Why I’m Excited About Voice

Today I’d like to talk about why I’m so excited about voice. I was sitting in traffic this morning on the commute into Boston. I work in Downtown Crossing doing client services, ad ops, big data research for an awesome startup focused. My commute lasts between 1–2 hours in each direction, and I do it a few times a week from New Bedford, a fishing (and once, whaling) city. While driving this morning, I was immersed in full realization that there were hundreds, if not thousands of cars around me and passing by me on the other side of the highway. Every single one of those people has a smartphone, and every single one of them has the need for what I am building. That thought made me smile, and grin, and smile. Cut scene and fade to me yelling out on 93N, ‘Wahoooooo!!!!’.

It’s not a small number of people either. Out of 140 million drivers in the US, they each spend an average of 47 minutes in the car daily. There’s a huge portion who spend 1–3 hours, and some who spend 4+ hours. I did the math and that’s a LOT of minutes. Yet no one’s created anything for them to do that is also safe. So we drive past people looking up at the road, down at their phones, up and down. Or they’re playing with their phones looking for podcasts and audio books and such. Yes, many are listening to radio, and podcasts have risen tremendously over the last several years. The point is that people have a LOT of time to spend here, and they are craving something to do with it.

I remembered back to a few months ago, sitting in traffic on a day just like this (well, colder and snowier), I felt a deep need, almost a pain. I felt that one huge thing missing in my life was the ability to do SOMETHING ELSE while driving or walking. I realized I could use software to address that problem, and felt an overwhelming desire that I had to be the guy to solve it.

But as this apropos screenshot from Speakway will show, it was time to be brave. Honestly, it took a lot of courage to get going. When you have a big family who relies on you, and you’ve had some harsh failures in the past (more on that some other time), you pay more attention to where you spend your exploratory time. Mine comes at night, usually between 9 and 2AM. It was only after a few weeks of tinkering and research that I decided to commit to the project.

At first, I wanted some buttons mounted on my steering wheel that I could press and trigger something in an app. The app would maybe blast some kind of noise through my car speakers, and pretend I was piloting a space fighter through the endless sea of brake lights and blowing them up. I prototyped a few hardware solutions, but all were outside my current cash flow to R&D and bring to market.

I also wanted to interact with the drivers around me, and came up with a half dozen or so ways to do that. This kind of exploration led me to do a ton of market research, and I set up a Facebook page to sort of track my entrepreneurial journey. I also set up a huge spreadsheet model of different kinds of revenue channels and verticals and it looked really promising. I ran some peer surveys to find out who was/wasn’t interested, or felt this pain, and was greatly encouraged.

Then, a good friend mentioned the possibility of using voice. It had been years since I tried to do speech recognition, and it was pretty bad back then. I respect his opinion a lot, so I played around a bit and realized it was actually good enough to use right now. Even better, I thought, it was bound to improve and get more accurate over time.

At this time, despite being pretty well connected to the tech world, I had still been focused on Hollywood, big data, interactive cinema technology and ad tech, which I had been doing for the last several years. I hadn’t even been paying attention to things like Amazon Alexa. In fact, Alexa and Echo didn’t even come up in the searches I was doing to find existing ways to solve my problem. I think this is because I was focused on ‘in-car’ experiences, and they are primarily focused on home automation and shopping. Yet about a month into my research I did learn about Alexa, and took the time to catch up about what they were up to, and also learn about the AI and chatbot space.

Creating a simple Google News Alert showed me that Amazon was quite busy, and there were already some huge names making moves in this space.. MY space. In fact, they were trying to integrate into some automobiles already. This was a massive slap in my face at first, but then that turned into a feeling of validation. After all, if it was attractive and deep enough to warrant Amazon’s attention, that was probably a good thing. However, they still seemed focused on shopping and home automation, and are rapidly iterating on products which are really great at those things. I tried my hand at building an Alexa skill, and saw some things I didn’t like from a developer perspective, flexibility, and especially monetization.

What I wanted to do was give people awesome things to do while driving. Games. Interactive Stories. Chat. Research. AI assistant stuff. Multiplayer stuff. I wanted to create a voice-first platform that could connect to other chat-bots as well, and I made a list of 125 or so different experiences I wanted to make (I’ll talk more about this list at another time). I ranked them by difficulty to create, size of market segment, profitability, and demand. That gave me a good 20 to prioritize, but I would first need to build the platform to run them all.

I dug right in and started putting the pieces together. I thought I’d end up sending out 10 or so experiences in the initial launch, and prototyped a bunch of things. Chat, casino games, personal productivity, strategy, card games, trivia, group word games, social games, etc. A good part of my hobby and game dev life has been dedicated to role-playing and fantasy games, so I also kind of HAD TO make an RPG style experience, I called it “Instant Combat” and my youngest son heard me say it so many times that he still walks up to my phone and tries to say it, though I’ve since changed the name to ‘Adventure’.

So after prototyping one each of lots of kinds of experiences, I talked to another friend who is always out looking at what’s cool and different. He suggested I stop trying to do everything at once, and focus on just the RPG thing. So I did. I still switched gears from time to time, but focused. I was already part of some large RPG groups, so I made some brief exploratory posts to gauge interest, and found what I was looking for: some very passionate people who were absolutely interested. I don’t know how many, but that’s a good start. I used to really have a tendency to just build what I wanted to use or play, but that’s a terrible way to make an app that lots of people will use.

I’m glad I did decide to focus, otherwise I never would have reached my goal of shipping in June. All the other experiences are queued up to be finished, but I’m shipping with ‘Adventure’. I just submitted it to the App Store for review. The Android version will come soon as it’s still a little buggy for speech recognition (my fault, not Android’s).

I’ve used Speakway a LOT. I’ve found lots of rough edges and things which needed to be smoothed out, and tackled them one at a time. I’ve used it on walks also, and that’s not only super fun but has made me walk longer than I normally would. This is a surprising but very welcome side effect. I ran a small alpha test and got some great feedback already, addressed many concerns. The ToDo and bug list are immense, but I’m putting it out there despite the nervous feeling of ‘everyone will hate it’ that comes with releasing anything. I’m sure my friends and family are also quite sick of seeing me rant and rave on Facebook about my project too, but that’s not going to end soon :)

I have some ideas on how to market it that I’ve been collecting over a few months, but I’m new to consumer marketing in general. I have no budget, and no idea what I’m doing. But I do know how to use data, and have been reading/listening to whatever I can. The goal of releasing now, though the first release is probably pretty buggy (and I’ve long ago forgotten how a new user might use it), is to gather information about what’s working and what isn’t from a content/flow perspective.

The interactive audio platform I ended up with today is actually very extensive, and will handle a huge portion of that big ole list of experiences I would like to put in it. I used some “off the shelf” components, and poured 3 months, many gallons of midnight oil, and a good ten thousand lines of code, server backend processes, database, analytics, and all kinds of yeck. I tried to keep it minimal but that’s hard for me. I constantly had to pull back to the core and question my focus.

So thanks for reading. Please check out Speakway if you see it on the App Store. Tell me what YOU’D like to be doing while driving, that you can’t do right now. I plan on posting more about the market research I did and some of the experiments I hope to do for getting the word out to people who might want to use Speakway.

I will leave you with this screenshot.. UI is not my strong suit ;)