Usability Research

Usability is about people and how they understand and use things, not about technology -Steve Krug

As designers and developers we deeply care about usability of products we build. As a basic premise of user-centered design, if those who are supposed to use it, can't use it the intended way, we consider that a big failure. Often the desire to reach market first in order to have the first mover advantage supersedes usability because of its poor framing as aesthetics or visual design. But that soon dissolves if some competitor launches the same useful product, which is far more usable. A more usable product gets better word of mouth and eventually wins as a prime differentiator over all other similar products in the market.

Evaluating usability is like assessing the merit of design, also known as heuristic analysis. At Obvious, we specialize in building large scale products that reach millions of users. At that scale, there are no edge cases. Every missed detail affects thousands of people. The cost of getting it wrong is disproportionately large. To avoid those mistakes, we attribute usability to the following five components –

✅How easy is it for the user to learn to use the design for the first time?

✅How efficient are the users while using the design?

✅How much do the users remember if they use the design after a long time?

✅How many errors do the users make?

✅How satisfied are they with the overall experience?

For example, a young user base can quickly adapt to an interface like Snapchat while for an older audience, using filters and understanding the icons can be a real struggle. This is because different users understand the same product differently and the interactions between humans and computers are always evolving. To tackle these complexities, we practice the three step process that follows-

Step 1: Design the experiment to gauge usability

Write an example as it is in the usefulness doc.

From dhruv's notes -

Define the business case, which is maybe to ensure that users double check their flight details before making a payment. This is likely to lead to less cancellation tickets.

From this flow, pick a small part that needs to be tested for usability. For instance, because confirmation is always a step built-in as a fool-proofing step, it can become frustrating. It does not have a stand alone utility. So maybe, you're trying a slightly different approach, where confirmation is built into the checkout flow, and you now want to test if people are able to easily use that interaction pattern. You will have to be a little creative here.

Step 2: Design a detailed, realistic prototype

With our experiment ready, we build a prototype for our users to complete the tasks we have designed. This is a manifestation of the final product and is used to gauge how users interact with our final designs. While creating our prototype, we focus on the following —

Make it as real as possible

To gauge how users react to our designs, we need our prototype to mirror the final product. If our prototype is of low-fidelity, we’ll be left with several false positives and negatives at the end of the study. Which is why, a prototype for a usability study needs to be populated with the details that matter — thoughtful copy, representative icons and crisp user flows. We have a variety of tools to do this — from very high fidelity prototyping tools like Framer to writing actual code.

Make it opinionated

While testing for usefulness, to see what users naturally gravitate to, we try to place an equal importance on all our different concepts and flows. On the other hand, in usability studies, we use icons, typography and colors to guide our users towards a certain path. We do this to test if our designs are working well and nudging users towards the actions we want them to take.

Improve tips here

Step 3: Prepare, Test and Synthesize

Irrespective of whether testing and synthesis happen in parallel, or one after the other, the process remains the same. In addition to our time-tested approach for preparing, conducting, and synthesising user studies, we are mindful of the following aspects when testing for usability.