In 2016, there was considerable focus on the evolution of the digital assistant (DA), and it makes sense. The prospect of being able to talk to our computers and achieve a level of productivity like never before is romantic. Unfortunately, we’re not there yet and that vision might not accurately reflect what people actually want. While each DA platform has continued to develop its capability, there is too much emphasis on party tricks and not enough on utility. In order to find out how the DA will be useful, we need to understand the jobs that it helps us complete.
In many respects, the DA is a solution in search of a problem. That's not to say there’s no value to DAs, but most of the big tech companies have yet to identify the utility. How do we solve this problem? I suggest we apply the theory of “Jobs to be Done” to this problem.
No, this isn’t Steve Jobs theory. Harvard Business School Professor Clayton Christensen, in his book "Competing Against Luck", outlines the idea of “Jobs to be Done.” In short, the theory argues that people hire products to accomplish specific jobs. He uses the word job instead of need intentionally. Needs are general and fuzzy - hard to pinpoint. For example, someone might feel the need to have a fulfilling career or a stable relationship. But, there are many factors that allow these to happen. A job is much more specific than a need. A person might hire a word processor (like Microsoft Word) to write his/her papers for a university degree. Obtaining the degree might lead to more fulfilling career, but the job is writing the paper. Jobs to be Done seeks to answer what caused that person to hired Microsoft Word over Apple’s Pages or Open Office, since any of those products could conceivably do the job of writing a paper. Jobs Theory seeks to discover causality.
Hiring a digital assistant
For which job would I hire a DA? I think tech companies are still trying to figure this out. I don’t have the research to explain the causality behind why people hire a DA. But from personal experience, and by observing others, I think people hire a DA to accomplish simple tasks when their hands aren’t free. I don’t think the DA is hired to be a companion or handle complex tasks.
When Apple unveiled Siri, they positioned it as an automation tool - for setting alarms, answering messages with dictation, setting reminders and calendar appointments, etc. Today Siri has gained a few more tricks, but it largely remains unchained despite Apple’s head start. Google has tried to make assistants smarter by incorporating the company’s machine learning and AI expertise. Google Assistant is better at providing factual answers than Siri, and at times it’s quite remarkable. The same is true when we compare DAs in the home. Amazon’s Alexa is great for adding alarms, setting timers, playing podcasts, and (ironically better at) adding items to Google Calendar. Google Home, in contrast, is better at providing factual information and answers to more conceptual questions.
But I would argue that factual information isn’t as useful. Rarely do I ask my DA to solve math problems or use Wolfram Alpha to answer a question. Google will read me the first paragraph of a Wikipedia article if I ask it about the 100 Years ‘ War. But what if I want to read past the first paragraph? Unfortunately, the machine learning aspects of the DA are largely party tricks. They’re great tech demos, but don’t help me accomplish key jobs that during my day.
The star of the Consumer Electronics Show 2017 was Amazon’s Alexa DA - originally a tabletop speaker called the “Echo” that is a voice activated DA. Today it’s being integrated into all sorts of devices from cars to pocket electronics. Some have argued it’s not as “smart” as Google’s Home competitor, which just came on the market. So why have so many individuals hired the Amazon Echo? It’s good at accomplishing simple tasks when your hands aren’t free.
Imagine you’re chopping onions for a soup. You don’t want to get onion juice all over your microwave to set a timer. So you say “Alexa, set a timer for 10 minutes.” Done. Imagine you’re washing dishes and you’re starting to get bored, but you don’t want to stop what you’re doing. So you say “Alexa, play my audiobook.” Now you’re enjoying your book while you wash up.
Apple’s iPhones also listen for the keyword “Hey Siri” and Google devices respond to “Ok Google.” I use these for the same purpose (because the Amazon Echo isn’t available in Canada). I hire Siri to make a note, set reminders, add calendar appointments, set timers, and play music when I can’t use my hands. Those are jobs that I routinely need to complete, and a DA does it well. Reading me part of a webpage isn’t a job I need done. I don’t need a robot to tell me a joke.
Jobs at home vs jobs in public
While I use both Amazon’s Echo (a home speaker) and Apple’s Siri (on the iPhone) in the same example, these are obviously two different devices. One is a home appliance while the other is a mobile device.
Unless I’m exercising outside, there’s more instances when I need a DA to assist me with a task at home. It’s not that I can’t ask Siri or Google Assistant to perform a task on the bus, but there is a social stigma associated with talking to oneself in public that doesn’t apply when at home. More often than not, I’ll just do the task manually in public.
It’s because of this social stigma (and the possibility of confusing those around you) that the DA is more likely to be hired at home than in public. This is where Amazon and Google have begun to understand the job to be done. Both the Amazon Echo and Google home are superior home devices - equipped with better microphones and speakers to accomplish simple taks via voice. So, when given the option, I think people hire DAs to automate simple tasks at home, when their hands aren’t free.
What the digital assistant needs to do better
Sure, there’s a laundry list of things DAs can’t do. I also appreciate the prospect of AI because I could have a more natural conversation with my phone (if you can call it a conversation). Memorizing keywords and specific phrases isn’t ideal.
But regardless of the language we use, the job we hire the digital assistant to do is the same: perform common, simple tasks quickly and efficiently hands free. Perhaps those tasks will become more complicated as DAs get better and as we become more comfortable thinking aloud. But, I doubt I will be asking Siri, Google Assistant, or Amazon’s Alexa to be my stenographer, transfer a file from one app to another, push a change to a webpage, or even read me an entire article.
There are certain tasks I will likely always do manually, and I think most people hire a DA to offload simple tasks to save time. So instead of improving party tricks or telling me lame jokes, Google, Apple, and Amazon should continue to allow me to automate simple tasks, but through the most natural language possible.