The Office for National Statistics is revamping its statistical bulletins in 2017. The following guidance is for those writing a bulletin. This guidance will evolve and be regularly updated.

What is a statistical bulletin? A statistical bulletin is a short summary of findings and essential commentary related to a new release of data. We encourage you to start writing each bulletin from scratch rather than reworking a previous release.

Editorial principles

Write your bulletin in a consistent format and style to help users navigate, consume, understand and use it.

To do this, follow our eight editorial principles for statistical bulletins:

  1. Bulletin means bulletin
  2. Tell a story through the data
  3. Find a topical angle
  4. Understand what your users want
  5. Don’t bury the story in stats
  6. Every section must have value
  7. Every section should be self-contained
  8. Write concisely in plain English

Bulletin means bulletin

Only include new and essential commentary unique to this release. Leave any information that doesn’t change from release to release in other parts of the website, such as Methodology.  

Avoid lengthy commentary. Use visual elements such as charts to help users understand the information at a glance.

Tell a story through the data

Your bulletin must tell the story that the new data reveals. Replicating the previous bulletin is likely to obscure what has changed or emerged.

Find a topical angle

If your data and its analysis can help shed light on current events then place it in the context of a topical news subject or story.

Understand what your users want

Keep in mind who the users are when writing your bulletin. What do they need to know and what is the best way to tell them?

Don’t bury the story in stats

Decide what the most important messages are and only include these. Include everything else in other, more detailed, articles and methodology reports.

Every section must have a value

Write and structure your bulletin so that every section justifies its inclusion. Will commentary help users or is the information better presented just as data?

Every section should be self-contained

Bulletins are not news stories with a beginning, middle and end. Structure your bulletin into sections which discuss different subjects or information. Users must be able to read and understand each section on its own.

You should not have a section that is simply used for information that doesn’t fit anywhere else, like Background notes. The new structure accommodates all your bulletin needs, and static information should be provided elsewhere on the website.

Write concisely in plain English

Put the most important information first. Follow our guidance for writing releases in plain English, and consider how users read information online.

Naming your bulletin

All release titles must include:

  • the name of the release
  • the geographical coverage
  • date or period the data covers, using the abbreviated month format (Jan, Feb)

An example of a title that includes all three elements is “Regional labour market statistics in the UK: June 2016”.

A title could include other information such as “provisional” or “final” where appropriate.


There are three sections all bulletins must have:

  • Main points
  • Things you need to know about this release
  • Quality and methodology

You can include other sections when needed, as explained in Writing your analysis.

Main points

This is where you communicate the most important information about your bulletin. Points should be ranked in order of importance from a statistical point of view. Each point should be a single bullet point and contain one message that is expanded upon in the release.

We advise a maximum of five or six main points; if you are getting close to 10 think about whether they are really needed. Avoid putting links in your Main points; it’s not good practice to point users away from your bulletin right at the start of it.

Each bullet point should be a single sentence starting with what’s happened, followed by the significance of this. Use a semicolon to split up the sentence if necessary. For example:

“The latest 3 month on 3 month figure shows an increase in growth of 1.6%; this is now the 31st consecutive period of 3 month on 3 month growth.”

Things you need to know about this release

This is where you make clear any vital information the user needs to know so that they don’t misuse or misunderstand the data.

This can include:

  • recommendations about how the data should be used
  • guidance on how to avoid misinterpretation
  • concise definitions of terms
  • any important methodology approaches or changes

Also include a standard bullet point for every release that has National Statistician’s designation.

We recommend that you break up this section with subheadings to make it easier to read and signpost users to the most relevant content.

Quality and methodology

Information about quality and methodology of the data, which helps the user decide how it can be used.

Use this section to communicate any minor changes to methodology or caveats that wouldn’t necessarily go in “What’s changed in this release” or “Things you need to know about this release”, but can help enhance the user’s understanding of the data.

Writing your quality and methodology

We recommend starting this section with a definition of the survey and what data are being collected.

This section must include a Quality and Methodology Information document (QMI) and can also include metadata documentation, if available.


Use the following standard text before linking to your QMI document:

The (add name and link) Quality and Methodology Information report contains important information on:

  • the strengths and limitations of the data and how it compares with related data
  • uses and users of the data
  • how the output was created
  • the quality of the output including the accuracy of the data

Other documentation to include

Include metadata documents which provide further information on relevant data quality, legislation and procedures. This can include documents such as user guides, technical articles and any dynamic quality information such as reference tables that changes every time your release is published.

Optional sections

Outside of the three mandatory sections and those you use to tell the story of your data, there might be additional areas you want to cover. For each of these, think carefully before using any of these common recurring sections; they should only be included if they are beneficial to the user.

Statistician’s comment

This section is for a comment aimed at the media, and if used it goes immediately after the Main points. The comment must be approved by the Media Relations Office. The heading can change to Analyst’s, Author’s, Researcher’s or Economist’s comment as required.

The comment should contextualise the main findings, drawing the information together to highlight something of primary interest to the media. It should not just be a figure that the user can read for themselves elsewhere in the bulletin.

The comment should be placed in double quote marks, followed by the name of the person quoted and their job title. If the person quoted has an ONS Twitter profile, include their handle as a call to action.

“Fewer people got married in England and Wales in 2013, the first decrease in marriages since 2009. The fall could indicate the continuation of the long-term decline in marriages since 1972 or could be due to couples choosing to postpone their marriage to avoid the number 13 which is perceived as unlucky by many cultures.”

Elizabeth McLaren, Vital Statistics Outputs Branch, Office for National Statistics follow @StatsLiz on Twitter.

What’s changed in this release?

Use this section to highlight changes to the release that are noteworthy but not crucial to the user’s understanding of the data the way content in “Things you need to know about this release” is, such as a change in methodology or a seasonal adjustment review.

Links to related statistics

Use for links to other publications. You should add context or further information about the link where relevant, explaining why the user might want to read it, rather than just provide a list of links under the heading. If you have a link that doesn’t need explaining, for example, in a marriages bulletin you may want to link to a divorces bulletin, then use the ‘You might also be interested in’ box instead.

You might also be interested in

This is a box on the right-hand side of the release page. It will only appear if content is provided. This is the place for links related to the publication and it only provides space for the link text with no context.

Upcoming changes to this bulletin

Use this section to highlight changes that will make future releases of your bulletin different from the latest version, from methodological changes to special events such as Chancellor of the Exchequer Budgets.

Whatever features in this section should appear in the next release’s “Things you need to know about this release”.

Writing your analysis

Once you’ve compiled the regular sections, the rest of your bulletin will consist of analysis of your data, containing only top-level information and essential analysis.

We recommend including no more than three or four analytical sections in order to keep your bulletin concise and focused.

Naming your sections

Each of your analytical sections should have a descriptive title that will tell the user a story at first glance. A good title will have an active verb, be a snappy, single sentence and encourage the user to read on. A generic title, such as “Introduction” or “June figures” is flat and uninformative.

Telling a story 

A good heading almost reads like a newspaper headline, and tells a complete story in itself.

England sees growth in private rental prices while Wales and Scotland remain static 

Labour productivity up in the first quarter of 2016, but the productivity puzzle remains

Asking a question

Alternatively, you can grab the reader’s attention by raising a question that will compel the reader to read your analysis in order to learn the answer.

What has driven the change to the long-term trend of falling producer prices?

At what age are people getting married?

Turning your data into a narrative

When writing a section, follow our guidance on writing for the web. Your text should be in concise, plain English and written with the user in mind.

Your analysis will expand upon the narrative established in your main points, giving the user more detail and context. When deciding what to write, consider what is the most important information that supports your main points, otherwise it can be left out of your bulletin. Users can still find the data in datasets and further detail can be explored in an accompanying article.

The first paragraph of each section must contain the most important fact. Use further paragraphs and figures to add more detail and context to broaden the user’s understanding.

Introduction: Productivity – as measured by output per hour – grew by 0.5% in Quarter 1 (Jan to Mar) 2016.  This leaves productivity 0.2% higher compared with Quarter 1 2008, just before the recent economic downturn.

Second paragraph: Quarterly growth of 0.5% is equal to the 1994 to 2007 average – but taken together with recent, weaker quarters, there is little sign of an end to the UK’s “productivity puzzle”.

Introduction to figure: This puzzle is illustrated by Figure 1, which shows two alternative measures of productivity – output per hour and output per worker – alongside their projected 1994 to 2007 trends.

Use of visual elements

Charts and interactive products

Use charts in bulletins to deliver a clear message visually. If a chart doesn’t tell its own “story”, consider whether it’s really needed. Annotating a chart can help add context to the data and reduce the need for text elsewhere.

Use interactive charts and maps sparingly when they can bring an extra dimension to your data. Contact the ONS Data Visualisation Centre on for advice and guidance on interactive charts.


Tables can capture a great deal of data but often aren’t the clearest or most transparent way of presenting the statistics or telling the story, such as a chart. If the user does want all of the data, they can download it. Only uses tables in your bulletin if there is a strong justification for doing so.

Further reading

Writing your quality information

Include quality information to help users understand data within your statistical bulletin and the quality implications for that data.

Help users avoid misunderstanding or using the data in the wrong context by making them aware of critical quality warnings or caveats on specific issues relating to the data.

There are three types of quality information within statistical bulletins:

  • Things you need to know about this release
  • Quality warnings within your analysis
  • Quality and methodology

Selecting and structuring your quality information

Your quality information should be:

  • structured in order of priority, with the most important item appearing first
  • what users need to know to avoid misusing data, rather than what they may want to know
  • regularly reviewed and updated to keep it current, relevant and helpful
  • the most important and relevant topics for that data – avoid “default topics” you update every month

Things you need to know about this release

Include the most crucial quality information, for example common pitfalls, that will reduce the risk of users misusing data, in ‘Things you need to know about this release’.

Read our guidance in the ‘What to consider including in your quality information’ section and Things you need to know about this release.

Quality warnings within your analysis

Include any critical quality caveats or warnings within your commentary alongside the analysis they relate to. These can also be included in the ‘Things you need to know about this release’ section.

Quality and methodology

This is the best place for any other quality information that enhances users’ understanding of the data and helps them to make decisions on its suitable uses, but doesn’t meet the criteria for ‘Things you need to about this release’ or relate to a specific area within your commentary.

This section should also signpost to further information on methods used to create the data and what the data is used for. Give a clear reason for the user to access the information so that they understand the benefit of doing so.

Read our guidance on Quality and methodology.

What to consider including in your quality information

Think about what users need to know to understand how to/how not to use the data within the bulletin. This could include, but is not limited to:

  • are there any common pitfalls or areas of misunderstanding from previous bulletins?
  • is there discontinuity; how does this affect the use of the data?
  • what is the most relevant and useful information from your users’ point of view?
  • is there uncertainty in the data and how does this affect its use?
  • what are the main data sources?
  • can you compare this data to other statistics?
  • what are the strengths and limitations of the data?
  • are there important issues affecting data, such as boundary changes requiring revisions?
  • what is the coverage, and what is not included?
  • what’s the periodicity?
  • what similarities and differences are there with alternative outputs?

This isn’t a definitive list of considerations. Different bulletins will have different quality information issues to raise with users.

Ordering your bulletin

A typical order for your sections would be:

  1. Main points (mandatory section)
    1. a. You might also be interested in (box-out in right hand rail, optional)
  2. Statistician’s comment (optional)
  3. Things you need to know about this release (mandatory section)
  4. Analysis section 1
  5. Analysis section 2
  6. Analysis section 3
  7. Links to related statistics (optional section)
  8. What’s changed in this release? (optional section)
  9. Upcoming changes to this bulletin (optional section)
  10. Quality and methodology (mandatory section)