From Scientific Depth to Clear Paths
- UX Researcher
- Information Architect
- Content Architect
The Challenge
The company’s website holds years of excellent knowledge: dozens of measurement techniques, a broad instrument catalogue, application notes, webinars and tutorials. But it is organised around three internal product pillars that no longer match the business, and its menus have grown beyond what a laptop screen could show.
The audience is unusually wide. Physics students and researchers, lab managers who buy instruments, decision makers scaling quantum computers, public funding bodies, existing customers looking for software, and future employees. Each arrives with a different question, and most of them know what they need, but not where to find it.
The company also wants to position itself as a reference voice in quantum technology, while still serving the full instrument portfolio.
My Role
I’m leading the user experience side of the project, from the first assessment of the existing site to the wireframes the client is now reviewing. I’m working alongside a project lead, a developer team and the client’s marketing, product and application teams.
So far I’ve run interviews with the client’s teams, mapped the customer journeys of nine audiences, and analysed traffic and search data to see how people actually reached the site. I’ve defined the information architecture, the taxonomies and the content of every page template, and I keep a decision log so every choice can be traced back to its reason and its source. Once the wireframes are approved, I’ll manage the handoff to the UI and development teams.
Deliverables
- Assessment of the existing site and its storytelling
- Nine customer journeys, scored against a validation checklist
- Information architecture and sitemap
- Mega menu and navigation rules
- Shared taxonomies for areas, techniques, categories, scale and content
- Content architecture: 22 page templates and a library of reusable modules
- Low-fidelity wireframes, desktop and mobile, in three releases (the first, covering 16 page types, is in review)
- Decision log and open questions, shared with the client
- Handoff to the UI and development teams (upcoming)
Discovery
The interviews and the data told a consistent story. Researchers think in terms of their experiment: the research field first, then the measurement technique, and only then the instrument. The current site asks them to start from the instrument.
Search data showed the same gap from the other side. Generic searches for an instrument type often landed on pages for discontinued products, because no page described the type itself. The most visited discontinued instrument still drew steady traffic from customers looking for manuals and software.
The journeys revealed the dead ends. Visitors from regions without an online quote reached a product page and found no way forward. Funding bodies, who approve large budgets, found no page written for them. Decision makers evaluating the flagship quantum system found a compelling story, but no clear page describing what they would actually buy.
Reframing the Problem
The current structure reflects how the company is organised, not how its customers work. Three internal pillars shape the menu, so the same knowledge appears in several places and the paths to a quote are hidden behind them.
That meant one page per piece of content, with relations generated from data rather than maintained by hand.
The question changed from “how do we fit more into the menu?” to “how do we let each visitor start from their own question and still reach the same, single source of truth?”
UX Goals
- 01 Let every visitor start from their research field, or from the instrument if they already know it.
- 02 Keep expert knowledge expert-level, while giving newcomers a way in.
- 03 Give the flagship quantum system both a story and a clear path to purchase.
- 04 Never end a journey in a dead end, in any region.
- 05 Make the site easy to run: one page per piece of content, links generated from data.
Process
I started from a qualitative review of the existing site and a benchmark of comparable scientific brands. Interviews with the client’s teams and the analysis of traffic, search and backlink data showed which pages carried real value and which could be merged or retired.
From there I built the information architecture in iterations, each reviewed with the client. Every decision is marked as agreed, recommended or open, and logged with its reason and source. Open questions go into a shared tracker, so the client can answer them asynchronously.
I then tested the structure against nine customer journeys, using a six-point checklist: entry point, clear next step, no dead ends, right content at each step, low effort, and a clear action at the end. Each weak point became a targeted recommendation, which raised every journey to the same level.
The Solution
Two entry points, one structure
The new structure leads with six research areas, each opening with a commercial layer (story, recommended products, proof) and leading to its measurement techniques one level below. Visitors who already know the instrument they need enter by category type, from the product menu or a guided finder.
A mega menu that fits the screen
The menu stops at the second level: sections, areas and pre-filtered selections, never long lists of techniques or products. Search, which understands abbreviations and legacy product names, takes visitors straight to what they already know.
Story and product, side by side
The flagship quantum system has two pages with two jobs: a landing page that explains why it matters, and a product page that shows what to buy, how it scales and how to request a quote.
Relations from data
Recommended setups, compatible options and related content come from the client’s own matrices and tags, so links stay correct as the catalogue changes.
Templates as a guideline
Twenty-two templates hold today’s content in consistent structures. Every module is marked mandatory, recommended or free, and has an ID shared by editors and developers. Editors can use the templates, duplicate them, or build new pages from the module library.
No dead ends
A local sales contact replaces the quote where it isn’t available, a page speaks directly to funding bodies, and customers find support for their instrument in one place.
Wireframes
With the structure agreed, I’m now working on low-fidelity wireframes, desktop and mobile, released in three stages so the client could review one part of the site at a time.
Each board shows the page, its mobile version, a note beside every block explaining why it is there, the alternative states (empty results, regional variants), and the modules the template needs. Every block carries the same ID as in the content architecture, so a decision made on one page is reused across templates and handed to design and development unchanged.
I’m testing each key page against the five main audiences (researcher, research group leader, student, existing customer, procurement) and scoring it from 1 to 10. Every low score becomes a concrete change, and the page is scored again. Some examples so far:
A menu tested on real content
Filled with today’s 48 technique links, the old applications menu was about 930 px tall on a screen with 770 px available. That test, shown to the client, settled the two-level menu.
A product page that informs, not sells
A clean hero with one outcome headline, a variant selector where a product has variants, and a quiet price line. Where prices are not public, the price box becomes a short request form, because lead generation is core to the business.
A finder that only asks what matters
A guided mode with up to four questions, chosen by the previous answer and skipped when they would not change the results, next to the classic filters.
A quote from anywhere
Measurement setups, product and option pages open the quote already filled; existing customers can ask for add-ons only.
Technical content kept, not rewritten
The client’s expert content moved into the new templates as it is; new copy is listed separately, so the client knows exactly what to produce.
Where It Stands
So far the client has received a complete, traceable structure: a sitemap, a two-level navigation, shared taxonomies, 22 templates and a module library ready for design and development. Most choices have been confirmed by the client along the way; the remaining ones are framed as recommendations or open questions, each with its reason.
Tested against nine customer journeys, the structure scored 7.4 out of 10 on average; with the recommendations in place, every journey reached 8.5. The project has now moved into wireframes, in three releases: the first is being presented to the client, and the other two will follow.
What I’m Learning
In a technical field, simplicity does not mean less content. It means a clear path to the right depth: the researcher, the student and the budget holder can all start from the same page and go as deep as they need.
It is also confirming how much an architecture depends on how decisions are made. Logging every choice with its reason and source makes it easy for a large client team to review the work, change their mind safely, and move forward together.
Scoring each page against the people who use it keeps the wireframes honest. A page that looks complete often scores low for one audience, usually the existing customer or the buyer, and the fix is rarely a new section: more often a single link, a clearer price or one less step to the quote.
Need to make a complex offering easier to understand?
Let’s connect research, architecture and accessible design into one clearer experience.
