Education web design serves several audiences on one site, and that is the whole difficulty. A school or college website has to work for prospective families comparing options, current students and parents looking for a timetable or a form, staff, alumni, donors and applicants, each arriving with a completely different question. Add legally required accessibility, a decentralised content culture where dozens of departments publish, and a governance process where nobody quite owns the homepage, and you have a project where the design is the easy part. The agencies worth shortlisting are the ones that talk about structure and governance before they talk about visual direction.
Design the information architecture before the interface
Most failing education sites fail structurally rather than visually. They are organised around the institution's internal departments instead of around what visitors came to do, so a parent looking for term dates navigates an academic division hierarchy to find them. The fix is to establish the top tasks for each audience first, from real evidence such as site search logs, the questions the front office actually fields and support ticket themes, then build navigation around those tasks. Ask a candidate agency how it determines top tasks and what evidence it uses. A shop that begins with wireframes and moodboards will produce something attractive that still buries the enrolment form four clicks down, and you will be paying to fix it within a year.
Accessibility is the governing constraint
Educational institutions serve the public, and the Department of Justice publishes guidance on web accessibility and the Americans with Disabilities Act explaining how inaccessible websites create barriers to access. In practice this decides the design system: colour contrast, visible focus states, properly labelled forms, meaningful heading order, captions on video and full keyboard operability all have to be built in from the first component rather than bolted on before launch. It also decides the training, because a decentralised publishing model means dozens of staff will add content for years after the agency leaves, and an accessible template filled with untagged PDFs and unlabelled images stops being accessible quickly. Ask what training and what publishing guardrails the project includes, and treat vague answers as a cost you will absorb later.
Governance, migration and the parts that overrun
Two things overrun education web projects more than any others. The first is content migration, because institutions accumulate thousands of pages across faculties, and deciding what to keep, rewrite or retire is slow work that somebody at the institution has to do. If the proposal assumes your staff will produce the content on schedule, say so out loud and plan for it, because that assumption is where timelines die. The second is approval, since a site touching admissions, communications, academic departments and senior leadership needs a named decision maker or every page becomes a committee. Ask candidates how they have handled both before. The ones that answer with a concrete process, a content inventory approach and a named approval structure have shipped an institutional site; the others have shipped a redesign for a small business.
How to compare candidates
Ask for two shipped institution sites at your scale and use them like a prospective parent would: find the term dates, find the admissions form, tab through it with a keyboard, load a page on a phone with a weak connection. Ask who owns the code and whether the content management platform locks you in. Ask how search visibility is preserved through migration, because a redesign that loses rankings on programme pages costs enrolment enquiries directly. Institutions commonly scope this as combined web design and SEO services so that redirects, page structure and content quality are one person's explicit responsibility. Finally ask what the site costs to run after launch, since hosting, support and accessibility remediation are real annual numbers that rarely appear in a build proposal.
Questions people ask about education web design
How long does an education website project take?
Institutional projects typically run several months to a year, and the pacing factor is usually content and approvals rather than design or development. Ask a candidate to show a schedule with your content production and review capacity built in, then ask what happens to the launch date if a department misses its content deadline by a month.
Should every department control its own pages?
Some autonomy is unavoidable and useful, but unrestricted publishing is how sites decay. The workable pattern is a locked template system with constrained components, a small number of trained publishers per area, and a periodic content review that retires what nobody reads. Agree that model during the project rather than discovering the need for it afterwards.
What about the PDFs?
They are usually the biggest accessibility liability on an education site and the least examined. Prospectuses, policies and forms accumulate as untagged documents that screen readers cannot navigate. Inventory them, convert the high traffic ones into real web pages, and set a rule for what may be published as a PDF at all going forward.
Do we need a new content management system?
Only if the current one genuinely blocks the work. Replacing platform, design and content simultaneously multiplies risk in one project. If your existing platform can support an accessible component library and your staff can publish in it, keeping it usually delivers a better site sooner. Ask candidates to argue against the platform they resell.