Talking with Yala, the coolest & mightiest social media bot on Slack.
At BOTCOPY we LOVE writing for the world’s coolest bots. We wouldn’t trade it for anything. Even a puppy in a teacup.
Why? Because when we’re watching results...and optimizing blurbs and logic flows...and suddenly retention and conversions spike, it’s dang euphoric.
Recently we had the chance to help out with copy for one of the most compelling and unique brands in the AI world: Yala.
We were stoked to work with Yala.
Yala enjoyed a well-earned stint as the #1 bot on Slack earlier in 2017. She is a spunky-smart social media bot that uses machine learning to post your stuff to Facebook, Twitter and LinkedIn at that perfect moment when the most of your followers are watching.
Yala solves a problem we all identify with: the vague feeling that a lot of what we post on social channels isn’t getting seen. Her scanning ability and meticulous timing of your posts to hit the most eyeballs makes all kinds of sense.
Aside from being smart and powerful, Yala has a fun, friendly, quirky brand voice. So when we helped write her tutorial, we made sure to bring that sense of fun into her language.
For example, an ordinary bot might say: “Our algorithm helps businesses post to social media accounts quickly and conveniently.”
We had Yala say things like: “My space-age Yal-gorithm will have you dancin’ the social media samba in no time.”
Distinct personality boosts bot performance.
It’s a subtle shift, but from what we’ve seen, it’s a vital one. Playful language used sparingly, combined with elegant, useful software, is the magic combo that seems to help bots soar to the top of the charts.
Yala’s a true pioneer that’s crushing it in the bot sector, so we couldn’t pass up the chance to ask them a few questions. Yoav @ Yala was kind enough to give us a glimpse into the mindset of a conversational AI company walking the walk and making a difference in this brave new category. Enjoy!
BC: What’s one thing you folks at Yala believe firmly about bots, or the conversational UI trend, that most of the rest of the industry disagrees with?
Yoav: There’s a common conception that bots are to apps what apps were to websites and websites were to desktop software: an evolution in software packaging and interaction. The way I see it now, bots aren’t quite a replacement for any other kind of software. Instead, they’re an extension of existing software interfaces. Take Alexa for example - her different “skills” are all more accessible versions of things you could’ve done on your desktop on mobile phone, but she can’t complete the really complex tasks. She’s more of an extension of your web browser than a replacement for it. I believe successful productivity bots will evolve to feature websites or apps as the “under-the-hood” interface, while having bots facilitate communication with that interface.
BC: What are some of your favorite bots (besides Yala of course) and why?
Yoav: Some bots continually inspire me to discover edge cases and craft appropriate responses for them. Meekan has best-in-class 404s - messages it uses when it doesn’t know what I’m talking about - and Poncho has a surprise waiting behind every corner, keeping me on my toes.
Howdy, Growbot and Todo are all just fantastic examples of how to build apps that actually make sense conversationally, and find product-market fit for them. They also know how to use all the little quirks of Slack as a chat platform to create a better user experience.
BC: Trailblazing seems like tough work! What is the biggest mistake the Yala team made early on, and have since learned from, with regard to building an awesome bot?
Yoav: We’ve spent too much time building workaround solutions for stuff that should be solved at the platform level. For example. before Slack introduced Message Buttons we built a complicated way to interact with Yala using emojis. It was colorful but unintuitive as heck. 😊
BC: What are some cool things you learned from reviewing user convos that surprised the heck out of you?
Yoav: Watching people interacting with Yala is BY FAR the best way to improve our product - better than user surveys, Google Sprints, anything! I can use these conversation logs to tell what are the most requested features, determine which points in our funnel are leaking, and find good spots to add a joke. To answer your questions: one thing that keeps surprising me is users’ expectations from the bot. They either expect it to know nothing at all, and are surprised by every single answer, or absolutely everything, and are annoyed when she can’t tell them Apple’s stock price. Go figure.
BC: You’ve blogged about the virtues of being inefficient as a company. In one sentence, how does being inefficient help you make better bots?
Yoav: At this point in time, bots themselves are inefficient. The tech isn’t there yet - discovery is difficult and GUI elements are almost nonexistent. Our willingness to be inefficient has helped us come up with solutions we wouldn’t have found otherwise.
BC: Do you have any favorite tools you use for programming your conversations?
Yoav: We use a tool called Walkie to design our Slack conversations that I find very useful. It mocks up the Slack environment so you can tell exactly how your messages might look in Slack. That’s all we use - mostly, we edit conversations using text editors like Atom or Sublime Text.
BC: Some people might be worried that AI will take over the world. If Yala took over the world, what kind of world would we find ourselves living in?
Yoav: That’d be nifty! I’m imagining a retro-futuristic middle of the road diner, and people/robots with blue skin and purple hair zipping around on flying motorcycles. Oh, and I would wake up every morning at the perfect time.
Get in touch to learn about making a better bot or making a bot better. And definitely check out Yala’s free trial.