TFT : A TOEFL Speaking Prep App


-1- If you have no professional product managers, don’t trust unconditionally in the demands of your boss.

I recieved a doc listing the most essential and required features from my boss at the very beginning. I didn’t quite realize why they were chosen because I was new to the company.

This start-up basically runs an online language school serving the students who are going to study abroad in China. In the doc I was told that we were going to build a go-to app for TOEFL learners. Actually, it’s the ambition of my boss, not the purpose that this app’s gonna achieve. What exactly will this app be like? A speaking tool.

According to the statics of ETS, the grades of Chinese is basically the worst among the asians. And what’s worse is that we have made no improvement during the last 5 years! Among the four sessions, speaking is definitely the most critical and difficult one but it’s strange that speaking is never truly emphasized by Chinese students. The main reason, according to my boss, lies in the absence of language materials. The ability of speaking is, in the essence, the number of topics that you can talk about fluently and clearly. So a web product has already been built — which allows learners to accumulate materials written by native speakers.

However, you still cannot make sure that a person can talk although he has remembered so many stories. Excercise and practice are also indispensable since you should turn those storied into your own language. I think that’s why that App shouldn’t be a simple translation from the web, but should serve a further goal.

In order to confirm that idea, we made a research basically with two questions. 1st, would you use your mobile phone to read and recite the speaking material? 2nd, would you use your phone to record your speaking answers? The answers are exactly what we had expected: the mobile screen is too small for long-time reading while it’s mobility and small size make it a great tool to record and review the audio answers.

-2- Design is a process of doing subtraction. If you have no idea which features should be cut, assume all of them are necessary and let users make the decision.

In my first draft, I have five icons on the bottom bar. “Learning”, “Speaking”, “Saved”, “Notifications” and “Profile”. Now, only three of them are left. Now it’s a pure tool to help you do speaking exercises. We know only three parts are essential because we whole group have achieved an agreement on what is our priority and what is the core of our product.

However, we still have numerous details that should be decided whether it’s important or not. There’s a fun story about the list view. Our boss proposed that maybe we should cut the answer list view and show users directly the best answers instead. I thought it was a genius idea for a MVP as long as the rule of “best answers” is well designed but my other colleagues didn’t agree. So we did a user research particularly about this feature and users told us that they preferred to pick the answer they want to read by their own. Sure, that made more sense.

Additionally, the statis can help identify the useless features because the clicks of users could be traced and calculated.

-3- If your company cannot afford user researches, having a well prepared chatty conversation with just a couple of users is also helpful.

One of the questions in the user research and how it works

“Being well prepared” is not easy because you have to set a goal, think of every question related to it, and stick to it. The method I used is visualizing your uncertainties through making diagrams like the picture in the upper left. You can create 2 or even more choices to figure out which one is the most liked.

In this case, it’s a question about if the list view of all answers is necessary. It aroused an argument because I believe that it’s super clean and straightforward to show only the best three answers by swiping left or right. However, it’s not a common practice that users would accept. Actually, most interviewees preferred to keep the list view in order to choose their favourate answer. I can understand that “the most liked” answer won’t asure it’s the one I’m looking for.

This method, to some degree, keeps the conversation organized and efficient. At the end, there will be a open question part which allows interviewees to express whatever they got from the conversation or about the product. This kind of private conversation also helps cultivate a group of royal and leading users.

-4- Defend your design, even it’s simple and unfinished.

In order to achieve the majestic vision in your mind, the MVP is the very first step that should be released asap. But the design of MVP is always so rudimentary that every team member seems to have a better design alternative.

My boss said that the value of our product is the content itself, so as long as we optimize the value of our product, it don’t even have to be an elaborated one. In a content-driven product, the cleaner and simpler design is , the better the content is served. Colorful illustrations or fancy transition animations are not an eye candy but a distraction in the MVP phrase.

You have the dazzle vision of what the design would be eventually like in your head but others don’t. So if your team members are not on the same page with you, you can either explain or insist on it. If it’s your boss, just start another user research.