Say what we do in words people already use

[drmowinckels]

Why

People arrive unsure whether we can help them, and the copy is a large part of
why. The site described itself in terms a reader only understands if they
already know what we do — "data engineering", "data science team", "FAIR
principles", "containerization", "integrated solutions", "standardized data
output", "registry data handling 55%".

This rewrites the language across the whole site so a researcher can recognise
their own situation without knowing what any of it is called.

What changed

Services — now problem-led. Each of the seven leads with the problem the
reader arrived with, lists the situations that produce it, and says what you
get. The category name stays as a small secondary label, so someone who does
know the term can still find it.

Before   Data Engineering
         Collecting and managing data and keeping track of its lineage
         is a critical aspect of scientific research. We can provide
         guidance and advice on obtaining, cleaning and securely
         managing your data...

After    Your data isn't ready to analyse.
         - Sixty spreadsheets with inconsistent column names.
         - A Nettskjema export nobody has opened yet.
         - Scanner output that has to be organised before anything runs.
         What you get: one documented dataset you can actually use, and
         the script that produces it, so it reruns when more data arrives.

The named systems do a lot of the work. Someone who has never heard "data
engineering" knows exactly what "I have a Nettskjema and can't get the data
out" means. Every name used (Nettskjema, TSD, Educloud, Saga, Quarto) already
appeared elsewhere on the site.

Hero and contact. The hero now names the concrete things we do instead of
"capture, process, manage and analyse". Contact says explicitly that you do not
need to know the technical name for your problem, because that is the real
barrier.

Feature section. "Integrated Solutions / Standardized Data Output /
Customization on Demand" is replaced by what working with us is actually like:
we use systems UiO already pays for, you own what we build, we can teach rather
than do, and you can start with a single hour.

Skills. The percentage bars are gone. "Registry data handling 55%" is not
something a visitor can act on, and the numbers were not sourced from anything.
They are replaced by three plain lists — languages and tools, UiO and national
systems, kinds of data — so someone with MRI data on TSD recognises themselves.

Pricing. Explains that the rate depends on how close you are to the
department and that computing time is applied for and paid for separately. The
bullets describe what you get rather than listing line items.

CTA, portfolio, team, testimonials. Same rule throughout: ordinary words,
concrete claims, and an explicit invitation to ask even if your problem is not
on the list.

Also fixed

• The English service list had Workshops but no Data capture; the Norwegian had
  Datafangst but no separate processing entry. The two languages advertised
  different services. Both now carry the same seven.
• A duplicated skipToContent block in both i18n files.
• Dead .progress-bar-track / .progress-bar-fill CSS, now that the bars are
  gone.

Notes

• The duration field from the earlier draft of this PR is dropped, not
  left blank — engagement length varies too much to state honestly.
• This branch was rebuilt on top of main after Strip landing page to essentials, move detail to sub-pages #23 renamed content/service/
  to content/services/.

Test plan

CI covers prettier, the content accessibility check, a strict Hugo build and the
CSS bundle freshness. What it cannot check:

• Are these seven problems the ones people actually arrive with? This is my
  best reconstruction from the services, portfolio and testimonials — you
  have the real record in the booking notes and inbox.
• The practical: lines describe how you work, and I inferred them from the
  site and testimonials (e.g. working alongside a student comes from the
  DRCMR quote about "daily sparring and guidance sessions"). Check them.
• Norwegian throughout is mine and wants a native read. "Finansiøren krever
  en datahåndteringsplan" in particular — is that how a Norwegian funder
  context would phrase it?
• The skills lists are what I could infer from the team pages and portfolio.
  Add anything missing, remove anything we don't really do.
• Pricing now states "the closer to the Department of Psychology, the
  cheaper" as plain English for the multiplier table. Confirm that is a fair
  description of the policy.

Since the first review round

Barbara's note was that the page held too much information and felt scattered,
and suggested structuring it around the four pillars in CAPRO_services.docx.

Those four — GOAT, MRI, statistical analysis, Bevex — are not a service
taxonomy. Two are named products, one is a research domain, and one is already
the "Processing and analysis" service under another name. Making them the spine
would leave work that fits none of the four homeless.

So they go in as a second layer instead. The seven problems stay the spine
and answer can you help with my problem. A short block under them, on both the
landing page and /services/, names the things a visitor already has a word
for, and answers do you do the specific thing I came for — which is how the
would-be clients in Barbara's account actually arrive.

What we run and support — GOAT, Nettskjema, MRI, behavioural experiments.
Nettskjema is in there although the document never mentions it, because it is
the one name clients were reported to ask about most.

Two of the document's claims are deliberately not made:

• No MR scanning. The document describes collection through LCBC's
  MRI-trained staff at Oslo University Hospital. That is their capacity, not a
  CAPRO service. The card starts from scans that already exist, whoever
  collected them.
• Bevex is not advertised. It is in testing, beta is expected spring 2027,
  and the site should not sell what it cannot deliver. The card describes what
  we can do today — build browser-based experiments in JavaScript with
  tachisto, exportable to Cognition.run, Pavlovia or JATOS — and mentions the
  UiO platform as in development, with no date.

Open source page

It said "Open source we have built" and listed four things. The organisation has
twelve worth showing. Someone judging whether we work in the open was judging us
on a third of the evidence.

Added: tachisto, uiothemes, tsdrtools, neuromat, ebrains.kg, nva,
the Nettskjema data inspector, and the R project-management workshop materials.
Two new categories, Web apps and Teaching.

The reason the rest were missing is that the card template required a local
image resource, and almost none of them ship a logo. The image panel is gone:
each card is now the categories, the linked name and a sentence. Four orphaned
logo files are deleted with it.

Accessibility and performance

Four WCAG failures fixed, 8.5 MB of unused bytes dropped, the logo hinted as the
LCP element, card images no longer upscaled, a real social card instead of a bare
logo, and card links that name their destination.

Bugs found while building this

• .svc__deliver p never matched: markdownify returns text without its
  wrapping <p>, so every "What you get" box rendered unstyled. The CSS was
  added earlier in this same PR.
• The portfolio blurb fallback reads .Plain, which carries the renderer's
  entities, so "Norway's" rendered as the literal ’. None of the
  original four entries had an apostrophe, so it had never surfaced.

Still to do before merge

• Norwegian needs a pass. The plan is DeepL from the English, which is
  the source of truth. content/offerings/_index.no.md is two revisions
  behind its English twin — its experiments card has neither the platform
  list nor the tachisto mention.
• tachisto links point at the old URL. capro.dev/tachisto/ and
  github.com/capro-uio/tachisto are both 404; capro.dev/jspsych-studio/
  is the only URL that resolves, so that is what the two link: lines use.
  Note that renaming the repo redirects the repo URL but not the GitHub
  Pages project path — capro.dev/jspsych-studio/ will 404 once the repo
  moves, so those links have to change in the same sitting as the rename.
• GOAT wording. "Built through UiO's HUB-node initiative" is the one
  phrase in the block an outside reader will not parse. Keep only if it
  earns its place on credibility.

CI fix carried here

themes/capro-tw/layouts/partials/nav.html walked site.Languages, deprecated
in Hugo 0.156. Under --panicOnWarning that is a hard error, and the strict
build in CI failed on it on the first push of this branch. Both language
switchers now walk hugo.Sites, which carries the same weight ordering.

It is unrelated to the copy work, but it blocked this PR, so it is fixed here
rather than in a follow-up.

🤖 Generated with Claude Code

[github-actions]

✅ Preview deployed at https://25--capro-uio.netlify.app
Please check it looks ok before merging.

[github-actions]

✅ Accessibility: no issues

axe-core over 19 routes × 2 viewports (one page per layout), plus 69 manual checks.

All 69 manual checks passed.

_{Automated tooling catches roughly a third of accessibility problems. A clean run is a floor, not a ceiling.}