Great digital experiences do not happen by accident. We test with real users, measure what matters, and turn evidence into clear design decisions. This is how OSKY plans, runs, and acts on usability testing for not for profit organisations and peak bodies across Australia.
Tailored, not templated
There is no one size fits all usability testing. We tailor the design process to your goals, budget, audience, and the current state of your website. Sometimes you need a quick pulse check on a few key journeys. Other times you need a deeper information architecture rethink before rebuilding templates. We design the scope to fit the outcome you want, then choose the right mix of research methods to match.
How we right size the work:
- Desired outcome – donations completed, forms submitted, members onboarded, content discovered faster, or fewer support calls.
- Current state – brand new concept, redesign in progress, or a mature site with specific friction points.
- Budget and timelines – from a focused two week sprint to a more comprehensive four week program.
- Risk and privacy – we mask data, exclude sensitive pages, and follow strict consent and retention settings.
What we use and when
Hotjar
Best for insights on live sites or staging sites. We use it to see how visitors behave and feel in the moment.
- Heatmaps for scroll and click behaviour
- Session recordings to spot friction at fields and menus
- On page surveys and feedback widgets to capture intent and sentiment
- Funnels and events to locate drop off points in key journeys
Optimal Workshop
Best for early discovery and information architecture validation before you ship changes.
- OptimalSort for open, closed, and hybrid card sorting to shape labels and groups
- Treejack for tree testing to validate navigation paths and labels
- Chalkmark for first click testing on wireframes or mockups
- Reframer for structured note taking and thematic analysis
Hotjar Engage
Best for recruiting and running interviews or unmoderated user tests when you need targeted voices fast.
- Built in recruitment with screening to reach niche audiences
- Automates scheduling, hosting, recording, transcription, and note taking
- Supports moderated interviews and unmoderated user tests for prototypes or live journeys
- Handy when you want to streamline logistics and keep everything in one place
We typically start with Optimal Workshop to get the structure right, then use Hotjar Engage for interviews or quick unmoderated checks, and finally run Hotjar to confirm that the new design works in the real world.
A general process most teams follow
Even though every engagement is tailored, most projects move through these steps.
1) Align on goals and success metrics
Agree on outcomes and translate them into measures such as task success, time on task, first click success, error rate, and qualitative confidence.
2) Set up privacy, security, and governance
Obtain consent, mask keystrokes, redact personal data, exclude sensitive pages, set retention windows, and run all scripts through our Cyber team.
3) Plan the sample and recruitment
- Card sorting: 15 to 30 participants per audience group
- Tree testing: 30 to 50 responses to validate labels and paths
- First click tests: 20 to 40 responses per screen
- Moderated sessions: 5 to 8 participants per key persona
- Hotjar passive data: run for one to two weeks or until agreed thresholds are met
Recruit via mailing lists, panels, social posts, or on site intercepts. We include staff only when staff workflows are in scope.
4) Design the studies
- Card sort to learn how supporters and clients group content
- Tree test to validate the navigation without visuals
- First click to check whether users know where to start on a screen
- Moderated tasks for deeper flows like Donate, Become a Member, or Find Services
- Hotjar setup with template heatmaps, event tagging for critical actions, and short intent surveys
5) Fieldwork
Time boxed to move fast. Optimal Workshop studies open for three to five days, Hotjar runs alongside, and moderated sessions are recorded and timestamped.
6) Analyse and synthesise
Combine quantitative and qualitative evidence. Identify confusing labels, failure paths, rage clicks, dead clicks, scroll depth issues, and drop offs. Pull quotes and short clips, score severity and effort, and prioritise fixes.
7) Playback and decision
You get a clear, actionable pack rather than a data dump.
- One page executive summary
- Annotated screenshots and heatmaps
- A recommendations with example labels and menu structures
- A prioritised fix list ready for your backlog
- A short highlight reel for stakeholders
8) Iterate and verify
Fold improvements into a sprint, then verify outcomes with simple split tests or first click re tests and a follow up Hotjar review. We can add accessibility spot checks in the same cycle.
Example scopes by budget and need
- Pulse check
One week, focused on two or three journeys. Hotjar heatmaps and recordings, one short intent survey, and a compact fix list. - IA validation sprint
Two weeks. Card sort, tree test, and first click tests on key templates. Outputs include a revised IA map and label set. - Conversion deep dive
Three to four weeks. Moderated sessions on high value flows such as Donate or Join, plus Hotjar funnels and events. Outputs include detailed form and content recommendations and a verification plan.
What you will receive
- Executive summary and clear next steps
- IA map with label recommendations
- Heatmaps and recordings with time stamped notes
- Survey findings and key user quotes
- A backlog of improvements grouped as quick wins, medium items, and larger roadmap items
- Verification plan to confirm impact after release
Frequently asked questions
What is a typical number of testers to recruit?
Short answer: enough to find the biggest problems fast, then test again after fixes. Here are solid starting points we use and tailor per project.
- Moderated usability sessions
5 to 8 participants per primary persona per round. Two personas usually means 10 to 16 total. If time is tight, 5 to 6 total still surfaces the majority of high severity issues. We prefer two smaller rounds over one large round so we can fix and verify. - Card sorting with OptimalSort
15 to 30 people per audience segment for open or hybrid sorts. Go toward the upper end when the content set is large or labels are unfamiliar. - Tree testing with Treejack
30 to 50 completed responses for the full tree. If you have distinct audience segments, aim for 30 per segment on the key tasks. - First click testing with Chalkmark
20 to 40 responses per screen or variant. This is fast to field and gives confident directional signals. - Hotjar heatmaps
Around 500 to 1,000 page views per template stabilises the patterns on most sites. For lower traffic pages, run for a set period and review alongside recordings. - Hotjar session recordings
200 to 500 recordings focused on the target journeys will usually expose repeated friction, form failures, and common backtracks. - On page surveys
50 to 100 responses to an intent or exit survey is enough to see clear themes.
These are guidelines, not quotas. We right size samples based on traffic, the number of personas in scope, and the decision you need to make now. When in doubt, run a smaller round, fix what you learn, then run a second check.
Which platform is best for usability testing?
There is no single best tool. Use the platform that matches your stage and the question you need to answer.
- Structure and labels
Use Optimal Workshop- OptimalSort to learn how people group content and which labels make sense
- Treejack to validate that your proposed navigation helps people find things quickly
- First clicks on early designs
Use Optimal Workshop Chalkmark to see where people would start on wireframes or mockups. - Interviews and unmoderated tests
Use Hotjar Engage to recruit targeted participants and run moderated interviews or unmoderated user tests. It centralises recruiting, scheduling, hosting, recording, transcription, and collaborative notes. - Behaviour on the live site or staging
Use Hotjar- Heatmaps for scroll and click behaviour
- Session recordings to spot friction in forms and menus
- On page surveys to capture intent and sentiment at the right moment
- Funnels and events to locate where people drop out
- Deep dives on complex flows
Use moderated sessions via Engage or a secure video call. When we need observers, timestamped notes, and clips, we keep everything in Engage or add a dedicated recorder when required. This complements, not replaces, Hotjar and Optimal Workshop. - Interactive prototype path checks
If we are testing a clickable Figma prototype and need unmoderated path data, we may add a lightweight prototype tester such as Maze or Useberry. We use these selectively when they provide speed without adding complexity to your stack.
Our default stack is Optimal Workshop for decisions about information architecture and first clicks, Hotjar Engage for recruitment and interviews or unmoderated tests, and Hotjar for real world behaviour and sentiment. We tailor the exact mix to your goals, budget, audience, and where your website is in its lifecycle.
Engage vs Optimal Workshop: how we decide
| Scenario | Use | Why | Example for not for profits |
| Rebuild navigation labels and groupings | Optimal Workshop (OptimalSort and Treejack) | Measures findability before visuals and reduces redesign risk | Help supporters find Programs and Services within three clicks |
| Validate first click on early designs | Optimal Workshop(Chalkmark) | Fast directional signals across multiple screens without heavy setup | Where would a donor start the gift flow on the new Donate page mockup |
| Interview or test with niche audiences quickly | Hotjar Engage | Built in recruitment and screening, automated scheduling, supports interviews and unmoderated tasks | Speak with case workers across regional areas about the Intake workflow |
| Prototype concept feedback in a day | Hotjar Engage | Run short unmoderated tasks, capture quotes and clips for stakeholder playback | Check if content labels make sense to parents seeking emergency relief |
| Observe behaviour at scale on staging or live | Hotjar (Observe) | Heatmaps, recordings, surveys, funnels to confirm changes work under real traffic | Confirm that the new Member Portal reduces support tickets |
Typical flow
Start with Optimal Workshop to design and validate structure, use Hotjar Engage for interviews and quick unmoderated checks, then use Hotjar on staging or live to confirm the changes and monitor over time.
Does Hotjar have a usability testing tool?
Yes. Engage is Hotjar’s research product for user interviews and unmoderated tests. It includes panel recruitment, custom screeners, automated scheduling, built in video calls, recording, transcription, collaborative notes, and easy sharing. We use Engage for concept and prototype validation and for speaking to hard to reach audiences, while Optimal Workshop remains our choice for card sorting and tree testing.
Is usability testing better done remotely or in person?
Short answer: remote is faster and reaches more people, in person gives richer observation. We choose the format that best serves your goals, audience, and budget.
When remote shines
- Distributed audiences or limited time and budget
- Early stage IA and first click tasks using Optimal Workshop
- Behaviour on a live or staging site captured with Hotjar
- Larger samples needed quickly with minimal logistics
- Participants who prefer the comfort and privacy of their own devices
When in person is stronger
- High stakes flows where micro behaviours matter, such as complex forms or identity checks
- Sessions requiring assistive technologies or specific devices
- Participants with limited digital access or skills
- Co creation workshops and stakeholder observation to build buy in
A practical hybrid pattern we often use
- Round one remote for breadth and speed
- Fix the top issues
- Round two in person with a smaller sample for depth and validation
- Ongoing Hotjar on the live site after release to confirm improvements hold up
Logistics and privacy basics
- The same consent, masking, and redaction rules apply to both formats
- Remote sessions use secure conferencing with a quick tech check before each interview
- In person sessions are run in a quiet space with screen recording and clear observer guidelines
- Incentives, scheduling, and accessibility needs are planned up front so participation is easy
Will recordings capture personal data?
No. We mask keystrokes, hide content in defined areas, and exclude sensitive pages. We also use short retention windows and respect opt out choices.
Do we need large samples?
Not always. A handful of moderated sessions plus focused Optimal Workshop studies often reveal the top problems. Hotjar then scales confidence across real traffic.
Can we test before we have a build?
Yes. We can test labels, menu structures, and first clicks on wireframes or mockups to reduce rework later.
How does this fit with accessibility?
Usability testing complements WCAG checks. We can add basic accessibility spot checks inside the same cycle or engage for a deeper review if required.
Why OSKY
We design and support websites for mission driven organisations. Our team blends research, design, engineering, and cyber security so testing is fast, safe, and practical. We focus on evidence, not opinions, and we make it easy for your stakeholders to see the why behind every change.
Ready to improve your user experience
Tell us the outcomes you care about and your constraints. We will tailor a right sized study using Hotjar and Optimal Workshop and deliver decisions you can ship with confidence.
