Comparing Sydney UX and Web Agencies for Government and Education

A government or education website gets judged by people who never chose to be there. A parent checking whether school's closed for the heat. A student on a bus hunting for a timetable. Someone filling in a council form in their second language, on a phone with two bars of signal. Any agency pitching for this work can call itself "user-centred". The real test is whether it can show you the users.

So here's a buyer's framework for comparing agencies on UX research and design capability for public audiences: what to ask, what a good answer sounds like, what a weak one sounds like, and the evidence worth demanding before you sign. It won't rank agencies for you. It'll give you a way to rank them yourself.

Key Takeaways

• Ask to see research outputs from a past project, not the process diagram. Interview findings and test reports aren't easy to fake.
• Citizens and school communities share a need for plain language and mobile-friendly pages, but a parent moving between school sites and a sector portal and a resident doing a one-off council task want different things from navigation.
• Findability is measurable. An agency that talks about navigation without a single before-and-after number has left a gap you should probe.
• Accessibility starts in the design brief. WCAG 2.2 Level AA is the current reference point, and your duty under the Disability Discrimination Act 1992 stays with you whatever the contract says.
• The real test comes after launch, when your own staff are editing. Compare the handover as hard as you compare the build.

Real User Research, or Decoration?

User research has a recognisable shape when it's real. Discovery workshops to agree what the organisation needs. Surveys to reach a lot of people cheaply. Interviews to find out why. User testing with actual users, not the project team. Journey maps built from what those people said and did.

Decoration has a shape too. Personas invented in a room full of internal stakeholders. "Testing" that turns out to be the comms team clicking through a staging site. A journey map nobody can trace back to a conversation. None of these is useless on its own (a stakeholder workshop's a perfectly good start), but none of it is evidence about the public.

A useful reference point is the kind of discovery described in the JMC Academy case study: surveys, interviews, user testing, competitor analysis and user journeys, carried out for the latest revamp of a site Webcoda has worked on across multiple iterations. For an education provider competing for students, that depth of discovery matters.

Questions to ask

• "Who did you talk to on your last public-sector or education project, and how did you find them?" A good answer names the user groups and the recruitment method. A weak answer describes a workshop with the client's own staff and calls it research.
• "What changed in the design because of something a user said or did?" Good: a specific decision that moved, with the reason. Weak: findings that happen to confirm the original concept in every detail.
• "Can we see a de-identified research output?" Good: a redacted test report, a journey map, a findings summary. Weak: only the polished case study page.

I'd give that last question the most weight. Anyone can write a methodology slide. Not everyone's got the working documents.

The Audiences Overlap, But They Aren't the Same

Government and education sites share a lot: plain language, task completion over browsing, trust (mostly, does this look official and current?) and pages that work on a phone.

The differences are where agencies get caught out.

Government audiences tend to arrive with one job to do and little patience: pay, apply, report, check. Some are stressed. Some aren't confident in English. Some are on older phones or patchy connections in regional areas, where a heavy page can be slow to load or fail to load at all. Navigation has to work for a person who'll visit once and never come back.

Education audiences come back often, and they're split. Students are on phones, between classes, wanting the one link they need. Prospective students are comparing institutions. Parents may be moving between a primary school site, a high school site and a sector portal in the same evening, so inconsistent labels across those sites cost them time. International audiences matter too: UTS College, a pathway provider to UTS, runs a separate Sri Lanka site alongside its main one.

And don't forget staff. Intranet work is UX work. The NSW Health 1 Reserve Road intranet includes a fully interactive floor plan you can filter by floor and search by keyword, which is a findability problem solved for people who work in the building.

Ask each agency: "How would you test with people who aren't confident in English, or who use a screen reader, or who only ever see the site on a phone?" If the answer's a shrug and a promise of "responsive design", keep looking.

Treat Findability as a Number

Navigation is often presented as taste. It isn't. It's measurable, and any government website developer worth shortlisting will have measured it.

Three published examples help. The Metropolitan Memorial Parks site uses navigation designed on cognitive-psychology principles, with automated testing and feedback from diverse user groups; its case study reports a 45% reduction in time-to content and a 30% decrease in administrative tasks. UNSW BusinessThink reports a 23% increase in visitors, a 31% drop in mobile bounce rate and a 70% increase in website searches. Visit Newcastle, a tourism site built for the research and planning stage of a trip, reports 40% more sessions, 23% more time on site and a 34% lower bounce rate, and won the Kentico best Government site award.

Treat any agency's figures, these included, as a starting point for questions rather than a promise. Ask what the baseline was, over what period, and which metric was agreed before design began. If they can't say, that's your answer. A good answer names a measure tied to a task ("time to find a form", "searches that end in a click"). A weak answer's "the client loved it". Lovely, but it isn't findability.

Accessibility Is a UX Decision, Not a Final Checklist

Accessibility bolted on at the end almost always costs more and works worse than accessibility designed in from the start. Colour, focus order, heading structure, error messages on forms: they're design decisions, and they're cheapest to get right in the design phase.

It's worth getting the rules straight before you compare quotes. WCAG 2.2 Level AA is the current reference point, though older contracts and policy documents still cite WCAG 2.1 Level AA, so establish which version each agency is quoting against. The Disability Discrimination Act 1992 binds your organisation, and hiring a supplier doesn't transfer that duty. Education providers also have obligations under the Disability Standards for Education 2005. The Commonwealth Digital Service Standard, part of the Digital Experience Policy since 1 January 2025, is mandatory for non-corporate Commonwealth entities only; NSW agencies sit under Digital NSW policy, which requires the current version of WCAG at Level AA, and councils set their own.

• Ask: "At what point in the project do you check contrast, keyboard access and form errors?"
• Good answer: during design and component build, tested with keyboard and assistive technology before content goes in.
• Weak answer: "We run an automated scan before launch." Automated tools catch some issues, not all of them.

Compare the Handover, Not Just the Launch

Here's the part most evaluations skip. The day after launch, somebody in your comms team (or a school administrator, or a faculty web editor) owns the site. If they can't edit it confidently, it decays. Fast.

The web content management system matters less than whether it's been configured for the people who'll actually use it. UTS College is a useful example: a 1,400-page site migrated to HubSpot CMS with its design preserved, then configured so non-developers could edit it. The case study notes that once the team was trained, Webcoda barely heard from them. That's the outcome you want. Infrastructure NSW on Umbraco is another: flexible content modules plus a bespoke project timeline with PDF export, so editors can build pages without a developer for every change. And UNSW BusinessThink was set up on Kontent.ai as a publishing engine for a large content repository with many collaborators, to streamline content creation and approval.

Questions worth asking any agency:

• Who'll edit the site after launch, and have you met them yet?
• What training do editors get, and is it recorded?
• What happens when we need a page type that doesn't exist?
• What are your post-launch arrangements, and what do they cost?

One honest caveat. If your site is small and your audience is narrow, you probably don't need a research programme on the scale of a university's. A focused round of interviews and some testing on the top tasks may be plenty. A good agency will tell you that, even if it shrinks the proposal. (It's an awkward thing to say in a pitch. Reward the ones who say it.)

Evidence to Demand Before You Shortlist

1. A de-identified research output from a past public-sector or education project.
2. One design decision that changed because of user testing, and why.
3. A findability or task metric, with its baseline and measurement period.
4. The WCAG version they're quoting against, and when in the project they test.
5. A handover plan naming who gets trained, on what.

Nobody gets public-sector UX perfectly right, us included. What you're really looking for is the agency that knows where its own gaps are and says so.

Where Webcoda Fits

Webcoda is a web development agency in Sydney, based in Surry Hills and working since 2005. We offer digital design (UX/UI) and website development, and our government work includes clients such as the Australian Institute of Health and Welfare, NESA, Newcastle City Council and NSW local councils, alongside education clients including UNSW, UTS College and JMC Academy. We build to the WCAG guidelines during design and development. We're a certified B Corp, a Microsoft Partner, an Umbraco Gold Partner, a Gold partner for Xperience by Kentico (with 110+ Kentico sites launched and 18 site of the month features) and a HubSpot Solutions Partner. If you already have an in-house research team, you may only need a build partner, and that's a fine conversation to have too.

Talk to Webcoda

Webcoda is a Sydney-based, B Corp certified digital agency that has been building websites and enterprise web applications since 2005, for organisations including Sony, BridgeClimb, the Australian Federal Government and NSW Health. We are an Umbraco Gold Partner, a Gold partner for Xperience by Kentico and a HubSpot Solutions Partner, so platform advice comes without a single-vendor agenda.

Get in touch to talk through your project, or browse our full range of services and platforms.