Every project, paper and person, behind one filter
Website

The Water Institute at UNC's Gillings School works on water, sanitation and hygiene with governments, UN agencies and foundations. Its site carries 316 pages: 41 people, 36 funded projects, 46 publications and recordings, and twenty years of a conference where, in their own words, WASH professionals of the world come together.
An institute that size has one problem online. Everything it has done is interesting to somebody and almost none of it is interesting to everybody. So most of these pages are a filter, not a list, and the people who run the institute change what sits on them without coming back to me. Handing the site back is the point of most of my university work.
The shelf
Every publication, filterable by keyword
Forty-six papers, recordings, webinars and digests in one table, tagged by media type and dated where a date exists, behind a single keyword field. Somebody hunting for one specific thing does not want a reading list. On a page like this the filter is the page.
The client edits this themselves, in the CMS

The backing
Every project, with the funder named
Thirty-six projects, and the left column of every row is who paid for it: the Gates Foundation, UNICEF, the World Health Organization, World Vision, national aid departments. Current and completed filter apart. For an institute that runs on grants, the funder beside the work is both the credential and the invitation.
The client edits this themselves, in the CMS

The people
Forty-one people, in three groups
The director, then faculty and staff, then students, each with a photograph and a page of their own. At a research institute the roster is the product. Somebody weighing whether to apply, collaborate or fund is weighing these specific people, and they should be able to see them.
The client edits this themselves, in the CMS

The record
Eleven years of conferences, still addressable
Every Water & Health Conference back to 2015 keeps its own page, with the Nexus and Water Microbiology conferences alongside them. A paper that cited the 2018 program still resolves to something. An institute whose currency is citation cannot afford an archive that rots.
The client edits this themselves, in the CMS

The scale
The numbers the conference runs on
Twenty years running. Six thousand registrants. A hundred and fifty countries, two hundred and fifty-eight universities, seven hundred and eighty-nine organizations. They sit in a strip straight under the hero, so the case for being in the room arrives before any of the detail does. The first question anybody has about a conference is who else will be there.

Who is behind it
The sponsor wall, kept current by the client
The Gates Foundation, the Conrad N. Hilton Foundation, Osprey, the Water Environment Federation, Columbia World Projects, PLOS, IDEXX, set directly above the ask to become one of them. The list changes every year and the team adds to it themselves. Above the wall sits the case for attending at all, which is the page's other job.
The client edits this themselves, in the CMS

The questions
Thirty-eight answers, in five groups
Basic details; registration, payments and logistics; abstracts; side events; the Emerging Voices award. Each one opens where it sits, with a search across all of them. A conference drawing people from a hundred and fifty countries gets the same thirty-eight questions by email every year, and this is the page that stops them arriving.
The client edits this themselves, in the CMS

The proof
Last year's conference, in photographs
A carousel of 2025 underneath everything else: people presenting, people listening, people talking in the gaps between sessions. Somebody deciding whether to fly to North Carolina is deciding whether that looks like a room they want to be in, and no amount of copy answers that question.

How an institute keeps its own site
Six pages that used to be code
The homepage, the about page, the conferences page, the conference landing, the history page and the join page were all hard-coded. They are now content the institute edits, with the layout held steady underneath so an edit cannot break the page.
Editing happens on the page itself
Staff click the thing they want to change on a preview of the real site rather than hunting for the matching field in an admin screen. For occasional editors that is the difference between updating a page and asking somebody to.
The FAQ groups are theirs to rename
The five categories are content, not code, so the conference team can add one, reorder them or rename them the week before registration opens without filing a request.
The program embed only accepts one source
The conference program is published as an interactive document and dropped into the page by link. The field refuses anything that is not from the publishing tool it expects, so a pasted link cannot become an untrusted frame on a university site while everyone assumes it is fine.
Publishing is the only step
Hitting publish rebuilds the pages that changed, so an editor never has to wonder whether the live site caught up, and nobody has to deploy anything to correct a date.
Check out these other projects
Adrial is my go-to designer for anything my business needs … which is why I have used him for three different companies over the last 20+ years.
Craig WoodCEO, DefinianYour great work deserves to look great.
Twenty-five minutes. Tell me what you're trying to do and who needs to see it, and I'll tell you what I'd build first and roughly what it costs.



