How I conducted a heuristic evaluation on the Google Home
Based on personal experiences with the Google Home
During the ubiquitous computing era, the idea of verbally communicating with computers was seemingly too farfetched and futuristic. It didn’t seem plausible. Humans have always dreamed of being able to speak to a computer for an automated voice in exchange. Decades later of dreams and wishes, and with the inventions of the smartphone, Apple released Siri. I remember the vast amounts of excitement and curiosity surrounding this. People were amazed that anyone with a Siri-featured iPhone could literally speak to it, and ask it questions. Of course with Apple continuing their progressive innovative identity, other large tech companies wanted in. Those companies, like Google, would start to implement a modern digital virtual voice assistant to their smartphones. It began a new era.
This era immediately expanded the potential of our ubiquitous computing technologies. Already, humans around the world held smartphones in their pockets. So, the only way is up. And that’s when the virtual voice assistant expanded even further, leaving it’s beloved smartphone, and becoming its new identity. In 2016, Google released their very own smart speaker: Google Home.
Google Home speaker has brought parts of my life to ease, allowing me to ask it a question that saves me those few minutes in the morning that I firmly believe are so very crucial. Instead of having to locate my iPhone and relying that my “Hey Siri, what’s the weather like outside?” is heard, I ask my Google Home instead. It’s reliable, and always listens to me. It’s almost as if I’m speaking to my bedroom.
Being a devout user of the Google Home, but also an intrigued student of innovative technology and user interface/experience design, I’ll be performing a heuristic evaluation on the Google Home. This is an assessment of a system addressing its usability, pinpointing on its user interface design to identify both errors and effectiveness. The Google Home does not have a digital user interface, but a virtual user interface. The well-known Nielsen Norman Group has The 10 Usability Heuristics for User Interface Design that experts and designers around the world have attributed to its credibility. I will also be using the heuristics originated in Heuristic Evaluation of Persuasive Health Technologies (p. 3). I decided to embrace these heuristics from Kientz and her colleagues from the University of Washington in my analysis because I consider it more adapted relative to modernized technologies.
Visibility of System Status
The system should always keep users informed about what is going on, through appropriate feedback within reasonable time.
This heuristic is inherently the most essential to a system. Users must rely on relatively instant feedback. This revolves solely on verbal communication, which is how you interact with the Google Home.
There are some physical interactions a user can perform on their Google Home. The main ones are volume control, play/pause, and a long press and hold for the voice command to start. But, why touch the device when it’s meant to be spoken too? Evaluating this step should stem from a verbal communication standpoint; if the Google Home system hears the input and is providing some sort of feedback in return. Well, when you start off by saying “Hey Google”, blinking lights appear on top of the speaker, as shown below.
Once a user makes their request to the speaker, it starts to blink more, clearly indicating that something has forced the speaker to perform this action. The speaker informs its user that it’s listening and is processing a request immediately. It then verbally responds to the request from Google Home’s virtual voice assistant. Maybe, the speaker could provide a verbal command ensuring that it heard the request. I’d suggest having it say “Listening…”. Sometimes, the buffer time is extended, but technologies like these deserve patience. All in all, the voice assistant provides reasonable and reliable feedback.
Match between system and the real world
The system should speak the users’ language, with words, phrases and concepts familiar to the user, rather than system-oriented terms. Follow real-world conventions, making information appear in a natural and logical order.
In the small and recent world of virtual user interface systems, there has yet to be concrete universal behaviors and familiar language. These smart systems need to understand any sort of complex-worded request from their users. In most of these virtual voice assistants, it starts off with the commands, “Hey/Ok Google”, “Hey Siri”, or “Alexa…”. The Google Home utilizes simple vocabulary, easy enough for the general user to understand. The responses are genuinely cognitive and behavioral relative to a human’s response.
Google designed their speaker to be able to make those commands not so complex. Requesting what the weather is like outside, to turn on the bedroom lights, and to play a movie on a television, can only be so complex. It matches technological real-world conventions; the requests matches the ability of doing so in a physical interaction. But it makes it quicker, and more convenient.
As of September 2019, the Google Home supports thirteen languages. They need to add more languages so it can be supported around the world.
User Control and Freedom
Users often choose system functions by mistake and will need a clearly marked “emergency exit” to leave the unwanted state without having to go through an extended dialogue. Support undo and redo.
The Google Home does a great job giving their user control and freedom. If a user waned to end a request, it’s as simple as saying, “Hey Google, Stop.” Any sort of request can stop or be prevented with this command. Again, since this is a virtual user interface and commands are made verbally, it’s much more effortless in comparison to a digital interface.
Customizability
Users should be able to customize aspects of the technology, for example, creating personalized goals and customizing product settings (public/private data, interface, etc.). However, customizability should not interfere with persuasive aspects.
The Google Home App provides full customization for the speakers features. It’s a fantastic app, really. There are tons of personalized settings a user can play around with to give them their desired user experience. A user can generate personalized phrases for certain actions. “Hey Google, goodnight” can turn off your bedroom lights. However, there are limitations. You cannot change how you greet the speaker, this is an untouchable feature. You also cannot have the speaker read your texts, which has been a big concern surrounding its features. Once you change the settings, the speaker will accustom to it instantly. You know have your own personalized unique relationship with your Google Home.
Consistency and standards
Users should not have to wonder whether different words, situations, or actions mean the same thing.
The Google Home doesn’t have a vast, intellectual set of vocabulary. But that’s ok. It doesn’t need one. Especially with a broad range of users, keeping it simple is probably best. The virtual voice assistant’s behavior remains consistent. If a request is too complex or inaudible, it’ll respond with, “Sorry, I didn’t really quite understand that.” But, this stays consistent. If it doesn’t not comprehend a request, it will give the user this feedback. It does a fascinating job being coherent and fluid with their user.
Error prevention
Even better than good error messages is a careful design which prevents a problem from occurring in the first place. Either eliminate error-prone conditions or check for them and present users with a confirmation option before they commit to the action.
As a high-end innovative technology company, Google relies on their designers to ensure error prevention is extremely emphasized in their final product. Releasing a smart home speaker, users are going to be fully reliant on it performing perfectly nearly at all times. Knowing that, and being released two years after Amazon’s Alexa, Google had time on their hands to ensure that it’s an error-proof system.
The way Google designers ensured error prevention is with trusted reliable feedback. If a user requests: “Hey Google, remind me to take out the trash tonight at 8:00 PM.” After a slight buffer, it responds with, “Ok, I’ll remind you at 8:00 PM.” Simple. This is their method of confirming with their user that there were in fact no errors, and that the request was successful. If you ask to create an event, and have your Google calendar linked with it, it’ll appear in there. Another example of confirming the request was successful, and no errors occurred. And lastly, when the speaker doesn’t understand a user’s request, it responds with “I didn’t quite understand that.” Feedback is crucial in their design, and they emphasize significance through these methods.
Protect Users’ Privacy
The system allows users to keep personal information private. Users can control what, when, to whom, how, and how much information is made public. Any public information is kept abstract.
This is a huge concern with smart home speakers. Alexa, Google, and Siri have all been scrutinized for their products possibly listening to their users at all times, especially when the user is not interacting with the product. Google does collect voice from their users to analyze improvements of voice recognition. However, the speaker needs to always be listening. This is inherently how it operates. You can go to its settings and not allow Google to collect your voice and store it in their servers. This is still a concern for the future of privacy and data, however.
Usable and Aesthetically Appealing Design
The visual design of the technology is attractive and appealing and adheres to basic usability standards. The design captures and sustains the user’s interest, enhances user engagement with the technology, and also adds to the credibility and usability of the product.
The Google Home has a very simplistic design to their speakers. The Google Home Mini comes in different colors, but the engineered physical design and structure attracts user engagement. Both Google Home and the Google Home Mini have a clean and smooth color, with a fabric touch wrapping it’s speaker.
A concern residing with the design is usability. Users do not know at glance that they can physically interact with the speaker. There are no clean indicators or signifiers, it’s something a user needs to spend time with to figure out.
Help users recognize, diagnose, and recover from errors
Error messages should be expressed in plain language (no codes), precisely indicate the problem, and constructively suggest a solution.
There are some constraints within the system that require you to have other products or services. For instance, if there is a unique song on Google Play Music and you want to listen to it, it maybe only available on a Google Play Music subscription. Thus, it will suggest a playlist or a station on Google Play Music based on the song that you asked for. Neat. If it doesn’t know an answer to a question, it usually resorts to the next best answer it has, if comparable.
Conclusion
In summary, performing a heuristic evaluation provides significant insight on understanding the system from a user’s perspective but by analyzing it as a designer. Following experts own interpretations of heuristic evaluations broadens the analysis even more. This evaluation is an excellent UX strategy to ultimately examine and improve on a system. Pinpointing weaknesses of the Google Home is difficult; trying to criticize a universally known reputable smart speaker isn’t easy, especially one from Google. However, it can always be improved. The smart speaker can develop a better method of confirming the user’s input was heard. The smart speaker can be more of a speaker rather than just a virtual assistant, having a better sound system and bringing another pleasure to its users. It’s not indicative regarding how you physically interact with the speaker. And, maybe the Google Home can be better at hearing any input rather than having to be at a specific distance. These are all possible improvements that can be considered.
