Developing a Responsible Business Compass for Ireland

4. Recommendations

Abstract

Chapter 4 presents recommendations for the future Compass to be a user-centric resource and a trusted reference. Recommendations cover the full data management cycle, from data collection, storage, curation, dissemination and protection, and are presented across three main areas: the frontend, the backend, and aspects related to co-ordination, co-creation and maintenance. These recommendations also include considerations, as resources and capacity allow, for the integration of artificial intelligence (AI) in navigation, data governance, and user interactions including with non-human, AI agents. They come with concrete guidelines or examples, drawn from the pilot tool and the analytical report.

In Brief

These OECD recommendations aim to support Ireland in developing a Responsible Business Compass to navigate the EU legislation on sustainability, climate neutrality and responsible business conduct. The future Compass aims to help Ireland’s policymakers improve policy coherence and reduce administrative burden, as well as help businesses understand their obligations and find available support. These recommendations draw on a mapping of EU legislation, Ireland’s business policy system analysis, policy and multi‑stakeholder dialogues, a public consultation, and a pilot tool for experimentation that was developed for user experience (user ex). The recommendations are intended for both policymakers and developers.

Overarching principles

• User‑centric design: tailor the Compass to different user needs and profiles, and enable new forms of information discovery boosted through artificial intelligence (AI).

• Trusted reference: ensure access to high‑quality, accurate, timely, clear and secure data, aligned with global data standards.

Frontend recommendations (user interface)

• Dual access: tailored navigation for policymakers and businesses.

• Layered navigation: high‑level overviews and options for deeper dives; consider linking to external technical resources.

• User-friendly UX: simple navigation, interactive elements, maximum three clicks to reach content, undo options, clear system messages; prepare for zero user interfaces (zeroUI) and AI agent navigation.

• Coherence of information and semantics: harmonised terminology and co-ordinated updates across Departments.

• Business profiling: clear definitions of business categories and alignment with other platforms; consider unique IDs or cross‑platform links.

• Visual identity: consistent style guide and logo.

• Performance and security: fast loading, secure navigation, no personal data collection.

• Accessibility: plain English and Gaelic, device‑friendly, compatible with assistive technologies.

• Guidance tools: combining online/offline guidance, through tooltips, demo video, email support; feasibility of an AI chatbot to be assessed.

• Legal compliance: proper use of copyrights and disclaimers.

Backend recommendations (data governance and system architecture)

• Data governance model: choose between centralised one‑stop shop | decentralised hub‑and‑spoke | hybrid approach (recommended for flexibility).

• AI readiness:

• Relational SQL data model: scalable, transparent, supports complex queries and data integrity.

• Rule‑based schema: consistent linking of legislation, institutions and policy initiatives.

• Data preprocessing: structured classification of qualitative data, supported by generative AI and manual verification.

• Technology choices: choose between open‑source or commercial data visualisation tools (e.g. Rshiny or PowerBI) easy to use but limited for customisation or navigation | Custom interfaces (JavaScript/Python) more flexible but costlier and requiring more expertise | Hybrid solutions that could balance needs.

• Security: ensure confidentiality, integrity, and risk management, especially if data exchange or profiling grows more complex.

• User analytics: integrate monitoring of usage patterns, errors, and user feedback.

Co-ordination, co‑creation and maintenance

• The Compass requires ongoing co-creation with the 13 Departments/agencies involved in Ireland’s sustainability and RBC agenda and business support, as well as the business community.

• Agree on respective roles and responsibilities; ensure clarity on how input could be integrated into the data cycle.

• Ensure up‑to‑date mappings of legislation and policy initiatives, and periodic updates of the Compass in sync with other relevant platforms and systems.

• Ensure cross-Department alignment with shared commons, definitions, taxonomies, and reporting rules. Consider automation and interoperability with existing repositories.

• Adopt a realistic scope reflecting institutional capacity to maintain high quality data, timeliness and trust.

• Appoint contact points in each institution.

• Enable revisions and upgrading in the frontend and backend; balance risks and costs of vendor lock-ins, internal technical expertise or external maintenance.

• Keep comprehensive, updated technical documentation.

• Involve communication teams, IT departments and administrative services in cross-functional co-ordination, throughout design, testing and deployment.

• Involve civil society, businesses and academia in cross-sectoral co-operation for user‑centric design, visibility and credibility.

Introduction

These recommendations aim to support Ireland in developing a Responsible Business Compass to navigate EU legislation on climate neutrality, sustainability and responsible business conduct, as well as ease compliance and promote competitiveness and sustainability in European and global value chains (GVCs).1

The future tool aims to make it easier for Ireland’s policymakers to identify policy duplication and gaps, reduce administrative burden and offer the right support to business. It also aims to help businesses operating in Ireland understand the scope and obligations of relevant legislation and what policy support is available to them. The recommendations were informed by mappings of EU legislation and Ireland’s business policy system, complemented with policy and multistakeholder dialogues, and a public consultation on the desirable features of the Responsible Business Compass. A pilot tool that enabled a dynamic exploration of the mappings was developed for hands-on experimentation.2 The OECD recommendations build on the OECD Smart Data Strategy (Box 4.1)(OECD, 2026[1]) and the OECD Data Lake on SMEs and Entrepreneurship (OECD, 2026[2]), that provide the principles of a data governance framework for the future Compass.

Box 4.1. OECD Smart Data Strategy

The OECD Smart Data Strategy aims to enhance the use of data to improve policymaking and societal well-being. The Strategy defines two broad goals: Integrating the Data Cycle and Embracing Smart Data. Initiated in 2019, originally to modernise OECD’s data and statistics, it provides guidance on how data are disseminated and managed, and encourages mutualisation, automation, and the use of new data techniques, including artificial intelligence (AI), along the data value chain. The stages of the data cycle include design, collect, process, analyse, disseminate, evaluate.

The OECD Smart Data Strategy proposes a renewed data quality framework. Data quality is assessed in terms of relevance, accuracy, timeliness, accessibility, clarity, coherence-consistency, integrity-transparency, cost efficiency and security-privacy. The Strategy also promotes horizontal work and collaboration for more efficient ways of data sourcing, and by expanding the OECD data governance “toward a governance of contents”.

Source: (OECD, 2026[1])

The OECD recommendations on the future Compass are published together with the interactive pilot tool and an analytical report that presents the key findings of the mappings, methodology and technical information (Chapters 1-2-3 of this publication). The recommendations cover the full data management cycle and are presented across four main areas: overarching principles, the frontend, the backend, and aspects related to co-creation, co-ordination and maintenance. The recommendations come with concrete guidelines or examples, drawn from the pilot tool (hereafter “the tool”) and the analytical report (hereafter “the report”).

4.1. Overarching principles

Know your user(s)

A user-centric approach. The design of the Compass should follow a user-centric approach that prioritises users’ (variety of) needs, abilities, and contexts of use. Developments should also prepare for new forms of AI-boosted data discovery and navigation, including through natural (human) language queries (prompts) and processing. This user-centric approach is typically implemented through design, functionalities and continuous improvement (i.e. maintenance, use analytics and user feedback).

Deliver high quality output

A Trusted reference. The Compass must be recognised as a trusted source of evidence for policy and business intelligence. The tool must comply with global data quality standards (of relevance, accuracy, timeliness, accessibility, clarity, coherence-consistency, integrity-transparency, cost efficiency and security-privacy). These standards underpin the data governance framework for the Compass.

4.2. Frontend

The front end is the visible and interactive layer of a website or application. It includes everything users see and use in their browser or device. OECD recommendations for the user-facing side of the future Compass focus on presentation and interaction rather than underlying processing. However, they bring considerations regarding the architecture of the Compass insofar as those affect navigation and user experience. OECD recommendations include the following.

Dual access, plurality of users

The Compass is aimed at two broad groups of users – policymakers and businesses – with differing expectations, questions and understanding of the EU legislation. The Compass should reflect this duality and provide tailored access to each group, with different views and navigation options, even to the same content (the tool).

Even within the same group, users may have different levels of expertise and understanding of the EU legislation or the business policy system. They may also look for different types of information. This should be addressed in the architecture of the Compass, which should enable surface navigation (across high-level information) and deeper dives. Some OECD recommendations for deeper dives are provided in Annex 4.A. For those deeper dives, stakeholders also suggested that the Compass could provide access to external – more specific or technical – resources, e.g. webpages or co-ordinated tools proposed by different Irish Departments and agencies, case studies or support pages, official pages of the EU legislation or the relevant European Commission pages where guidance is provided etc. (see further elaboration about data governance and co-ordination).

User-friendly navigation

The Compass should propose simple and intuitive navigation across the mappings, high-level and technical information, and different profiles of users.

• Multi-dimensional navigation requires interactive components, such as tabs, buttons, tooltips and mouse-overs, as well as a simplified navigation bar.

• The number of clicks/options to access relevant information should be limited to three to account for the limited time businesses have for surfing content.

• Likewise, users should be offered options to undo, unselect or correct actions to avoid impasses.

• The Compass should provide clear system status messages, such as loading indicators, confirmations and error messages etc., to reinforce transparency and avoid user frustration. To the same extent, the Compass’s developments should anticipate and prevent any possible disruption in navigation, such as broken links.

• Stakeholders who provided feedback on the tool have stressed the following components as particularly efficient: the word cloud (“Institutional and policy landscape for RBC” tab), the Sankay diagrams (“Policy support for compliance” tab), the “Overview of EU legislation” tab, the profiling of businesses (“Which legislation can affect my business?” tab), and the breakdowns along a simplified taxonomy of business obligations (multiple tabs). These visualisation/navigation options should be prioritised in the Compass’s developments.

• Stakeholders have also suggested additional navigation with keywords. The Compass could include a taxonomy of key topics (keywords) and provide access to targeted information based on this list. The list should be updated as needed.

• Alternatively, the Compass could include a search engine using natural language processing (NLP) for accessing relevant information (see further elaboration about the backend, and data governance and co-ordination). The feasibility of integrating NLP into the Compass will have to be assessed as part of future developments.

The Compass should also prepare for AI-enhanced navigation, e.g. through zero user interface (zeroUI) and AI agents.

• ZeroUI refers to systems without traditional graphical interface (such as screens, buttons or menus). Interactions rely on natural, implicit, or ambient methods, e.g. voice interfaces, such as conversational assistants, or seamless, background operation, such as notification systems when deadlines are approaching, new support schemes become available, or regulations or requirements are changing. A zeroUI could for instance generate step-by-step compliance guidance.

• The Compass should also make information machine consumable, either by AI agents or large language models (LLMs), which in turn could enable more intuitive and implicit navigation for human users. AI agents can for instance embed Compass outputs into business processes, suggest documentation, recommend relevant funding or advisory services, pre-fill forms or guide users during reporting etc. AI agents can even link the Compass with internal company data, regarding own operations or supply chains, or with external data sources, regarding risk profiles or public sanction lists. In AI-boosted navigation, a robust data governance is even more critical, as generative AI does not replace the need to modernise data processes, it reinforces it (SIS-CC, 2025[3]) (see further elaboration about data governance).

Coherence of information and semantics

The Compass must be developed on harmonised content and agreed terminology and wording, based upon a consensus on the text to be used across Departments and agencies, and policy areas. These harmonisation efforts will contribute to building consistent semantics and concepts for AI integration.

The consensus should include a general agreement on the sequencing of updates and alignment with other relevant resources, so that users can access (the same) timely and coherent information through different channels (see further elaboration about the backend, and data governance and co-ordination). For transparency, the Compass may display the date of latest update.

Profiling of businesses

It is important to allow for navigation according to business profile, and the Compass should ensure clarity and accuracy on this.

• Defining categories of businesses. Legislation applies to different types and sizes of businesses, for example, large enterprises, small and medium-sized enterprises (SMEs) or small and mid-cap enterprises, or entities that import, export from or place goods on the EU market (see Annex 1.A in Chapter 1 for detailed overview). Businesses, especially smaller ones, may need particular guidance to find out which category they are in, and therefore which legislation applies. The Compass should offer tailored profiling options along these lines. Likewise, as compared to the tool, it could provide more options and clarity on how businesses integrate into GVCs, e.g. the scope of their import activities.

• Aligning profiling across platforms. Stakeholders that provided feedback on the tool have suggested linking the Compass to existing profiles already developed for, or in place of, other platforms and one-stop shops. The feasibility of aligning business profiling across platforms could be assessed depending on the architecture and technical environment of those. Different approaches could then be explored, such as a unique ID identifier, a central website serving as a hub to connect across the other platforms, cross-link between platforms or a cross-platform structured approach, e.g. through schema markup. The profiling in that case should include considerations on security-privacy (see further elaboration about the backend, and data governance and co-ordination).

Visual identity and consistency

For ensuring consistency across the Compass and its components, and raising its public profile, a style guide is needed. This charter would list detailed layout, including a palette of (primary and supporting) colors, typography and hierarchy (font, sizes, headings etc.), icon and imagery style (shapes, patterns, photos etc.), a grid system for spacing and alignment, consistent interaction patterns, and related rules, to be applied across the Compass and in external references to the Compass (e.g. other web pages, documentation, flyers etc.). To build the visual identity of the Compass, a logo could also be helpful.

Performance and security

The Compass should be optimised to reduce time and barriers to navigation, e.g. through fast load times, efficient rendering and minimum interactions (e.g. compressed images, simplified architecture etc.). It should also provide secure conditions for navigation, especially if connected to other platforms. The Compass should not aim to collect personal data and should protect session information (see further elaboration about the backend).

Accessibility

Content should be readable and accessible to all users. This includes:

• Delivering messaging and documentation in plain English and Gaelic, avoiding cultural or exclusionary expressions, as well as jargon and complex phrasing, and privileging short sentences.

• Enabling keyboard, mobile and tablet devices versions.

• Optimising navigation with assistive technologies and web content accessibility standards, including for users with differentiated visual needs (colour-only navigation, scalable text and zoom, contrast, light/dark modes etc.), or diverse interaction modes (e.g. gesture-only navigation, but instead mouse, keyboard and touch, including for larger touch targets).

• Testing the Compass across different technical environments (e.g. mobile/desktop devices, Android/IOS, screen sizes, etc.) and different profiles of users (including small businesses).

Online/offline guidance

Users should be able to access guidance at any time during their navigation, e.g. through a short demo video on how to use the Compass, tooltips on mouse-over providing definitions and explanations, email contact etc. Stakeholders who provided feedback on the tool have proposed a guide or manual, or deploying a hotline or chatbot for assistance. The feasibility of developing a 24/7 chatbot using generative AI will have to be assessed.

Copyrights

The public facing tool should present all relevant legal disclaimers and acknowledgements, as well as comply with any intellectual property (IP) requirements. A list of IP requirements, disclaimers and acknowledgements will have to be defined and prepared in advance, to inform the Compass’s developments, e.g. in terms of length, placement, sensitivity etc.

4.3. Backend

The backend refers to the part of a website or application that operates behind the scenes, handling data, processing and communication with databases, servers and other platforms. It is not visible to users but supports the functionalities of the frontend. OECD recommendations on the server-side layer of the future Compass focus on data management, business logic and governance (i.e. rules, processes and workflows), interfaces and access control, and back-office communication. OECD recommendations include the following.

(De)centralised data governance

It should be decided, upstream in the development process, what level of information depth the Compass would give access to. This questions the core role of the Compass, as a comprehensive one-stop shop or as a hub-and-spoke. Challenges for development and future maintenance will differ between these two options and possible variants. The feasibility of using generative AI for regular data collection and classification will have to be assessed, as well as the modalities for enhancing data exchange protocols and interoperability between administrations.

• As a comprehensive one-stop shop (centralised governance), all information will be collected, classified, stored and maintained by the central unit in charge of the Compass, that will also be responsible for overseeing the full data cycle and backend/frontend developments. Relevant Departments and agencies for supporting business compliance will be involved through co-ordination and reporting mechanisms.

• As a hub-and-spoke node (decentralised governance), the Compass will orientate users across Ireland’s ecosystem of business resources. Each relevant Department and agency will be responsible for collecting, classifying, storing and maintaining information, and overseeing the full data cycle and quality. The central unit in charge of the Compass will be involved in co-ordination and backend/frontend developments to access information from external systems.

• A variant (hybrid governance) would combine the two data governance approaches above, by maintaining a one-stop shop for higher-level information and non-expert navigation, and adopting a hub-and-spoke model for more technical and granular information, especially on EU legislation and details about Ireland’s policy support and targeted contact points. This hybrid governance could allow the co-existence of different platforms and systems, and more user-centric policy delivery (e.g. at local level with proximity to users, or in specific policy areas with greater business acumen). It could also enable more timely and accurate input from the policymakers shaping business policies and practitioners directly supporting businesses.

  Relevant Departments and agencies would be responsible for managing their data cycle in relatively similar ways as it is currently, but integrating shared guidelines and rules. The central unit in charge of the Compass would develop data exchange protocols and commons, provide vertical and horizontal co-ordination mechanisms, especially to ensure alignment in data sharing across the board, and oversee part of (high-level) data management as well as the backend/frontend developments of the Compass.

  The hybrid model requires consistent semantics (concepts, codifications, identifiers) across platforms and systems. Data access is then enabled through application programming interfaces (APIs) that serve as doors to other Departments and agencies’ systems, to query their data without need to download, store or reconfigure. An API-centric approach implies the prior adoption of open data standards, i.e. principles for a consistent, logical structure to all data files. SDMX standards (standards for Statistical Data and Metadata eXchange) provide an example of reference framework that was developed to enable statistical agencies to implement the most appropriate governance architecture and improve the quality and maintenance of their structural metadata and business process (SDMX Statistical Working Group, 2023[4]).

AI readiness

For the Compass to be AI-ready, it should be treated as a policy intelligence system, not only as a website. Information (data) must be harmonised, structured, multilingual, and enriched with high-quality metadata (i.e. contextual information) to be usable by AI agents and LLMs, or contribute to train or fine-tune them. Short, traceable statements work better, supplemented with short explanatory text. Linkages between EU legislation, institutions and policy support measures need to be encoded, along a policy taxonomy. Every rule should also be documented (validity dates etc.). The Compass could then be linked to authoritative external data sources through governed connectors. Finally, an AI-powered Compass should be able to distinguish between low-risk AI tasks, such as search, synthesis, and higher-risk tasks, such as legal interpretation or compliance scoring, and it must be trained and evaluated. The AI-readiness agenda will be shaped by both strategic alignment and practical experimentation (SIS-CC, 2025[3]). Ireland’s policy community will therefore have to invest not just in the future tool, but in standards, skills, and shared strategies.

Relational data model and rule-based schema

The tool has been developed on a Structured Query Language (SQL) data model, with a predefined schema articulating tables, columns and relationships using keys. Transactional rules enable users to retrieve, update and join (link) data. The Compass should scale up this data model and build on the proof-of-concept and lessons learned during experiments.

• SQL data model. The SQL model offers several advantages compared to noSQL approaches, such as transparency in how data is stored and organised and how relationships and rules are built, and easiness to understand and manage complex datasets. The SQL model is a well-documented standard, that does not require advanced data science skills for deployment or maintenance. The relational model is also appropriate for handling connections between data (i.e. interactions across and between EU legislation and Ireland’s policy mixes), for enabling complex queries (e.g. though joint tables) and for maintaining integrity in future updates (in institution, policy or legislation mappings).

• Rule-based schema. A rule-based schema defines interactions between selected pieces of EU legislation and Ireland’s support measures, their nature and where they take place (see detailed description in Chapter 3 of this report). This schema provides the foundations of the relational data model. The classification of information is consistent across EU legislation, and across Ireland’s institutions and policy initiatives. Under this schema, Ireland’s institutions are linked to EU legislation through their objectives, mandate, and strategic orientations. Policy initiatives are linked to institutions through existing governance arrangements. Policy initiatives are also linked to EU legislation through the target populations and scope of business obligations. The schema was elaborated to align with the data governance framework of the OECD Data Lake on SMEs and Entrepreneurship (OECD, 2026[2]), and its mappings of business policy mixes and taxonomy of policy target population and beneficiaries, to possibly support international data sharing and peer learning. It was also elaborated to reflect expert knowledge on the EU legislation on sustainability and RBC, and findings from the legislative mapping, especially regarding entities in scope and their obligations (see Chapter 1 of this report), as well as the OECD RBC standards (see Chapter 2 of this report).

Data collection and pre-processing

The SQL model and rule-based schema require a pre-processing of data. Raw information was collected from the official websites of relevant institutions (see the list of institutions in Chapter 3 of this report), Ireland’s National Enterprise Hub (Ireland, 2026[5]), and the OECD Data Lake on SMEs and Entrepreneurship (OECD, 2026[2]). The mapping of policy initiatives was refined based on feedback received from Ireland’s policymakers during the policy workshop and successive consultations on the tool and the report. Raw data, qualitative by essence, was classified in Excel along the SQL relational data model above. Data collection, classification and quality check were conducted using a generative AI with access to the Internet (i.e. upgraded versions from ChatGPT 4.0 deep research to ChatGPT 5.2). Information was controlled manually, as well as relationships in the data model.

Open-sourced data visualisation solutions and proprietary developments

The Compass’s developments should be guided by a strategic, integrated, user-centric vision of ambitions, data governance, and cost and technical feasibility. The Compass could leverage open-sourced, public data visualisation (data viz) products, for fast and ready-to-use solutions. Alternatively, future developments could aim to design and deploy proprietary solutions for more tailored navigation and visualisation. A feasibility study should inform decisions, accounting for the cost-benefit of proprietary solutions, and the respective ease of use, flexibility, scalability, technical requirements (including security and monitoring) and possible integration with existing platforms and systems, of different solutions.

• From public data viz solutions. The tool has been developed using Microsoft Power Business Intelligence (PBI), to leverage publicly accessible PBI data viz solutions. But other data viz solutions, such as Tableau, are also available on the market. PBI is typically used for corporate reporting, self-service analytics and interactive dashboards. Easy to use, it however requires a strong data modelling (Power Query and DAX codes) and may be more limited for customisation, complex storytelling or online publication. Limitations stem from the native, pre-defined, built-in visualisations provided by Microsoft PBI, which restricts flexibility in customising, responsiveness for complex visuals, and integrating external code. For instance, the tool includes data viz that were developed offline, with Python/R programmes applied to native visual options provided by Microsoft PBI, and underlying data and SQL model above, for exploring linkages between business obligations and Irish policy support measures (“What government support for what obligation?” tab). Yet, the online PBI version does not support Python/R languages, preventing the upload of the dynamic data viz developed using these languages. Publishing PBI reports to the web also brings data in the public space, being unsuitable for sensitive or internal data.

• To programming-based custom interfaces. Custom interfaces, using JavaScript or Python, provide a higher degree of control and flexibility over how data and functionalities are presented. Interfaces can be tailored to exact requirements with no restriction on predefined components, and they can support complex layouts, animations or dynamic filtering. They can also more easily connect to Application Programming Interfaces (APIs), databases, backend services and existing platforms, and support end-to-end workflows (not just visualisation). However, this comes with trade-offs in terms of complexity, cost, and maintenance. Initial development investments are larger, the time for deployment is longer and the level of expertise required is higher. Custom solutions may also exceed actual needs.

• Hybrid data viz options. Hybrid solutions exist that can combine the strengths of off-the-shelf tools (e.g. Power BI) with programming-based custom interfaces, and help balance requirements. The tool was for instance embedded into a custom-built frontend OECD webpage. Integrating custom-built components, and other programming languages (such as Python or JavaScript), into BI platforms can extend their capabilities and enable more advanced data preprocessing, tailored visual development (e.g. double-sided Sankay diagram) as well as more intuitive navigation.

Security and data protection

The Compass should offer appropriate security and data protection to users, which engages the liability of the central unit in charge of its deployment and maintenance. The tool currently relies on by-design Microsoft PBI and OECD web security features. It also handles public, freely accessible data only, and is labelled as pilot and experimental. Depending on decisions taken for the future Compass, the conditions for data confidentiality, integrity and protection may have to be reinforced. This may be true in case of non-public data processing, data exchanges across platforms and systems increasing the attack surface and risks of breaches, or if the Compass proposes the use of unique ID identifier and profiling etc. A list of possible risks and technical solutions could be developed to inform strategic decisions before developments start.

User analytics and retro-feedback

Ideally, the Compass could include features -embedded in the future tool by design- to monitor navigation patterns and areas of interest by user profiles. It could also track system errors or underperformance. Both approaches could help improve the Compass’s design and responsiveness to user needs. Real usage feedback could also inform the tailoring of a user-centric tool, and interface and system improvements. For instance, in case an assistance chatbot is deployed, conversation analysis could provide insights on user groups, points of confusion, or training data and self-improvement. Alternatively, depending on the complexity of the future tool and technical solutions underpinning it, user opinion surveys could also help gather viewpoints and suggestions on a regular basis. These surveys could be conducted online from the Compass, and/or through the relevant Departments and agencies and their communication channels.

4.4. Co-ordination for co-creation and maintenance

The Compass is a user-driven co-creation, that need to reflect the views of the business community and the various stakeholders involved in policy support. The institutional mapping has identified 13 Departments and agencies that are at the forefront of the sustainable, circular and RBC transition in Ireland, with responsibilities in supporting businesses at national or subnational levels. These 13 institutions overlap in their mandates that cut across several policy domains, from enterprise, trade and investment promotion, employment, climate, energy, regional development and to consumer policy.

To be and remain a trusted reference, the Compass must also deliver time-proof quality output (in terms of relevance, accuracy, timeliness, accessibility, clarity, coherence-consistency, integrity-transparency, cost efficiency and security-privacy). These quality standards, and how they are met, should be agreed among policy stakeholders.

Co-ordination will be critical for developments and maintenance. The stakeholder dialogue that was engaged in 2025 with Ireland’s policymakers, the business community and academia, was instrumental in building consensus, advancing the report and the tool, and informing these recommendations. OECD recommendations for future co-ordination include the following.

Maintaining updated legislation and policy mappings

The legislative and policy mappings underpinning the Compass should be maintained and up-to-date, including through co-ordination across different Departments and agencies, and across different levels of government.

• Any relevant change in Ireland’s business support system should be reflected in the Compass in a timely and accurate manner, e.g. the release -or phasing-out- of policy initiatives, Ministerial overhaul, the creation of new governance institutions, or a shift in policy mandate and portfolio across Departments and agencies etc. In addition, stakeholders who provided feedback on the tool suggested extending the mapping of available support measures to those proposed at local level by local authorities or chambers of commerce.

• Future relevant changes in EU legislation should also be reflected in the Compass. The Compass could be designed to accommodate new pieces of EU legislation as relevant, including possible revisions in the policy mapping, rule-based schema and frontend/backend.

The Compass should be synchronised with related platforms and systems to avoid confusion. Information should also be kept harmonised across the different Departments and levels of government involved. Approaches may differ depending on the data governance model chosen.

• Shared vision of roles and responsibilities. Departments and agencies, including the central unit in charge of the Compass’s development and maintenance, need to agree on respective roles and responsibilities, and be informed of their place in -and impact on- the data management cycle of the Compass. Part of the Compass’s maintenance will require input on the changing policy mix, being automated (top-down) or manually provided (bottom-up).

• Shared commons. Policy stakeholders should be involved in the definition of data and governance commons, or at minimum be consulted on those before they are adopted. Those include taxonomies, standard definition of policy terms, rules for reporting policy changes, rules for managing the platform and the data cycle (including updating), or the level of granularity the Compass will reach. The level of information depth the Compass will aim to is key at this stage, to set how responsibilities are shared and how new input is integrated. Updates (e.g. to the mappings) should be submitted in standardised formats to ensure consistency across-the-board (e.g. via online or excel forms with a common structure, formats etc.).

• Automation. The Compass could also access existing repositories automatically, such as the National Enterprise Hub. This would require implementing – or strengthening – interoperability rules between institutions, e.g. though a mapping of data fields between systems, shared taxonomies or reference data, transformation rules where data sources and systems differ, open standards etc. Alternately, the Compass could access information from related platforms and systems and re-process information to create this consistency, e.g. using generative AI. Commons and rules, as well as a general understanding among policy stakeholders, will be equally necessary.

• More is less. The ambitions of the Compass, in terms of comprehensiveness and depth, should be sized to the capacity of Ireland’s institutions to inform and maintain it. As not all Departments and agencies may be able to provide the same level of details, or have platforms and systems in place that could communicate with the future Compass, a consensus should be reached upon a common baseline of information, realistic system interoperability improvements, or the possibility for different data exchange protocols to co-exist. A state-of-play across the 13 relevant institutions would help support co-ordination and prepare the Compass’s developments.

• Contact points. The designation of contact points in each of the 13 institutions could support co-ordination and ensure closer follow-up.

Updating the frontend and backend

The technical environment of the Compass may require maintenance, upgrading or revisions (e.g. in the relational data model or the rule-based schema). Approaches may differ depending on the data governance and technological choices made. The use of public data viz tools may increase risks of vendor lock-ins. The design of custom interfaces will require proper technical skills internally or imply the procurement of maintenance services. Keeping a technical documentation that is up-to-date, covering the frontend and backend, will help promote continuity in maintenance, even if outsourced.

Cross-functional, cross-sectoral coordination

The Compass co-creation and maintenance should involve communication teams, IT departments, and administrative services at different stages of the development, from design to testing to deployment. Consultation and testing with the civil society, business community and academia will also contribute to reinforce the user-centric features of the future Compass, its visibility and reputation.

References

[5] Ireland (2026), National Enterprise Hub, https://www.neh.gov.ie/.

[2] OECD (2026), OECD Data Lake on SMEs and Entrepreneurship, https://www.oecd.org/en/data/dashboards/oecd-data-lake-on-smes-and-entrepreneurship.html.

[1] OECD (2026), OECD Smart Data Strategy, https://www.oecd.org/en/about/projects/oecd-smart-data-strategy.html.

[6] OECD (2018), Due Diligence Guidance for Responsible Business Conduct, https://doi.org/10.1787/15f5f4b3-en.

[4] SDMX Statistical Working Group (2023), A Reference Framework for SDMX Structural Metadata Governance. Version 1.0. 27 March 2023, https://sdmx.org/wp-content/uploads/SDMX-Structural-Metadata-Governance.docx.

[3] SIS-CC (2025), “Beyond the Buzz. What does AI readiness really mean for official statistics”, SIS-CC Newsletter, https://siscc.org/beyond-the-buzz-what-does-ai-readiness-really-mean-for-official-statistics/.

Annex 4.A. OECD recommendations for deeper dives on EU legislation including due diligence related measures

1. Depending on the technical solutions chosen for developing the future Compass (centralised, decentralised or hybrid governance, that will determine the how), users could be offered access to more technical and granular information about EU legislation, as follows:

• Timeframe for implementation. More detailed breakdown of phasing-in timeframes could be provided for each EU legislation, potentially even a view where businesses and/or policymakers could filter by year over the next 10 years to see which legislation requirements will be fully in force and applying to them and when.

• Modalities of legislation application. Complementary descriptions of EU legislation could include lists of key modalities of enforcement (penalties, investigations etc.) and support (guidance, risk indicators etc.).

• More diversified user profiles. Adding filters or access options relating to the supply chain or sustainability issue in scope, in addition to the business profiling, could be useful to different audiences.

• References and links. References to the key legislative articles could be added within information about Requirements, Provisions for SMEs and Scope of application. Exact articles are provided across Chapters 1 and 2 of the report. Links to final legal text, Commission databases and other relevant guidance could also be helpful for businesses.

2. In addition, a more granular navigation across due diligence requirements could be proposed to users, for example:

• Distinguishing between due diligence related measures and other pieces of EU legislation

• For due diligence related measures, mapping information across the OECD six-step due diligence framework (OECD, 2018[6]), and key topics such as the role of stakeholder engagement and sustainability initiatives (see annexes of Chapter 2) to show commonalities and divergences identified in the report.

• Tailored and simplified overviews of the due diligence obligations and due diligence steps for each business are (see Annexes and summary boxes in Chapter 2).

Notes

← 1.  This Project is funded by the European Union via the Technical Support Instrument (TSI), and implemented by OECD, in co-operation with the European Commission. The opinions expressed, arguments employed and conclusions drawn in this Report should not be considered as representative of the official views of the European Union. Neither the European Union nor the granting authority can be held responsible for any use that may be made of the information contained therein.

← 2.  Developing a Responsible Business Compass for Ireland webpage, accessible at https://www.oecd.org/en/data/tools/developing-a-responsible-business-compass-for-ireland.html.