Siri’s Job

This past week, I was camping in Tahoe with around 50 friends and family. This group comprised a lot of diversity around age, jobs, and income class. In our group were doctors and nurses, construction workers, police and fireman, software engineers, teachers, students, and those fortunate enough to be retired. While I try not to avoid any talk of work or tech when on vacation, I’m generally the person this group comes to with their questions on things present or in the future. For whatever reason, Siri was a hot topic this year.

Through a range of conversations around Siri, I heard things like “Siri is stupid,” or “Siri doesn’t know anything.” Some variation of this theme came up numerous times, as people shared their frustration with Siri and wondering what she is supposed to be good at doing.

As I dug deeper into what was going on, I realized a key issue is certain expectations that are placed upon Siri. To use business terminology I’m fond of, the question at hand is why consumers are hiring Siri. It became clear that most of the folks in our group used, or tried to use, Siri to search for things on the web and other general information tasks. This was the primary driver of their sentiment toward Siri. When I asked if she did things well like get directions, set a timer, set a reminder, etc., they all said Siri worked perfectly for those tasks.

These conversations and anecdotal data line up with our primary research on this subject. We recently conducted a study on smartphone virtual assistants like Siri and Google Assistant and found that searching for something, or general information queries were among the top five most common behaviors by both Siri users and Google Assistant (Hey Google, Ok Google) users. Interestingly, searching the Internet was the number one thing people do with Google assistant where it is the fifth most common behavior of Siri users. Siri users mostly use Siri to automate tasks that would take them time on their smartphones. Like setting a timer, setting a reminder, calling someone and texting someone.

The key behavior usages between both assistants is not a surprise since Google is, and will always be, better than Siri at searching the web. Google’s mission is to organize the world’s data and they will do that better than anyone. Therefore their AI agent will always be the best at search. The challenge we have today with consumer sentiment around Siri is with its weakness for general Internet search. If consumers do this regularly and want to use a voice assistant for searching and general information queries, Siri is not going to beat Google. Apple’s challenge is to help consumers understand the jobs where Siri is best.

In part, this seems the goal with Apple’s recent short video ads with Dwayne Johnson A.K.A The Rock. Both our data, and my qualitative conversations confirm Siri delivers on the job of getting things done or automating certain tasks inherent to iOS. Knowing Google’s mission is to organize the world’s data, and that mission will manifest itself with their AI agent, I’d interpret Apple’s mission with Siri to organize your life.

Apple is going to have to pick the battles Siri will fight, and I’d argue that organizing your life, and being the assistant of choice to assist you with life itself is the battle I think Apple wants to take on with Siri. Siri has the potential to understand me deeply and perhaps intimately. Siri will know about my family in ways in other digital assistants can. Siri will also understand more completely things like my preferences, core behaviors, likes, and dislikes, places I’ve gone, the food I like (since I take so many pictures of it), and much much more.

While Google will battle here as well in some cases, the reality is they have made their mission clear and are not (yet) trying to turn their assistant into a true personal assistant. We know this because Google Assistant has no name, which is required to build trust and intimacy. This is why Apple has Siri, Amazon has Alexa, and Microsoft has Cortana. Without a name, Google assistant users will only use it to a point but will likely not develop a deeper more trusted relationship with users of other assistants will. Given Google’s business model this makes sense. They would rather not lock themselves out of the iPhone, or Microsoft’s ecosystem and want iPhone users, Windows users, and even Amazon customers to still use Google services. Playing a cross platform game means Google is focusing Google assistant on information, and this is wise.

So this brings up a key hurdle I’m still not sure how we get over. All of these assistants will need to talk to each other because they all solve different problems and are hired for different reasons. Google is my search of choice, Amazon is my shopping platform of choice and Siri is my life assistant. I’d prefer to have Siri talk to Alexa (or the Amazon ecosystem at large), and Google’s search products and get me what I need by using the best mechanism to do so. Perhaps this is asking too much, but this scenario will yield the best customer experience.

The other scenario, is I have to use each assistant independently for different things. This is basically how I use them today. When I want to search for something on the Internet or do a general information query, I use Google Home. When I want to shop or get product information I use Amazon and the Echo. When I need to organize my life and get information on events, calendar, email, contact friends or family, etc., I use Siri.

At least for now, and perhaps the foreseeable future, none of these assistants is poised to succeed well and be hired by a consumer to do everything. They each have their glaring weaknesses and together are pretty powerful. I have a hunch consumers will have more delightful experiences with these assistants if they are hired for the right reasons. The challenge these assistants face currently is consumers do not yet know what those reasons are.

These assistants will ultimately become platforms. Right now there is a land grab for the relationship with a consumer by Siri, Alexa, Google Assistant, and Cortana. My thesis remains, these will succeed best if they remain specialized, however, they will all get along in an ideal world.