UX design that tests the flow before you fund the build
Every project rests on assumptions about what people will understand and tolerate. Research does not remove the assumptions. It tells you which ones are wrong before you build on them.
What is UX design?
UX Design is the research and structural work that decides how a product or site should behave: user research, journey mapping, information architecture, task flows and usability testing. It suits Australian organisations about to invest in a build who want the expensive assumptions tested while they are still cheap to change.
Get a fixed written quote- Typical timeline
- 5 to 10 weeks
- What drives cost
- The number of audiences, the number of methods, participant incentives, and whether sessions are remote or on site.
- Best for
- Builds where a wrong structural decision would be costly to reverse
- You own
- Recordings, transcripts, analysis and every research artefact
- Built with
- Interviews, journey maps, tree testing, moderated usability sessions
Your handover
Recruiting Australian participants who resemble your users
A study is only as good as the people in it, and the fastest way to get comfortable, useless results is to test with colleagues, friends of the team and clients who already like you. We write a screener that excludes people who work in your industry, and we recruit against the segments the project actually cares about, including the awkward ones such as first time buyers or people who abandoned you last year.
- 01Research plan tied to named decisions
- 02Screener and recruitment of matched participants
- 03Interview and observation findings with evidence clips
- 04Journey map covering front end and back office steps
- 05Information architecture validated by tree testing
- 06Task flows for the primary and recovery paths
- Usability test reports with severity ratings
- Prototype iterations between test rounds
- Prioritised recommendations with effort against impact
Geography and timing matter here more than in other markets
Geography and timing matter here more than in other markets. Sessions have to work across a three hour spread between the eastern states and Western Australia, and regional participants often have connection quality that itself becomes a finding. We run most sessions remotely for that reason, and travel for contextual work where the environment is part of the story, such as a warehouse, a clinic reception or a job site. We also handle consent and data properly, since recordings of identifiable people are personal information under the Privacy Act 1988 and need a clear consent, a stated retention period and secure storage.
Each finding gets a severity rating and a location, so the team can triage rather than debate.
Journeys and where the failure actually sits
Journey mapping has a reputation for producing beautiful posters that change nothing. That happens when the map describes an idealised path rather than the real one. We build maps from evidence, marking each step with what the person is trying to do, what they actually experience, and what it costs them in time or confidence. The useful output is not the diagram. It is the shortlist of moments where people give up or ring you.
A booking flow may be flawless while the confirmation email lands in spam
Those moments are frequently not on the screen at all. A booking flow may be flawless while the confirmation email lands in spam. An enrolment form may work perfectly while the document upload requires a scanner nobody has at home. A quote request may convert well and then sit in an inbox for three days because no one owns it. Mapping the back office alongside the front end is what makes the difference, and it is why UX work often ends up recommending a process change rather than a redesign.
How the engagement runs
How many people you need to test with
Five participants per round finds most of the serious usability problems in a given flow, and a second round after changes finds most of what the first round hid. Teams often want thirty participants and one round, which feels rigorous and is worse: you gather a large pile of evidence about a design you then change, and you never learn whether the change worked. Small and repeated beats large and singular for anything diagnostic.
- 01Stage 1Define the tasks the product must support and what success looks like for each
- 02Stage 2Write a screener and recruit participants who match the target segments
- 03Stage 3Pilot the session with one participant and fix the script before the rest
- 04Stage 4Run five to eight moderated sessions, recorded with consent
- 05Stage 5Rate each finding by severity and pinpoint where in the flow it occurs
- 06Stage 6Change the prototype and run a second round against the same tasks
- 07Stage 7Report with clips, so stakeholders watch the problem instead of reading about it
Two decisions on your side that keep the project moving
We run sessions as tasks rather than demonstrations. Participants attempt something real while thinking aloud, and we resist the urge to help, which is the hardest discipline in the room. Each finding gets a severity rating and a location, so the team can triage rather than debate. Findings that block a task get fixed before launch. Findings that cause hesitation get logged and revisited with live data once traffic exists, which is where conversion testing takes over.
Choose the right level
Which research method answers which question
Research goes wrong when the method does not match the question. Asking people what they would do produces a prediction, and predictions about your own future behaviour are unreliable. Watching people attempt a task produces evidence. Survey data tells you how widespread something is but never why. Analytics tells you where people stopped but not what they were thinking when they did.
Method
01
Depth interviews
The question it answers
What are people actually trying to achieve, and what gets in the way
Typical scale
6 to 12 participants per audience
02
Contextual observation
The question it answers
What really happens in the workplace, including the workarounds
Typical scale
3 to 6 sites or sessions
03
Analytics and search review
The question it answers
Where do people drop out, and what were they looking for
Typical scale
Existing data, no recruitment needed
04
Card sorting
The question it answers
How do users group and name the things we offer
Typical scale
15 to 30 participants, run unmoderated
05
Tree testing
The question it answers
Can people find a specific item in the proposed structure
Typical scale
30 or more participants for reliable task scores
06
Moderated usability testing
The question it answers
Where exactly does the design fail, and why
Typical scale
5 to 8 participants per round, repeated
How we work this out during scoping
So we choose the method from the question rather than from habit. The table below is the shortlist we work from during scoping. Most projects use two or three of these, not all of them, and the temptation to run a large programme should be resisted unless the decision genuinely warrants it. The point of research is to reduce the risk on a specific decision, and if we cannot name the decision a study will inform, we do not run the study.
Information architecture, tested rather than argued
Navigation is the single most argued about element of any site and the easiest to settle with evidence. Internal teams name things the way the organisation is structured, which is invisible to the people outside it. A council names a section after a department. A health service names a service after a funding stream. Buyers search for the problem in their own words and never find it.
More on information architecture, tested rather than argued
We run a card sort to see how users group and name your content, draft a structure from the result, then tree test that structure with a larger group to measure whether people can find specific things. Tree testing gives a task success rate, which turns an opinion into a number and ends most internal debates in one meeting. We repeat it on the revised structure until the scores hold. This work also produces the URL and labelling plan, which matters commercially because structure influences how a site is crawled and understood.
When UX research is the wrong spend
If the problem is small, known and reversible, do not research it. A team debating a button label on a page that gets forty visits a month should choose one and move on. Research earns its cost when the decision is structural, expensive to reverse, or contested between people who each have plausible reasons. Anything else and the study becomes a delay tactic with a deliverable.
The rest of the answer
We also decline when there is genuinely no access to users and no willingness to create it. Research with the wrong participants is worse than none, because it produces false confidence with evidence attached to it. And if you already have meaningful traffic and a live product, the fastest learning may not be a study at all: instrumenting the funnel properly through analytics implementation and watching real behaviour for a month will often tell you more, and cost less, than interviewing people about what they think they do.
How we scope it
Four ways to scope your UX Design project
We do not publish package prices, because the same brief can be a short build or a long one. These are the shapes the work usually takes. Tell us which one sounds like you and you will get a fixed written quote that spells out exactly what it covers.
UX Core
The essentials, done properly, not a template
Fixed written quote, agreed before work starts
- Research plan tied to named decisions
- Screener and recruitment of matched participants
- Interview and observation findings with evidence clips
UX Full identity
A complete identity your team can apply without us
Fixed written quote, agreed before work starts
- Everything in UX Core
- Journey map covering front end and back office steps
- Information architecture validated by tree testing
- Task flows for the primary and recovery paths
UX Brand system
A system that holds up across products and campaigns
Fixed written quote, agreed before work starts
- Everything in UX Full identity
- Usability test reports with severity ratings
- Prototype iterations between test rounds
- Prioritised recommendations with effort against impact
UX Brand care
Applying and extending it as the business grows
Rolling monthly, quoted in writing
- New assets and applications as they come up
- The system extended rather than reinvented
- Files and source artwork kept in order and in your name
- Rolling, cancel with 30 days notice
These are shapes, not menus. Most quotes end up somewhere between two of them, and we will say so when the honest answer is the smallest one. Describe the problem and we will tell you which it is.
Questions buyers usually ask
Frequently asked questions
Ownership and handover
Who owns the research data and the recordings?
You do. Recordings, transcripts, analysis and every artefact transfer to you. We handle the material as personal information under the Privacy Act 1988, with participant consent covering how it may be used and a retention period agreed up front. If participants consented only to internal use, we mark those clips so nothing ends up in a public deck by accident.
Can we test with our own customers instead of recruited strangers?
Partly. Existing customers are excellent for understanding current experience and terrible for judging first impressions, because they already know your vocabulary and have learned your quirks. We usually mix them: some customers, some people who match your target audience but have never used you. The differences between those two groups are often the most useful part of the report.
Detail and edge cases
How long does UX design work take?
Typically 5 to 10 weeks. Recruitment is the usual bottleneck rather than analysis, particularly for narrow professional audiences where finding eight qualified participants can take a fortnight. We start recruitment in week one, in parallel with planning, and we run the analytics and search review while we wait so the first findings arrive early.
What drives the cost of a research programme?
The number of audiences, the number of methods, participant incentives, and whether sessions are remote or on site. Recruiting hard to reach professionals costs more than recruiting general consumers. Contextual work involving travel costs more again. We recommend the smallest programme that answers the decision in front of you, then quote it in writing.
We are pre-launch and have no users yet. Is research still worth it?
Yes, with a different shape. You cannot test loyalty or retention, but you can test comprehension, findability and whether the proposition makes sense to someone hearing it cold. We recruit against the intended audience and test the prototype on realistic tasks. That catches the structural mistakes, which are the ones that are expensive to unwind after launch.
Do you test with people who use assistive technology?
Where the product warrants it, yes, and we recommend it for anything public facing or delivered under government or health funding. Automated accessibility checks and expert review catch technical failures; testing with someone who navigates by screen reader or keyboard every day catches the things that technically pass and are still unusable. Providers working in health services usually include this.
Related services
Tell us what decision you need evidence for
Describe the build you are about to fund and the assumption that worries you most. We reply within one business day with a research plan and a fixed written quote.