EORI Number

[shabana-ali]

EORI Number

Overview

During accessibility audits by the @hmrc/dias team, we are seeing variations in how EORI numbers are referenced, checked and asked for by service teams.

Establishing a pattern will ensure that:

• Service teams will know how to present and code when:
  • referencing a EORI number
  • checking for EORI number
  • asking for EORI number

We ask the HMRC design system working group to look into this taking into consideration what has already been done by service teams (see below for examples and recommendations we made with regards to accessibility issues identified) and develop the best practice.

• Hints and error messages are consistent across services. These should:
  • describe the expected format, e.g. this starts with ‘GB’ followed by 12 or 15 numbers
  • provide a clear example, e.g. for example GB123456123456

Content designer @fred-jackson has put together hint text for EORI numbers from the UK.

• Input and validation is handled correctly, that is:
  • use a single text input
  • allow the user to enter uppercase or lowercase characters
  • allow the user to enter numbers with or without spaces

Other components or patterns

There is existing content style guidance which is not always adhered by service teams. The EORI number pattern can reinforce or raise awareness of this:

EORI number There is no need to expand the acronym the first time it’s used. If you’re describing a specific EORI number for Great Britain or Northern Ireland you should use either:

• an EORI number starting GB
• an EORI number starting XI

Do not refer to GB EORI number or XI EORI number. Do not use EORI without suffixing it with “number”.

EORI number would potentially follow the patterns:

• Accounts Office reference
• Employer PAYE reference
• Unique Taxpayer Reference
• VAT registration number
• Match an organisation to HMRC records

Services with reference to EORI number

Check UK Destined status (within project UK Destined Validation service)

Example of a clear hint:

EU Subsidy Compliance

Checking EORI number (GitHub issue #3524):

The EORI number should be present in questions so that the information is available upfront for all users, i.e. a screen reader tabbing or using the form shortcuts may not be made aware of this information when needed. The page title may be kept as ‘Is this the EORI number you want to register?’ (TBD).

Current implementation:

Proposed solution:

Asking for EORI number (GitHub issue #3533):

When a user is asked to input an a EORI number, clear hint/error messages should be provided and accommodating validation to avoid confusion:

• The "GB" prefix is present but hidden from screen reader users so they may add this as part of their input. A user copying and pasting may have the "GB" prepended to the number so causing them additional work to remove this. Users should be allowed to type/paste the EORI number in full (including the GB). Prefixes and suffixes are of most use when they are part of a common term such as currency, weight or percentage, not where they form part of a reference number. We would recommend not using a prefix.
• Validation should allow spaces and uppercase or lowercase characters - this should be catered on the server prior to validation. For example, voice control technologies may add spaces when a user pauses reading out long numbers.
• Reference to ‘GB’ EORI number, which is not as per the content style guide.
• Hint/error messages are inconsistent and should provide clear guidance to correctly input or resolve the problem.
  • in a few scenarios the error message "GB EORI number must be 12 or 15 digits" is presented. It will be beneficial to add more information or an example format to the hint text to minimise the chance of showing an error.
  • there are some instances when the error message "Enter the GB EORI number in the correct format (12 or 15 digits), like GB497294577697 or GB729415776974533" is presented. The example format is not a valid input currently on this service.

EORI Common Component

Prefix added (GitHub issue #3235):

Common Transit Convention (CTC) Arrivals - also known as NCTS

Variation of the hint text and error messages used (spaces and lowercase not accommodated) when asking for EORI number (GitHub issue #961):

CDS EORI

Variation of the hint text and error messages used (spaces and lowercase not accommodated) when asking for EORI number (GitHub issue #1286):

C-18 Disclose an Underpayment of Import Duties or VAT

Example of a variation in hint text and inappropriate text field length on a smaller viewport (GitHub issue #2056):

Check your guarantee balance

No hint provided and unclear error message when asking for EORI number (GitHub issue #3014, GitHub issue #3013):

Claim for reimbursement of import duties (C285 form replacement)

Hint copy for EORI numbers from any EU country (GitHub issue #2296):

Must be between 14-17 characters and start with the first two letter of the ISO country code, GB for example: GB03152858027010

Example of checking EORI number along with Name. This is the only instance where EORI number has been grouped with another key value (GitHub issue #2288):

Fulfilment House Due Diligence Scheme (FHDDS)

Check an EORI number

NVDA and TalkBack read "XI" as "11" (for reference only, no action required, GitHub issue #949)

[Jon-Rowe-HMRC]

Thanks for the detailed submission, appreciated. Shabana and I have discussed this.

Planned next steps: The DR team will put together a draft guidance page to be shared with Shabana and the relevant service teams for feedback and iteration. If positive we will then present to the working group for further feedback with an aim to publish.

[terrysimpson99]

The guidance at: https://design.tax.service.gov.uk/hmrc-design-patterns/eori-numbers/ contains the text: "If an invalid character is entered Say ‘EORI numbers can only have letters and numbers’."

Can I suggest we say something similar to that said elsewhere relating to unwanted characters?

Here are a selection:

• "Use validation to ignore unwanted characters, ..."
• "You should ignore unwanted characters entered: ...as part of numbers and codes,"
• "ignore any unwanted characters before validating"
• "It’s easier for users if you accept and ignore unwanted characters."

[Louisehmrc]

Thanks, @terrysimpson99. This advice is already available in the guidance on recovering from validation errors, so I don't think we need to go into the same level of detail here. If this becomes a more widespread problem, we can consider how best to include or refer to that guidance in-page.

[terrysimpson99]

The reason I mention this is because it does appear to be a widespread problem. I see it frequently.

It's understandable: if the code is only letters and numbers, then it gets build to reject anything else. Users often don't report this type of error because they think it's their fault.

Not everybody knows about the guidance in 'recovering from validiation errors'. That's why it's repeated in other places where validation is mentioned. For example https://design-system.service.gov.uk/patterns/national-insurance-numbers/ includes the prompt: "ignore any unwanted characters before validating" There are other examples where something equivalent is said.

I was suggesting EORI does something similar.

[terrysimpson99]

Do XI EORI and GB EORI numbers sometimes/usually/always have identical digits?

[Louisehmrc]

Hi @terrysimpson99, I've added 'You should ignore any unwanted characters.' to the 'Asking for an EORI number' section, and also removed the invalid character error message as it's no longer needed.