Website usability testing with real people, on a small budget
You know your site too well to see it clearly. Website usability testing on Tasklify puts your pages in front of strangers who follow a short script, record what they do and tell you where they got stuck, so you can fix the flow before it costs you customers.
Who uses this kind of user testing service
Tasklify suits teams that want answers this week rather than a research report next month: founders launching a landing page, designers comparing two layouts, e-commerce owners with a checkout that leaks, and developers who want to know if the new sign-up form is broken on Safari. If you have ever typed “test my website” into a search bar at midnight before a launch, this is for you.
Problems usability testers can find
- Broken flows: a form that will not submit on mobile, a coupon field that clears the cart, an email that never arrives.
- Confusing copy: headlines that do not say what you sell, pricing tables that hide the real cost, jargon in buttons.
- Navigation dead ends: pages people cannot find from the homepage, or menus that behave differently on phones.
- Trust gaps: the moment a visitor hesitates because there is no returns policy or contact page.
- Cross-device bugs: layouts that break on small Android screens, iPhone Safari quirks, or desktop browsers you do not use.
Designing a test script that produces useful answers
Good unmoderated tests give people a realistic goal rather than a list of clicks. Tell them what to achieve, not where to press, and ask about their expectations. Here is a brief we would be happy to receive:
Title: Find and buy a gift under $40 on our shop (desktop, 30 min)
Steps: 1. Start a screen recording. 2. Open the link below. 3. In your first 10 seconds, say or write what you think the shop sells. 4. Find a gift for a friend who likes cooking, under $40. 5. Add it to the cart and go to checkout. 6. Pay with test card 4242 4242 4242 4242, any future date, any CVC. 7. Stop the recording.
Proof: Screen recording of the whole session (upload up to 20 MB or share a link). Text answers: What did you expect the shop to sell? Where, if anywhere, did you hesitate? What almost made you leave?
A few rules of thumb: keep one scenario per task, set the estimated time honestly, and always ask at least one open question. Split a long study into several shorter tasks instead of one exhausting one; attention drops sharply after about 20 minutes.
Choosing who tests your site
- Countries: all, Tier 1 markets only, or a specific list (US, GB, CA, AU, DE, FR, IN, BR, ID, PH, NG, PK). Match the markets you actually sell to.
- Device: Desktop for complex dashboards, Android or iOS for mobile checkout, or Any to see the spread.
- Minimum approval rate (80, 90 or 95%) and minimum completed tasks (10, 50 or 200) to favour workers with a good track record on long tasks.
- Repeat set to once, so each tester sees the site fresh. First impressions only happen once.
- Plus-only if you want a smaller pool of regular, committed workers.
- Time limit of 1 or 6 hours for a 30-minute session, so nobody holds a slot all day.
- Review window of 24 hours to 7 days, and a scheduled start if you want sessions to begin right after a deploy.
Cost of a round of usability tests
For thirty recorded 30-minute sessions at the suggested $2 reward:
| Line | Calculation | Amount |
|---|---|---|
| Reward budget | 30 sessions × $2 | $60 |
| Service fee | 12% of the reward budget | $7.20 |
| Pre-funded when you publish | Budget + fee | $67.20 |
| Optional Boost | +20% of the reward budget | +$12 |
| Total with Boost | Budget + fee + Boost | $79.20 |
A lighter first-impression test works too: 60 people answering “what does this page offer?” in about five minutes at $0.35 each comes to $23.52 including the 12% fee. On your first deposit you get a 10% bonus up to $5; on a $67.20 top-up that is $5 of extra task budget. Top-ups start at $10. See all fees on the pricing page.
Reviewing sessions and keeping quality high
Recordings take time to watch, so plan your review before you publish.
- Pick a review window long enough to actually watch the sessions: 72 hours or 7 days for recorded tests.
- Skim first. A recording that is 40 seconds long for a 30-minute task, or that shows a different site, can be rejected with a written reason straight away.
- If only the text answers are missing, use request changes; the tester gets one chance to add them.
- Bulk-approve the sessions that clearly followed the script, then take notes on the rest.
- Check the duplicate-file and duplicate-text flags; copied answers across testers are a strong signal of low effort.
- Block anyone who submits fabricated proof. If a tester disputes a rejection, you can respond, and Tasklify decides, normally within 48 hours.
If you know you will be slow to review, switch on auto-approve so testers are paid when the window closes.
What doesn’t belong in a usability test
Do not use a “test” as cover for something else. Asking testers to post reviews of your site, leave comments or likes, sign up for accounts with invented names to pad your numbers, or click your ads are all prohibited under the Acceptable Use Policy. So is asking for real card details or ID documents.
Frequently asked questions
How many usability testers do I need?
For spotting the biggest problems in one flow, five to ten people often surface most of them; the same issues start repeating quickly. Use larger batches of 30 to 100 when you want to measure something, such as how many people complete checkout, or to compare two versions of a page.
Is this moderated usability testing?
No. Testing on Tasklify is unmoderated: people follow your written script on their own time and send back recordings, screenshots and answers. There is no live video call. If you need to interview participants in real time, use a dedicated research tool alongside it.
Can testers use a staging site?
Yes, as long as the link works for someone outside your network. A password-protected staging URL is fine if you put a shared test password in the brief. Never ask testers to use their own real payment card; give them your payment provider’s test card details instead.
Will testers think aloud on the recording?
Some will, if you ask, but many workers record silently, so do not rely on narration. Ask for written answers to specific questions as well, for example what they expected to happen when they tapped a button.
What if my website collects personal data during the test?
Tell testers to use made-up or throwaway details wherever possible, and only collect what you have a lawful basis for. Tasks that harvest personal data without one are not allowed under our Acceptable Use Policy.