Witlingo VUI Reviews

Below are UX reviews from the Witlingo UX team of deployed Skills/Assistants from Alexa, Google Assistant, and Cortana.  The purpose of these reviews is to highlight best practices that are being used in deployed far field voice experiences as well as areas of potential improvement in functionality and usability.  Here are the best VUI practice guidelines that are used in our analyses. Needless to say that all of this, just like the emerging voice space, is an evolving work in progress. Please feel free to send us your feedback at vui@witlingo.com or via the form at the bottom of this page. 

 


The Reviewers


luciana brielle vinita_linkedin
Luciana Morais Brielle Nickoloff Vinita Atre


The Reviews


Jeopardy

vui_jeopardy

Overall Score: 3.6/5.0

Date Published: 10/03/2017

Platform: Amazon Alexa

Summary Review: A solid, enjoyable skill that nevertheless leaves the user unsatisfied.  Given that Jeopardy is the type of game that one could easily binge on, limiting users to six questions per day, and only during weekdays(!) is off-putting and comes across as if the publishers of the skill are not taking this skill seriously and are not interested in maximizing usage. Keeping scores across days as the user plays is good, and so is providing where the player stands against other players. The skill is also well designed and easy to use. All of this makes one want to use it more, and so is ultimately frustrated and left unsatisfied because they can’t continue playing.

 

Key Potential Improvement: Unless the user is allowed to play for a much longer stretch, this skill will disappoint.

 

Additional Analysis: Detailed Scorecard


Open Table

logo_open_table

Overall Score: 3.8/5.0

Date Published: 10/03/2017

Platform: Amazon Alexa

Summary Review: A solid skill all around, when it works. The two biggest weaknesses with this skill are the speech recognition and the pronunciation of the restaurant names. Otherwise, Alexa comes across as patient, helpful, resourceful (as she searches through other possible restaurants the user could be talking about), and effective, completing the entire reservation process. In addition (or perhaps as a result of) its speech recognition issues, Alexa doesn’t seem to use location services very well. Even if you specify a restaurant in a certain city, she sometimes picks the correct restaurant, but in the wrong city (which is probably the top hit). The skill also doesn’t remember the user’s past restaurants. Otherwise, a highly usable skill that delivers value.

 

Key Potential Improvement: Better speech recognition and better pronunciation of restaurants.

 

Additional Analysis: Detailed Scorecard


Capital One

logo_capitalone

Overall Score: 4.0/5.0

Date Published: 10/03/2017

Platform: Amazon Alexa

Summary Review: The skill is well designed and is a good fit for voice and for daily use and repeated quick interactions.  At times the skill does have the feel of a phone-like directed dialog like IVR, but otherwise it is effective and is highly usable.

 

Key Potential Improvement: The skill can go to the next level if it were to support mixed-initiative experiences (the user doesn’t answer the question asked of them necessarily but can ask for something else in response).

 

Additional AnalysisDetailed Scorecard


Fandango

vui_fandango

Overall Score: 3.1/5.0

Date Published: 10/03/2017

Platform: Amazon Alexa

Summary Review: A good skill that executes well on a challenging task: enabling users to discover information about  nearby movies and also book tickets for movies.  The skill is discoverable without needing to speak the Fandango brand name (“Alexa, I want to buy some movie tickets”) and enables the user to select from a spoken list by referring to the position of the item (“the second one”).  The inability to say “Repeat,” however, leads to a breakdown in the experience when the user wishes to listen to the options again, which is common while listening to a list of movies or screening times.  Fixing that one flaw would make this brave skill far more usable.

 

Key Potential Improvement: Implement the ability for a user to ask for information to be repeated.

 

Additional Analysis: Detailed Scorecard


The Magic Door

vui_magic_door

Overall Score: 4.4/5.0

Date Published: 10/03/2017

Platform: Amazon Alexa

Summary Review:  A solid skill all around.  The skill is engaging and makes effective us of audio cues and effects.  Long stretches of Alexa speaking in her native Text to Speech (TTS) voice can get tiresome while she narrates the story.

 

Key potential improvement: Using a voice talent or crafting the TTS more carefully would take the skill to the next level.

 

Additional Analysis: Detailed Scorecard


Kayak

vui_kayak
Overall Score: 3.6/5.0

Date Published: 10/03/2017

Platform: Amazon Alexa

Summary Review: Booking flights or hotel rooms via conversational speech is tedious.  The skill does respect many design best practices (it is concise, opens well and closes well, provides help, and flows well), but the linear nature of voice hobbles the experience.  Booking flights and rooms is something that is best done via a visual interface that enables the user to make changes and review the results quickly, which they cannot do with voice.  The skill does the best that it can within the bounds of an interface that does not lend itself well to a multi-step task.  The core flaw that the skill suffers from is its assumption that users begin booking knowing exactly what time they are leaving, what hotel they want, etc.  A strong element of search is usually part of the booking experience, and searching by voice is not ideal.

 

Key Potential Improvement: The skill should re-think its purpose.  For instance, a more compelling and useful skill would be one that provides the user with information about rooms and flights in a way that the user can express naturally: e.g., “What’s a cheap room near the Moscone center in San Francisco?” or “Any flights out tomorrow from National Airport to Boston that are less than $500?”

 

Additional Analysis: Detailed Scorecard


Your Feedback