Guidance

Data quality action plan: implementation guide

Published 11 November 2025

What a DQAP is

A data quality action plan (DQAP) is a systematic approach for enhancing and sustaining the quality of data in your organisation.

It involves:

• identifying critical data
• setting quality standards
• assessing current data quality
• prioritising areas for improvement

The plan includes specific actions to address issues, targets for improvement, and ongoing monitoring. The aim is to ensure that your data consistently meets users’ needs and supports your business processes and effective decision making.

Your organisation should create a DQAP for each data asset that you identify as containing data that is critical to meeting your objectives.

This guide covers 7 steps for implementing your DQAP from start to finish, and includes relevant information from the:

• Government Data Quality Framework
• Government Data Quality Framework: guidance
• Government Data Quality Hub training modules

We recommend that you use the DQAP template and other resources to complete your DQAP, while the terms and definitions explain more about data assets and data-related roles.

You may already have tools, processes or resources in your department that support some of these steps. Once you’ve assessed your data quality, you can prioritise issues, set goals, identify the root cause and communicate results (steps 4 to 6) in any order.

What matters is that you:

• have a record of the actions taken
• report your results
• define how you’ll continue to monitor data quality for your data asset over time

You should use this guidance alongside, and aligned with, any data policies and guidance in place within your own organisation.

Who this guidance is for

Central government departments should follow this guidance.

It’s also recommended for other public sector organisations including:

• arm’s length bodies
• government agencies
• local government

Devolved administrations

Data management is a devolved issue, so administrations in Scotland, Wales and Northern Ireland have their own approaches to maintaining data quality. However, there are many benefits to aligning data quality initiatives across the UK, and it’s important that administrations share good practice and learn from each other.

Who should be involved

Your organisation should have one or more data owners or process owners that are responsible for actively monitoring the quality of your data assets over time.

A data owner or process owner should be responsible for producing your DQAP. They will need help from:

• business and technical subject matter experts (SMEs)
• business and data analysts
• information asset owners (IAOs)

The RACI Matrix in the DQAP template shows which roles should contribute to each step of the plan.

These roles align with the data ownership model which defines the main data ownership roles in government:

If these roles do not match your organisation’s named roles, for each activity you should still allocate the most appropriate individual that has similar responsibilities for data. Across government, data management and data governance roles and responsibilities are carried out in various ways. In smaller organisations, one individual may perform several roles.

Prioritise DQAPs that support ESDA submissions

You can use this guide to support all your data quality improvement initiatives, but you must prioritise Essential Shared Data Assets (ESDAs).

ESDAs are the most important data assets that are shared across government. For example, where the data from one organisation helps another organisation deliver a public service or evaluate a policy.

If your organisation is in scope – including all central government departments – you must identify the ESDAs you share with other government organisations. Your lead data owner should be coordinating this activity.

For each ESDA you identify, you should have a DQAP in place to assess and report on its quality. You should then list each ESDA in the Data Marketplace. This service includes a catalogue of data listings and information on how to request access if you work for a government organisation.

Step 1: Identify critical data

Critical data means the data that your organisation considers essential for delivering its business objectives. For example, data related to:

• public service delivery
• legislative, legal or contractual requirements
• policy development and evaluation

Having this data defined and documented is an important part of good data management practice. To determine whether the data is critical, consider if business outcomes rely on the data, its role in decision making, and the impact of poor quality.

You need to:

• identify a critical data asset for your DQAP and define its primary purpose – this will focus your plan on addressing specific data quality issues that are most critical to your organisation’s success
  • start by identifying your ESDAs and determine the critical data in those before anything else
• identify and document critical data items in the Data Criticality Matrix in the DQAP template
• log any known issues with the data in the Data Issues Log in the DQAP template
• prioritise critical data for measuring – use the data quality issues framework to prioritise fields or tables that most affect operations and users
• identify relevant technical and business SMEs

Who in your organisation should be responsible for this

Data stewards or business SMEs with an understanding of the data they create or use.

Build a data catalogue

This is an effective way to document all the critical data in your organisation. You can then record the underlying critical data elements for each of your critical data assets.

You might do this as part of a wider data mapping activity to support a range of data management activities such as:

• data architecture impact assessments – for example, as a result of new legislation or regulations, or new or revised government policies
• security planning
• interoperability design
• master data management
• metadata, data quality and risk management

Evaluate data quality

Evaluate data quality based on its specific use, recognising that importance may vary.

The quality of data can also vary depending on the specific purpose for which it is used. Different purposes may rely on different parts of the data or have varying expectations for those parts.

Your organisation’s risk appetite should inform any variations in data quality. For example, you may have a risk framework that sets a tolerance for data quality.

Step 2: Identify data quality rules

You should have a comprehensive set of data quality rules, aligned with user needs and business goals, to assess the quality of a data asset. Focus your assessment on the main use of the data rather than any secondary uses. 

To determine what rules to apply, you need to understand both the data and how it’s applied in a business context. There may also be specific data standards that data quality rules must abide by or be validated against. 

Defining rules is often an iterative process, where the initial data quality assessment reveals data quality rules or thresholds that need further refining. It’s unlikely that you’ll correctly define all the rules on the first attempt.

For the data asset in your DQAP, you need to:

• identify its users and their data quality needs – this will ensure that the rules define data that is fit for its intended purpose
• define data quality rules for critical data items – the Rules section of the DQAP template covers what to measure, identified methods, agreed targets and more
  • make sure you keep the measurement criteria and details up to date
• define the data quality dimensions and measures for data items in scope
• define, as a percentage, the threshold for acceptance

Who in your organisation should be responsible for this

Data stewards and SMEs with business knowledge should define the rules for the data asset.

They may need help from:

• data custodians or technical SMEs to help translate the business requirements to technical application
• business analysts – for example, to define user requirements

Differentiate from business rules

Data quality rules focus on measuring and maintaining data quality. This is separate from the business rules used in data processing.

Set realistic expectations

Define what ‘fit for purpose’ looks like, recognising that rules describe typical quality, not absolute perfection. For instance, a rule might set a 98% accuracy target, accepting occasional exceptions.

Use data quality dimensions

Use measurable data quality dimensions as a guide for setting rules, prioritising them based on user needs.

To evaluate data quality, start with the 6 core data quality dimensions defined by DAMA UK:

• completeness – ensuring that all required records and critical values in a data asset are present
• uniqueness – ensuring that each record is unique, with no duplicates in the data asset
• consistency – ensuring that data values do not contradict each other within or across data assets
• timeliness – ensuring that data is up to date and reflects the appropriate period for its intended use
• validity – ensuring that data is in the expected range and format, conforming to predefined rules
• accuracy – ensuring that data correctly represents the real-world entities or events it is intended to describe

You may adapt or expand on these dimensions based on your organisation’s specific needs.

Align data dimensions with your purpose

This will ensure you only measure what’s truly relevant. Not all dimensions will be necessary for every purpose in your plan.

Focus on the specific data fields that are most critical to your purpose and avoid the temptation to measure everything. For example, depending on the context, the completeness and accuracy of a postcode or house number might be crucial.

If you’re not familiar with the data fields in detail, involve a technical expert who knows the data set well to ensure you’re focusing on the right aspects.

The data life cycle

You must assess and manage data quality across all stages of the data life cycle – from planning and collection to dissemination and archival. This will help you identify quality issues early on. Focus on improving data quality as early in the life cycle as possible – for example, by implementing validation rules or ensuring consistent standards.

Different stages of the data life cycle require specific quality assessment measures. Make sure you have appropriate quality rules and measures for each stage.

Find out more about the data life cycle in the:

• training modules
• Government Data Quality Framework

Step 3: Assess current data quality

This allows you to baseline the current state of the data and then measure further assessments against this. Focus the assessment on critical data linked to its primary use.

You need to:

• acquire the data for measuring
• run quality checks
• document your findings

Who in your organisation should be responsible for this

Data stewards or business SMEs with a business understanding of the data they create or use.

They may need help from:

• data custodians or technical SMEs to translate the business requirements to technical application
• data analysts, or other business SMEs that use data analysis techniques, to perform the initial querying and analysis for the initial assessment

Run quality checks

Apply relevant metrics to measure how well your data meets its defined quality rules. Different checks may require different metrics – such as percentages, counts, true or false, or ratios – to track trends and identify issues.

Automate assessments

Automating your data quality checks using software or scripts can make the process faster, more consistent and easier to maintain. While manual assessments are valid, automation can reduce effort and improve accuracy. Consider the potential return on investment when deciding which approach to take.

Identify and log data quality issues

Use the data quality issues framework to determine when a suspected issue is a data quality issue, and how to assign it a priority.

You can then log any issues in the Data Issues Log in the DQAP template.

Document findings

You should keep a well-documented record of your data quality assessment findings. This will provide a benchmark for tracking improvements, identifying trends and addressing issues over time. 

Documenting results also helps future users understand past challenges, improvement areas and data quality limitations.

Keep your DQAP up to date with identified data quality issues and the targets set for each quality measure. Review results regularly with data users and communicate any issues and their potential impact to stakeholders.

Relevant training modules

The training will show you how to:

• use evidence-based criteria to measure and monitor data quality, helping you quantify quality, track changes and assess the impact of improvements
• link quality to purpose to understand how issues affect users and why certain problems matter
• take a proactive approach by checking quality regularly throughout the data life cycle to identify and resolve issues early
• communicate results clearly so users can choose appropriate data sources and adjust their work as needed
• plan and document actions in your DQAP, setting review dates to track progress
• account for delays where actions depend on factors such as funding or IT changes
• make a strong business case by comparing the cost and risk of action versus inaction to show the value of improving data quality

Step 4: Prioritise improvements and set goals

After carrying out your initial data quality assessment, you should:

• identify and prioritise areas for improvement
• investigate the most critical issues first to maximise potential benefits
• set realistic, well-defined goals for improving data quality, considering the priority of issues and available resources

Note that steps 4 and 5 may have dependencies on each other, so you do not need to complete them in a particular order. For example, you might need to re-prioritise issues after doing a root cause analysis.

Who in your organisation should be responsible for this

Data stewards or business SMEs with an understanding of the data they create or use.

If your organisation has data governance in place, your data governance board might establish working groups to agree the priority for improvements on a temporary or ongoing basis.

Data owners or process owners are accountable for the quality of the source data. They should share responsibility for setting the improvement goals with data stewards.

Identify improvements and decide priority

The results of your data quality assessment will indicate where you need to make improvements based on the issues you’ve identified.

Set a priority for the data quality issues based on the importance, impact and performance of the data asset. You can use the data quality issues framework to help with this.

Each issue should have an agreed action, target timeframe and assigned owner for resolution. It’s also important to define a review date for ongoing monitoring. You can record this in the Actions section of the DQAP template.

Define goals for data quality improvements

Focus on practical fixes that will address each data quality issue and align your data asset with the rules you’ve set for it. You can record these in the Data Issues Log in the DQAP template.

Agree high-level goals for improving the data asset’s data quality, for example:

• improving data quality by 5% a month
• minimising costs by prioritising process improvements over system changes where possible

Keep your DQAP up to date with target percentages based on the baseline results and realistic improvement goals.

When setting goals, consider the:

• importance of the affected data
• extent of the issue
• risk posed by poor data quality
• cost of improvements

Create a data quality dashboard

A data quality dashboard can provide an overview of data performance. It can help identify where to focus improvements and investment, and support decisions on balancing quick, tactical fixes with longer‑term, strategic remediation efforts.

Step 5: Identify root cause and take action

You should identify and document the root causes of the data quality issues you identified and prioritised in your initial assessment (step 3).

You then need to establish actionable steps to address the root cause of each issue, rather than just treating the symptoms. Cleaning bad data is only cost effective if you address the root cause first. This will prevent the same issues recurring.

You need to:

• identify and document the root causes of data issues – refer to the Causes section of the DQAP template
• assess and capture the resolution options for each data quality issue – refer to the Data Issues Log in the DQAP template
• assign an owner to resolve each root cause, and cleanse the data, in the Actions section of the DQAP template – include a target date for resolution

Who in your organisation should be responsible for this

Business analysts or data stewards should perform the root cause analysis.

They may collaborate with:

• technical SMEs or data custodians to identify system solutions or make corrections to data
• process managers for process improvements

Differentiate between systemic and one-off issues

Determine whether the problem is systemic or a result of a specific event, such as an incorrect data migration.

Select appropriate actions

Based on the root cause, choose actions that align with your organisation’s needs and will have the most impact.

Possible actions include:

• improving data management strategies and standards
• introducing data validation
• enhancing data storage
• system changes
• automating processes
• providing better guidance and training to staff
• addressing team culture

Evaluate costs and benefits

Assess the costs and benefits of potential solutions and develop an implementation plan, considering the trade-offs between fixing the issue and the value of high-quality data.

Step 6: Report on data quality

You should establish a communication strategy for reporting data quality to all relevant stakeholders. Regular and clear communication about data quality is essential to:

• ensure transparency
• ensure the data is used appropriately
• make users aware of any issues so they can determine if the data meets their needs
• help stakeholders who rely on data to make decisions to understand how reliable the data is

You need to:

• agree a list of stakeholders and stakeholder groups you need to communicate to
• draw up a communication plan explaining how you’ll communicate the results of your assessed data asset to each group, and in what format
  • include your DQAP in your communication plan
• communicate the results to all stakeholders, including target resolution dates and actual resolution dates
  • highlight any issues that were identified in the DQAP
• ensure that the quality of the data for the assessed data asset is known and understood by all stakeholders – they should understand both the strengths and limitations of the data
• schedule regular meetings to review progress and decision making – it’s important that you keep users informed about the improvements you’re making, and how changes may affect the data’s quality and usage

Who in your organisation should be responsible for this

Data stewards or business SMEs with an understanding of the data they create or use. Data owners and data users will also need to be involved.

Data owners should provide data users with the necessary, up-to-date information to make informed decisions.

Data users should:

• review the quality information provided
• consult with data owners
• consider all relevant caveats
• report any quality problems they encounter

What to communicate about data quality

Include details on:

• known quality issues
• the results of quality assessments
• cleansing or standardisation actions and goals
• data processing changes
• caveats
• context
• infrastructure

Use multiple communication routes

Employ various methods to communicate data quality, such as:

• data sharing agreements
• regular updates
• data asset documentation
• metadata

Responsibilities of data owners

Data owners must ensure their communication is clear, engaging, up to date, and provides users with the necessary information to make informed decisions.

Step 7: Repeat measure over time

You should have a process to continually assess the data asset, using consistent data quality methods to track changes over time. This will improve the data asset and identify any new data quality issues.

You need to:

• review and refine your initial assessment criteria
• establish the frequency of data quality checks – you can record the measurement frequency in the Details section of the DQAP template
• put feedback loops in place for data users
• regularly record and analyse the results of data quality checks
  • use repeated assessments to identify trends and improvements or deteriorations in data quality
• establish a schedule for taking appropriate action
• record the progress of issues – for example, by using the Data Issues Log in the DQAP template
• automate or streamline data quality checks where possible
• check you’ve completed all of the DQAP

When taking these actions, you might need to prioritise new improvement measures, refine your goals or identify the root cause of new data quality issues.

Who in your organisation should be responsible for this

Depending on the tooling and capability, one or both of the following:

• data stewards or business SMEs with an understanding of the data they create or use
• data analysts

Prevent future data quality problems

You cannot predict all future data quality problems, but you can prevent them through:

• proactive communication
• effective change management
• addressing quality issues at the source
• engaging with data architects to thoroughly test any new data assets, processes or IT systems you’re designing – considering data quality from the outset

Plan for and manage changes

Changes in data assets, such as shifts in collection or processing methods, can affect data quality. Data users and managers must adapt and plan for these changes.

If significant changes occur, make sure you assess the data quality of the data asset before, during and after the change. Compare the results to ensure the data remains fit for purpose.

Cultivating a strong data culture in your organisation equips people with the skills and knowledge to handle changing data needs, new sources, and system changes.

To manage and anticipate data quality changes, consider:

• root cause analysis
• regular communication with users
• proactive impact consideration
• quality integration in new systems
• thorough metadata documentation

Supporting resources

Training

If you’re using this guide for the first time, we recommend that you complete the Data Quality Action Plans e-learning course in the Government Data Quality Hub.

The hub also includes the following courses:

• Introduction to data quality – this introduces the content of the Data Quality Framework and principles you can apply to your data within your organisation
• Introduction to data quality assessments – this covers the importance of good quality assessments, how this should look in practice, and some concepts behind DQAPs

We recommend these courses to anyone with an interest in, or responsibility for, the quality of data within a business area or for a specific data asset. Typically, this will be people in your organisation who have a defined role that includes data responsibilities, such as:

• data owner
• process owner
• data steward
• business SME
• operational manager

DQAP template

You may wish to use the DQAP template to capture all the relevant information in your DQAP for each data asset.

Data quality issues framework

The framework defines a data quality issue, and explains how to identify issues when they occur and assign a priority to the issue.

Terms and definitions

Data asset

A container that holds one or more data resources. The most common resources are data sets and data services.

Data set

An individual, structured collection of data that can be accessed in one or more representations. Data comes in numbers, text, pixels, imagery, sound and other multimedia – any of which might be collected into a data set.

The scope and description of a data set will generally be meaningful to a business user of the data, containing fields and ranges of values they would recognise from their use of the data in its business context. For example, a project-based organisation might use an IT system to assign resources to projects. A data set called ‘Assignments’ that contains records of assignments extracted from this system might be meaningful to a resource manager.

Data service

A data service is a specialised service or application that provides access, management, processing or integration of data – usually through an Application Programming Interface (API).

A data service enables users and systems to retrieve, update or analyse data in a structured and secure way. Examples include:

• DEFRA Data Services Platform
• Office for National Statistics (ONS) Integrated Data Service

Data user

Anyone who interacts with data, even if their use differs from its original purpose. Users can include internal teams, other government organisations, businesses and the public.

Data owner

Data owners are generally senior leaders. They’re responsible for setting the data priorities and policies, and approving standards for data, within a specific data domain or business area they manage. For example, the domain might be financial data, or the business area might be finance. They’re responsible for the quality of data within their remit but are not responsible for individual data assets.

Find out more about data owners.

Data steward

Data stewards are business SMEs with delegated responsibility from the data owner. They manage the day-to-day operational activities to support data governance and implement the policies, standards and processes to improve data quality.

Find out more about data stewards.

Data custodian

A data custodian is responsible for the technical management and safeguarding of data assets. They collaborate with data stewards, business SMEs and analysts to resolve data issues, specify technical quality rules and run data quality checks.

Find out more about data custodians.

Data analyst

A data analyst provides specialist professional support to data quality initiatives. The scope of their responsibilities will vary. Generally they:

• are responsible for the initial profiling of data
• assist with the configuration of technical rules, analysis of data quality results to identify issues and identify trends over time

Information asset owner (IAO)

An IAO manages the risks to personal information and business critical information held within an organisation.

Find out more about the role of IAOs in government.

Process owner

An individual with responsibility for the end-to-end management of a specific process within an organisation. They ensure processes run efficiently and optimise data usage and quality within the process.