data collection to separate unconditional from conditional endorsement (Wave A: 44 responses, Wave B: 58); the two versions were harmonised for analysis. Source: ESG4GLAM framework validation survey, 102 respondents, data collected in September 2026.
93.2%

endorsement among the 73 stakeholders outside the consortium, against 86.2% among partner staff. The positive result does not come from the project’s own networks.

57.8%

answered yes to all five criteria, and 78.4% to at least four. Only 9.8% were positive on two or fewer.

Results by criterion

Where the framework scores highest

Share of “yes” answers to each validation question, with the 95% confidence interval. Pick a group to see how its view compares with the 90% benchmark.

All
Organisation
Country
Affiliation

All respondents · n = 102

  • The framework is a useful basis for the 35-hour training course

    94.1%96 of 102

  • Overall, it is a valid reference tool for the European GLAM sector

    91.2%93 of 102

  • The 31 competencies and 4 domains cover what GLAM professionals need

    88.2%90 of 102

  • Descriptions and examples are clear for non-ESG specialists

    82.4%84 of 102

  • The 5 levels (L1–L5, EQF 3–8) match real roles and career paths

    82.4%84 of 102

  • My institution can apply the framework in practice

    74.5%76 of 102

Practical applicability is the main challenge: 26 respondents (25.5%) say time, budget or staff are too high a barrier. Yet 22 of them still see the framework as a sound basis for training.

View data as a table
Positive answers by criterion and respondent group (yes answers / group size)
GroupnTraining basisValid toolCompleteClearLevelsApplicable
All respondents102969390848476
Library38373637333532
Museum26252320171815
Archive6665644
Other32282828282725
Latvia34343233313128
Finland23201817141513
Italy22192018191718
Serbia14141414131413
Other countries9998774
Partner staff29262523222420
External stakeholders73706867626056

Confidence intervals: Wilson score method. The survey is not a probability sample, so intervals show the precision of the estimates, not representativeness. Source: ESG4GLAM framework validation survey, 102 respondents, data collected in September 2026.

Who took part in the survey

The whole GLAM ecosystem

By type of organisation n = 102
102respondents
  • Library3837.3%
  • Other3231.4%
  • Museum2625.5%
  • Archive65.9%
By country of organisation n = 102
  • Latvia
    34
  • Finland
    23
  • Italy
    22
  • Serbia
    14
  • 8 other countries
    9

Spain, Estonia, Sweden, Poland, Czechia, Germany, Bosnia and Herzegovina and Australia.

Inside the “Other” group n = 32

Not peripheral, but the wider cultural value chain the project set out to reach:

  • 14Higher education and research
  • 7Cultural associations, foundations, public bodies
  • 6Art market and fine-art services, incl. a gallery
  • 4Consultancies
  • 1Independent professional
View data as a table
Respondents by country and type of organisation
CountryLibraryMuseumArchiveOtherTotal
Latvia3031034
Finland4140523
Italy3141422
Serbia0301114
Other countries (8)15129
Total3826632102

Country was derived from the organisation name and e-mail domain, so read it as a close approximation. Source: ESG4GLAM framework validation survey, data collected in September 2026.

Training priorities

Which domain should the course tackle first?

Share of respondents naming each domain (96 answered, more than one choice allowed). Both the survey and the earlier needs analysis put the Cross-Cutting foundation first.

  • I · Cross-Cutting51.0%

    ESG literacy, links between E, S and G, digital sustainability and responsible AI

    49 of 96 · gap rank 1 of 4
  • III · Social41.7%

    Inclusive access, community co-creation, accessible communication, staff wellbeing

    40 of 96 · gap rank 3 of 4
  • II · Environmental25.0%

    Energy and climate control, sustainable exhibitions, emergency preparedness

    24 of 96 · gap rank 2 of 4
  • IV · Governance24.0%

    Ethics, transparency, ESG in planning and reporting, data management

    23 of 96 · gap rank 4 of 4

Social ranks second although its measured gap was small: libraries name it as often as Cross-Cutting (18 each). Priorities follow each institution’s public mission as well as its gaps, so the course will offer a common foundation followed by differentiated pathways.

Gap rank: mean competency gap score from the ESG4GLAM needs analysis (importance minus current proficiency). Source: ESG4GLAM framework validation survey, data collected in September 2026.

Part 2 · Focus groups

Ten conversations across four countries

Five partners each ran two sessions in September 2026. Every dot below is one participant; moderators and note-takers are not counted.

Finland

13

Humak HU-1 in person · HU-2 online

Forum Marinum FM-1 · FM-2, in person

Museum training, galleries and artist associations, city libraries, cultural-policy research, museum curators, educators and audience engagement.

Latvia

10

Jelgava City Library JC-1 · JC-2, online

Library, museum and archive professionals from national, municipal and regional bodies, including senior and branch managers.

Serbia

6

University of Novi Sad UN-1 · UN-2, in person

Curriculum and micro-credential designers, a festival PR coordinator and other cultural professionals.

Italy

5

Mazzini Lab MZ-1 · MZ-2, online

An expert outside view: sustainability and governance academics, a heritage restorer, an ESG accountant and an impact evaluator, rather than GLAM practitioners.

Jelgava City Library confirmed that its 10 participants also answered the survey; for the other groups this cannot be verified. Survey and focus group figures are therefore reported separately and never added together.

What the focus groups agreed on

Keep the architecture, make it easier to use

Number of focus groups, out of ten, that raised each point. No group asked to add a domain or to remove a competency as irrelevant.

  • Judged the architecture comprehensive and relevant4 domains, 31 competencies · all ten groups

    10 of 10

  • Tied adoption to funders, planning cycles, existing standards or endorsementR11 · HU-1, HU-2, FM-2, JC-1, JC-2, MZ-1, MZ-2, UN-1

    8 of 10

  • Asked for levels that follow function or proficiency, not job title or seniorityR1 · HU-1, HU-2, JC-1, JC-2, MZ-1, UN-1, UN-2

    7 of 10

  • Asked for real-life examples, templates and toolsR5 · HU-1, HU-2, JC-1, JC-2, MZ-2, UN-1, UN-2

    7 of 10

  • Asked for plainer terminology and a fuller glossaryR6 · HU-1, FM-1, JC-1, JC-2, MZ-1, UN-1, UN-2

    7 of 10

  • Saw overlaps between competencies, above all in ESG dataR8 · HU-1, HU-2, FM-1, JC-1, JC-2, UN-1, one MZ-1 participant

    7 of 10

  • Found the framework heavy and asked for an entry layerR3 · HU-1, HU-2, FM-1, FM-2, JC-1, UN-1

    6 of 10

  • Found the level descriptors too brief or too alikeR2 · JC-1, MZ-1, MZ-2, UN-1, UN-2

    5 of 10

  • Asked for broader crisis preparedness: civil protection, conflict, natural hazardsR9 · JC-1, JC-2, HU-1, HU-2

    4 of 10

  • Questioned or widened the scope of CC-06, digital sustainability and AIR10 · FM-2, HU-1, JC-1, UN-1

    4 of 10

Counting groups shows the breadth of support, not its prevalence in the sector. Practitioner-based evidence comes from at most six groups (HU-1, HU-2, FM-1, FM-2, JC-1, JC-2). Source: ESG4GLAM focus group reports, 10 sessions held in September 2026; analysis in D2.5, Section 4.

Levels L1–L5

The most contested element, and the priority for revision

82.4%of survey respondents say the five levels match real GLAM roles
Not supportedin its current form, conclude the focus groups

The survey shows broad acceptance of the idea of progression; the focus groups show what should change. Every one of the ten groups raised a point on the levels. Six accepted progression in principle, but seven asked to decouple levels from job titles and seniority and five found the descriptors too thin.

  • A director may know little about sustainability while an entry-level employee knows a great deal.
  • In a small institution one person may work “perhaps even from L1 to L4”.
  • The EQF mapping itself was not assessed by any group: it is neither validated nor challenged.

The level recommendations (R1, R2) remain provisional until comments are checked against the full version of the framework, which is the first next step.

Today the levels read as a ladder of job titles: a department manager is “L3” in every competency. Proposed: a level per competency, defined by responsibility, autonomy and observable evidence. Job titles and EQF ranges become illustrative. Illustrative example of one professional.

Assessment by dimension

What the evidence supports, and what not yet

How the focus groups judged each dimension of the framework. This describes what participants said; it is not a formal approval or a claim of consensus.

  • Relevance to the GLAM sector

    Supported at overview level

    All ten groups judged the architecture relevant; none asked to remove a competency.

  • Completeness of scope

    Supported, with targeted extensions

    No domain missing. Extensions asked for: broader preparedness, digital scope, heritage-aware environmental criteria.

  • Clarity

    Supported, with refinement

    Clear to ESG and policy experts, harder for small-institution practitioners; English-only text is a barrier in Finland and Serbia.

  • Level structure L1–L5

    Not supported in current form

    Every group raised a point; seven want per-competency levels decoupled from job titles. R1 and R2 provisional.

  • EQF mapping

    Mapping not assessed

    Only its presentation was commented on: show EQF as learning outcomes, not rank.

  • Internal coherence

    Needs refinement

    Overlaps seen by seven groups, mainly the ESG data cluster; no agreed merge yet.

  • Proportionality

    Needs development

    Endorsed in spirit, but six groups warned of volume: an entry version and pathways are needed.

  • Practical applicability

    Conditional

    Clear uses named in every group; uptake depends on examples, tools, funders, planning and endorsement.

Source: D2.5, Table 14. Focus group evidence is based on ten partner-written reports; participants reviewed the framework, they did not apply it.

From feedback to action

84 change requests, eight high-priority recommendations

Change requests logged, by partner 84 in total
34Humak 21Univ. of Novi Sad 18Mazzini Lab 6Jelgava City Library 5Forum Marinum
  • Jelgava City Library 6
  • Forum Marinum 5

Counts reflect how detailed each report was, not how critical each group was. They include four suggestions by a Humak moderator, which carry less weight than participant evidence. All 84 will be logged in one register with a decision (adopted, deferred or declined) and a reason.

High-priority recommendations from the focus groups

R1Define levels by responsibility, not job title Provisional

Define levels by responsibility, autonomy and demonstrated evidence; treat job titles and EQF ranges as illustrative; let self-assessment return a profile with a level per competency; allow non-sequential access. 7 groups, 4 partners

R2Write distinct level descriptors Provisional

Distinct descriptors per competency, each with one observable output or decision; separate individual competence from institutional maturity; keep L1 for support staff and volunteers. 5 groups, 3 partners

R3Add an entry layer

A one- to two-page entry summary, a modular-use statement and a starter pathway across all four domains, with an advanced tier. 6 groups, 4 partners

R4Pathways by sub-sector and size

Turn the proportionality principle into pathways for museums, libraries, archives and galleries of different sizes, including volunteer-run bodies and local public administrations. 6 groups, 5 partners

R5Examples, templates and tools

Examples in every domain, including one-person institutions and failed approaches, an end-to-end worked example, and a workbook, templates and self-assessment in the toolkit. 7 groups, 4 partners

R6Terminology, glossary and naming

A fuller glossary (CSRD, ESRS, EU Taxonomy, materiality, GRI…), one-line definitions with a GLAM example, one naming pattern, and titles aligned with descriptions. 7 groups, 5 partners

R10Decide the scope of CC-06

Keep digital sustainability and responsible AI as one wider cross-cutting competency, or split digital preservation from AI; add born-digital preservation, AI copyright and digital wellbeing. 4 groups, 4 partners

R11Position the framework among existing systems

Show how it relates to ICOM, IFLA and ICA codes, the SDGs and schemes such as Spectrum and EcoCompass, and how it plugs into planning and reporting without adding a report. 8 groups, 5 partners

Validation indicators

Targets from the project application

  • Achieved

    Stakeholders giving feedback

    Target ≥ 100

    102

    survey respondents from 70+ organisations in 12 countries, plus 34 focus group participants

  • Achieved

    Approval rate

    Target ≥ 90%

    91.2%

    endorse the framework as a valid reference tool; 93.2% among external stakeholders

  • Partially achieved

    Utility and applicability

    Target ≥ 90% positive

    94.1%

    as a training basis, but 74.5% for practical application in respondents’ own institutions

  • Achieved

    Diversity of settings

    Diverse cultural settings

    All 4

    GLAM types took part, plus universities, cultural associations, art services and consultancies

  • In progress

    Feedback integrated

    Target ≥ 80% of key suggestions

    18

    recommended adjustments, tracked in a Version Comparison Matrix released with the final framework

  • Achieved

    Dissemination

    Target ≥ 5 channels

    6

    channels used or planned: websites, social media, newsletters, press, professional networks, Multiplier Events

What happens next

From validation to training

  1. 01

    Decisions on the change requests

    The Steering Committee reviews the feedback-to-action matrix. Every request gets a decision and a reason.

  2. 02

    Version check, then a new level model

    Comments are checked against the full framework; levels are then redesigned as per-competency profiles with richer descriptors.

  3. 03

    Final framework

    Plain-language entry layer, glossary, broader example roles and boundary statements, released with a Version Comparison Matrix.

  4. 04

    35-hour training course

    Seven modules of about five hours: a Cross-Cutting foundation first, then Social, Environmental and Governance, with elective pathways.

  5. 05

    ESG Knowledge Box

    A self-assessment checklist, a glossary, an annotated toolbox linked to each competency, and case studies from GLAM of every size.

  6. 06

    Pilot and second validation

    The course and revised framework are tested in institutions of different sizes and types, with more archives, libraries and small organisations.

The full report

All the evidence, in one document

Methods, full results by group, the thematic analysis of comments, all focus group recommendations and the feedback-to-action matrix behind the final ESG4GLAM Competency Framework.

Download D2.5 (PDF) See the competency framework

ESG4GLAM is a Cooperation Partnership in Vocational Education and Training funded by the Erasmus+ Programme (KA220-VET, 2025-1-IT01-KA220-VET-000364271).