Issue No.3

I've never really liked AI

Here's how Sona changed my mind

When I was younger, I thought having a computer I could talk to would be cool, like the Enterprise computer in Star Trek or KITT from Knight Rider.

You'd think, then, that I'm the sort of person who would be really into AI.

But I'm not. I have a genuine and long-held dislike of AI.

To me, even the best generative AI writing sounds stilted and forced.

I find AI videos to be mostly soulless, interesting only because they remind me of other things made by actual human beings (that the AI is more or less ripping off).

I think most generative AI "illustrations" look like little more than an improvement on Snapchat filters from a few years ago. And Snapchat at least has the decency to make partnership agreements with the brands they base their filters on.

I'm not shy about this. I talk about it all the time. So when I heard OpenPhone was launching an AI agent to help our customers answer calls, I was open to it, but a little skeptical. I thought it might be just a "we have AI, too!" add-on to our product.

If it was, that would have been ok. I don't have to be a big fan of all our launches. Not everything in the world is for me. Still, I wanted to give it a fair shot — technically, I had to give it a fair shot since I work here — and I'm glad I did.

As part of our big Sona launch, I made an announcement video starring our CEO and co-founder Mahyar. It's a fun video. If you haven't seen it, you can watch it below:

In the video, Mahyar stands on a loud street and talks to Sona to request an appointment for his dog Bella at a fictional San Francisco dog grooming company called Muttropolitan Groomers.

The conversation was real and recorded live. To make this work, I had to create the Muttropolitan business from scratch and train Sona to answer the kinds of questions a real dog groomer might get.

Despite having zero experience with Sona and a general aversion to AI, I had a great time setting it all up.

Such a great time in fact, that Sona completely won me over. What really impressed me was how quickly and easily I was able to get Sona to provide the sort of answers I wanted, in a tone and style that I felt matched our own brand, as well as the branding of my punnily named dog groomer.

I am no longer skeptical of our plans and had so much fun that I've begun questioning my larger dislike of all AI. I even downloaded the ChatGPT app on my phone so I can play around with it more. Maybe the Star Trek future I always wanted is closer to reality than I thought.

If you think Sona might also win you over, Justine and Kat can help you get started with your own version of Sona. Continue reading to see how.

Justine and Kat explain Sona

For this month's edition of Justine and Kat Explain, we thought we'd show instead of tell.

A few weeks ago, Kat, alongside our co-founder Daryna (who has her own section of this newsletter below) and our Senior Product Manager, Andrew, hosted a live event that details why we built Sona and how it works. We recorded it and you can watch it right here:

If a 40 minute demo sounds like a little too much at the moment, Justine made a 2 ½ minute video walking through how to set up your own version of Sona. Here it is:

Founders Corner

Becoming a multi-product company

These last couple of weeks have been a big deal for us. As you know by now, after seven years of building and refining our business phone system, we're officially becoming a two-product company with the launch of Sona, our voice AI agent.

I'm not here to write about Sona itself, Richard, Justine, and Kat already talked about it plenty up above. Instead I want to share my thoughts on the challenges launching a second product. (Note that this is an abridged version of a longer article I wrote for my own Founder-to-Founder newsletter).

The inevitable second product question

As a founder of a SaaS company, there's a point when you start wondering if you should introduce a second product.

Jason Lemkin talked about this many times, including recently in this LinkedIn post: Every founder of a SaaS company will eventually consider adding a second product. Some do it too early, and it becomes a distraction. Others do it too late, and the growth stalls.

Personally, I feel like we're doing it at the right time. If we did this earlier, we probably wouldn't have the internal infrastructure to do it right. And if we waited longer, it would feel like we missed the boat.

I guess only time will tell, but I think we've hit that sweet spot.

As a communication platform, we've had the opportunity to introduce AI in a meaningful way. Seeing our customers being hungry for AI inside OpenPhone has certainly sped up our timeline.

I've been thinking about all the things we need to figure out as we navigate this transition. Here's a list of the biggest ones:

Pricing in the world of AI

We've always aimed to offer simple and clear pricing, with plans that come with US & Canadian calls & messages included. AI comes with inherently different economics that lend a lot better with a usage-based model.

Our go-to-market strategy is flipping

For years, our marketing efforts have been focused on being discovered as a "business phone system." That's been our identity. But now it's possible new customers will discover us through the route of voice AI first, then realize they need the phone system too.

Support gets more complex

We've been self-serve for years, but this is a new product with AI in the call flow.

The concept of training your AI agent is fairly new to a lot of businesses we serve. So we're investing in content creation and offering more hands-on support.

Partnerships become more valuable

As OpenPhone gets more configurable (API, Sona, etc), there's a lot more value an agency or solutions partner can offer. Sona adds such an exciting dimension to what they can do on behalf of their clients. Just like we saw automation agencies create Zapier and Make workflows for businesses, we'll see them creating workflows inside OpenPhone.

Becoming a multi-product company is new territory for us. We've gotten advice, we've planned carefully, but there's still so much we're figuring out as we go.

Every SaaS founder (and many other founders) will face this decision eventually, and I hope sharing our process helps when it's your turn to do something similar. Read my entire post about this to learn even more.