EDUKATE SINGAPORE · GOLD STANDARD PRACTICAL CASE STUDIES · CASE 50
A booking system can be technically successful while the community service becomes harder to reach.
A neighbourhood library moves its community literacy workshops to online-only booking. The new dashboard fills quickly and saves staff time. At the service desk, however, some prospective participants cannot use the form, verification process or device it assumes they own. The Gold Standard question is not simply whether the website works. It is whether the service remains accessible to its intended users.
This is a fictional library and a fictional rollout. It does not describe the National Library Board's actual booking policies or claim that any identified public institution operates a digital-only system. Public accessibility and digital-inclusion guidance is discussed for context, not as jurisdiction-specific legal advice.
Chapter index
- 1. The launch report says 94 per cent success
- 2. Capacity used is not the same as people served
- 3. Three barriers are not one accessibility problem
- 4. What public digital-service guidance supports
- 5. Restore an authorised path while investigating
- 6. Design the service as one route with several entrances
- 7. Test with the people the launch did not count
- 8. A filled workshop does not automatically prove equality of access
- 9. Check the repaired route during a busy week
- 10. Practice: the workshop is full; did the service succeed?
1. The launch report says 94 per cent success
A fictional community library runs four weekly reading and digital-literacy workshops, each with twenty places. Staff move registration to a new website. It records seventy-five confirmed registrations for eighty places in the first week, and the project dashboard labels the booking rate 94 per cent after rounding.
The librarian in charge can see why management is pleased. Paper slips are no longer being recopied. Participants receive confirmation messages and staff spend less time searching for names. The booking records are clearer than the old spreadsheet.
But the counter staff tell a different story. A retired reader who uses a basic phone cannot complete the online verification. Another visitor using assistive technology cannot finish the date selection. A parent with limited connectivity arrives in person and is told that the counter can explain the form but cannot submit a booking. None of these attempts appears in the 75-of-80 figure. The system appears nearly full while the service quietly loses some of the very people it was created to help.
2. Capacity used is not the same as people served
Seventy-five divided by eighty is 93.75 per cent, rounded to 94 per cent. That number accurately describes filled places relative to offered places, assuming the bookings are genuine. It says nothing by itself about the proportion of eligible people able to access the service, the fairness of registration opportunities or whether those who booked actually attended.
An invisible denominator matters here. The system counts successful online submissions, not everyone who tried to enrol. A person who cannot reach the first screen is absent from the conversion funnel entirely. Someone who encounters an inaccessible control may abandon the form before an error is logged.
A Gold Standard review separates occupancy, completed applications, assisted applications, technical failures, unsuccessful eligibility checks, waiting-list decisions and attendance. Each measures a different stage. The library should keep the valid 94 per cent occupancy figure without treating it as proof of inclusive success.
3. Three barriers are not one accessibility problem
The first visitor lacks the assumed technology: an online account, suitable device or reliable connection. Better colour contrast on the website will not give this person an internet-capable phone. A genuinely accessible service may need an appropriate supported route that can complete the same legitimate booking without requiring personal digital independence.
The second visitor has technology but cannot operate the date control with their assistive setup. This is a web accessibility problem requiring genuine keyboard and assistive-technology testing rather than a visual inspection alone. The third user has temporary poor connectivity; an expiring one-time code may turn a valid attempt into repeated failure.
Treating all three as 'needs training' would move responsibility from service design to the user. The correct response is to diagnose where the user journey breaks, then choose a route that preserves security, privacy and fair access without making staff invent unofficial exceptions.
4. What public digital-service guidance supports
The World Wide Web Consortium's Web Content Accessibility Guidelines 2.2 provides testable recommendations addressing a range of disabilities and access needs. Its role here is to inform how the online interaction should be assessed. Passing a website audit alone does not demonstrate that a person without connectivity can enter an online-only service.
The United Kingdom's Government Digital Service Standard states that services should consider users who lack internet access, skills or confidence and provide suitable assisted digital support. That is a UK service-design reference, not a claim about rules governing this fictional library.
In Singapore, IMDA's digital-inclusion programmes, including Seniors Go Digital and Tech Connect, provide real examples of support for people learning or using digital services. These examples illustrate why an organisation should not treat every user as someone with the same device, skills or confidence. They do not establish the imaginary library's actual policy.
5. Restore an authorised path while investigating
The library does not cancel every legitimate online registration or suddenly promise unlimited places. It preserves confirmed bookings under the existing published rules and offers an authorised review route to people who could not apply because of the system transition.
The service owner assigns trained counter staff to provide accessible assistance or capture applications by an approved alternative channel, subject to the same eligibility and capacity rules. If the rules need to change because the opening disadvantaged some applicants, the responsible decision-maker must approve and communicate the remedy; a helpful volunteer should not quietly change booking priorities on their own.
The team records the problem type and outcome without storing unnecessary sensitive details about individuals' disabilities, devices or household circumstances. Assistance should not require public disclosure of a private condition. The immediate Gold Standard response is a controlled, fair way forward, not a public list of people who struggled.
6. Design the service as one route with several entrances
An online form, a supported telephone booking and an accessible in-person process can all reach one authoritative capacity register. The booking must be accepted, acknowledged and given the same relevant information whichever permitted route the participant uses. Otherwise a new paper list will recreate the old duplicate-booking problem.
The owner defines the shared states: enquiry, application started, application received, eligibility confirmed if relevant, place offered, offer accepted, waitlisted, cancelled and attended. A counter enquiry should not be mistaken for an accepted place, and staff should not promise a seat simply because a visitor has told them their details.
One clearly responsible service manager controls exceptions. The digital team owns accessible interaction and reliability; librarians own the practical participant journey; the person responsible for the workshop owns capacity. The solution is coordinated service design, not competition between 'digital' and 'traditional' users.
7. Test with the people the launch did not count
The original test involved staff using modern phones on stable connections. Unsurprisingly, the form completed smoothly. A fairer retest includes people with different digital confidence, common assistive technologies and access constraints. Participants should help test actual registration and acknowledgement, rather than being asked merely whether the homepage looks easy.
Useful scenarios include a keyboard-only booking, a forgotten password, a failed verification message, a partially completed application, an in-person supported booking and an appropriate cancellation. The team also checks whether the staff assistance queue is practical when several users need support at once.
A test should capture successful end-to-end completion, accurate placement in the shared register, clarity of confirmation and the ability to seek help when something fails. Anonymous usability observation can reveal barriers that an error log cannot. Designers should not assume that one successful assisted registration proves the route will work for everyone.
8. A filled workshop does not automatically prove equality of access
After the repair, the workshop may still be fully booked. That is not itself a failure. Finite places have to be allocated somehow. The real governance issue is whether eligible users had a reasonable, rule-consistent opportunity to apply and whether the selection rules were communicated before the service opened.
The library should report how many people entered through each channel, where applications stopped and whether assistance was available when advertised. It should look for patterns that reveal a channel-specific barrier, while avoiding unwarranted conclusions about demographic groups from small or incomplete samples.
Neither a high digital take-up rate nor a small number of complaints proves that an excluded group is absent. People who have stopped trying are unlikely to appear in a complaint dashboard. Responsible managers ask how they would learn about those missing voices.
9. Check the repaired route during a busy week
A quiet test with one helpful librarian may pass. The system needs to operate when the workshop is popular, a counter worker is absent, the web form is temporarily unavailable or a participant needs help close to the booking deadline.
The retest checks who takes responsibility for the request, how capacity remains consistent, whether confirmation is delivered and how an unsuccessful application is explained. A paper backup is not useful if nobody knows how to reconcile it with the main register before seats are offered.
The library can credibly say it improved access only within the tested conditions. It should maintain a documented exception route and revisit the measures when the service or its audience changes.
10. Practice: the workshop is full; did the service succeed?
You are reviewing a programme with eighty places and seventy-five online registrations. An unknown number of visitors could not submit the form, and the counter lacked authority to register them. The manager proposes: 'Our online launch delivered an inclusive booking service with 94 per cent success.' What would you accept, challenge or investigate?
The defensible finding accepts 93.75 per cent place occupancy if the register supports it. It rejects the larger inclusion claim until the library has evidence about accessibility, assisted access, end-to-end booking and equitable treatment under the published rules. The unknown number of excluded applicants must remain unknown until investigated; it cannot honestly be set to zero.
Your corrective plan should define an authorised assisted channel, a single capacity register, accessible testing, fair remediation for affected applicants and evidence of how the routes behave at peak demand. The Gold Standard is not 'digital versus human'. It is a service that people can genuinely reach and complete.
Related reading and reference boundaries
- W3C · Web Content Accessibility Guidelines 2.2
- GOV.UK · Make Sure Everyone Can Use the Service
- IMDA Singapore · Tech Connect
- IMDA Singapore · Seniors Go Digital
- Gold Standard · Public Sector Digital Inclusion
- Gold Standard · Public Sector Assisted Digital Services
- eduKate · WCAG and Digital Accessibility
- Case 22 · An Accessible Route That Stops at a Locked Door
- Gold Standard · Core Requirements and Evidence
- Gold Standard · Casebook and Decision Practice
- The Gold Standard · Master Hub
All operational details are fictional. W3C WCAG 2.2 is web-accessibility guidance, not proof of offline inclusion; UK service-design guidance and Singapore digital-inclusion initiatives are cited only for their respective contexts. The GS-R01–GS-R12 requirements are eduKate's educational reference programme.