Tithi Mondal

Guide6 min readUpdated

Usability Testing on a Small Budget: A Step-by-Step Plan

Learn usability testing on a small budget: a five-person plan, a ready script, guerrilla testing tips and what to do with results, for Indian founders.

By Tithi Mondal, UI/UX Designer

What usability testing is, and what it is not

Usability testing is simple. You give a person a task, such as "order a meal for two", and you watch what they do. You do not help. You do not explain. You note where they pause, tap the wrong thing, or give up.

It is not a survey, and it is not asking people whether they like your design. Opinions are easy to give and often wrong. Behaviour is harder to fake. If someone says "looks nice" and then cannot find the checkout button, the second fact is the one that matters.

You do not need a lab, a big budget or a research team. A phone, a quiet corner and five patient people are enough to learn things that will change your product.

When to test

Test early and test small. The best time is when you have a clickable prototype and before developers have built it. Changing a screen in Figma costs minutes. Changing it in code costs days.

You can also test a live app, a competitor's app, or an old version of your own. Each is useful. If you are not sure how a prototype differs from other design files, see wireframe vs mockup vs prototype.

How many people do you need

For a first round, five people is a good starting number. You will notice the same problems repeating by the third or fourth person, and that repetition tells you what to fix. If your product has very different user types, such as buyers and sellers, test a few of each rather than ten of one.

Run a second round after you fix things. Five people, then fix, then five more, teaches more than twenty people at once.

A step-by-step usability testing plan

Follow these five steps in order.

Step 1: pick three tasks

Choose the three actions that matter most to your business. Write each as a goal, not as instructions.

  • Good: "You want to send 2,000 rupees to a friend. Show me how you would do it."
  • Bad: "Tap the Send button, then enter the amount."

The bad version gives away the answer. Examples of good tasks: find a product and add it to the cart; register as a seller; book a vendor for a wedding; find your last month's report.

Step 2: find the right people

Test with people like your real customers, not your team, not your relatives who want to be polite. If you sell to shopkeepers, test with shopkeepers. If your app is for students in smaller cities, test with students in smaller cities.

Cheap ways to find them:

  • Ask existing customers on WhatsApp for 15 minutes of their time.
  • Message a few people from your enquiry list.
  • Ask friends to introduce one person each who fits your customer type.
  • Visit the place your customers already are, which is guerrilla testing, covered next.

A small thank-you helps: a UPI transfer for a snack or tea, a discount code, or free credit. Keep it modest and tell them upfront.

Guerrilla testing

Guerrilla testing means approaching people in public places and asking for five minutes. It works well in India because people are generally willing to help, and your customers are often easy to find: a market, a college canteen, a clinic waiting area, a co-working space.

Keep it short and polite. Introduce yourself, say you are testing an app, not them, and ask for five minutes. Offer a small thank-you. Do not test in a loud place, and always ask permission before you take notes or record the screen.

Guerrilla testing is rougher than a planned session, but it costs almost nothing and gets you real reactions fast.

Step 3: the script

Read the same words every time, so results can be compared. Here is a script you can use as is.

Opening

"Thank you for helping. I am testing the app, not you, so you cannot do anything wrong. Please think aloud: say what you are looking at and what you expect. If something is confusing, that helps me, so please say so. I will not be able to help during the tasks, but you can stop at any time. Is it all right if I take notes?"

Warm-up questions

  1. "What do you do, and how do you usually use apps like this?"
  2. "When did you last buy or book something on your phone? How did it go?"

Tasks

Give one task at a time, read it aloud, and hand over the phone. When they ask "Is this right?", reply "What would you expect to happen?" If they are stuck for more than a minute or so, move on and note it.

Closing questions

  1. "What was the easiest part?"
  2. "What was the most confusing part?"
  3. "If you could change one thing, what would it be?"
  4. "Would you use this? What would stop you?"

Run the session in the language your user is most comfortable with. For many users that is Hindi, Bengali, Tamil or another language, not English.

Step 4: run the session

  • Sit beside the person, not across, so you can see the screen.
  • Stay quiet. Silence is uncomfortable but useful.
  • Do not defend the design. If someone struggles, the design needs to change, not the person.
  • Note what they do, not only what they say. Tap errors, hesitations, and words like "hmm" matter.
  • Keep each session to about 20 to 30 minutes.
  • If you can, have a second person take notes while you talk.

Step 5: turn notes into decisions

After five sessions, put your notes in a simple table.

ProblemHow many people hit itHow serious
Could not find the cart4 of 5High, blocks purchase
Unsure what "Verify" means3 of 5Medium
Liked the green colour2 of 5Low

Fix in this order: problems that block a task and affect several people come first. Cosmetic comments and one-person opinions come last. Then change the prototype and test again with fresh people.

Common mistakes

  • Leading the user. "Do you see the blue button?" ruins the test.
  • Testing too late. After launch, every fix is slower and costlier.
  • Testing only on a laptop. Most of your users are on phones, often mid-range Android. Test on one. My article on designing apps for Indian users explains why.
  • Only listening to friendly people. Ask the sceptics too.
  • Fixing everything. Choose the top three problems and fix those properly.

Budget you actually need

A basic round can cost close to nothing: your time, a phone, and a few small thank-you payments. Paid testing tools and recruitment panels exist, but you do not need them to begin. If you also want an expert to review your product without users first, a quick UX audit checklist is a useful partner to testing.

If you are planning an app and want the prototype ready for testing, I build these as part of my UI/UX design service, and you can reach me through the contact section.

Frequently asked questions

How many users do I need for a usability test?

Five people from your real customer group is a good first round. You will see the same problems repeat, which shows what to fix first. Then fix and test again with new people.

What are good usability test questions?

Ask about tasks and experience rather than opinions: "How would you order for two people?", "What was most confusing?", "What would stop you from using this?" Avoid questions that hint at the answer.

Can I test a prototype before any coding?

Yes, and it is the best time. A clickable Figma prototype opened on a real phone is enough to find most navigation and wording problems.

Is guerrilla testing reliable?

It is rougher than a planned session but very useful for finding obvious problems quickly. Use it early, and follow up with planned sessions for important decisions.

Have a product in mind? Let us design it properly.

Tell me what you are building. I will reply with questions, a rough timeline and a quote.