3 Word Review

A fast new approach to writing and reading reviews

The brief

During the time at home caused by the coronavirus pandemic, I took the opportunity to further educate myself in a set of industry standard design tools. One of which was Adobe XD, known for it’s powerful prototyping functionality. I wanted to explore the full possibilities of the tool whilst working on an app idea I had for a while.
Online reviews of places and products often require users to enter both a review of varying lenths and some form of star rating. I wanted to create a product that increased the likelihood users leave a review by reducing time spend writing them to at least 75%.
Playing on the concept of shortform texts on Twitter and Summly, I had the idea to launch an app that hosts very short reviews with dynamic ratings assigned from them.

3 Word Review Prototype


Sky Office

Intuitive UX would sit at the heart of the product, streamlining the user journey to find places near them by implementing a set of search methods helping users explore bars and restaurants in their area.
The review mechanism was a key consideration and needed to be as short and minimal as possible, urging users to share their thoughts in a concise way. To avoid having to enter a star rating, I envisaged using an AI algorithm that analysed a 3 word review to decide whether it was positivie, neutral or negative.

My role and process

I designed all aspects of this product myself, to breathe life into my idea as well as to showcase my design skills. I focused my thoughts around two key areas of the proposal; finding a restaurant nearby, and making reviewing easy and quick.
While exploring search functionality on other apps, I felt that location based search options would work best for people on the move, exploring dining options near them. A map and accompanying list of nearby restaurants would cover that user group. If the user wanted to do a more specific search, a simple Search input field could be used for that.
Thinking more about how the AI rating algorhitm would work, I imagined that combining the result (positivie, neutral or negative) with other reviews of the same location would dynamically create a percentage rating. For example, if the algorithm established that 9 out of 10 reviews were positive, the rating would be 90%. This would be the equivalent of an average star rating, but without any user ever having to enter it manually. On top of this, additional AI could extract key words from all reviews on a single place and visualise them at the bottom of the page in the form of clickable tags.


3 Word Review on Mobile

Creating this app has helped me think deeper about the challenges and opportunities an app idea like this could bring. Even after setting some of the key requirements and moving into the design stage, new ideas often cropped up, such as sharing and direction options.
This app was primarily created to explore new prototyping tools and build on an idea I set myself, and having spent a few weeks creating this helped me to see the full potential of Adobe XD. I believe that the app concept touches on an interesting idea of dynamically generated ratings that could also be applied as a back-end solution into existing apps, setting new expectations for the users of how long a review process would take. By removing the manual entry of star ratings and reducing review lengths to a mere 3 words, I would argue that the time a user spends on writing a review has been drastically reduced; if an average Google review contains 30 words, my mechanism is already 10 times faster.