Ryan Restivo wants your journalism to be found.
He built an AI-powered tool to demystify search engine optimization. The product, YESEO, is a free Slack bot that reduces the time it takes to write SEO-friendly headlines with relevant keywords.
“How do we lower the barrier to entry for people to get the information they want at the time they need it? We tend to think about that in terms of our audience, our readers, but journalists also have to worry about how they are writing headlines and content and whether that is optimized for a search engine,” Restivo said. “I want to make sure we’re giving our newsrooms as strong a treatment as we give our users.”
Today, YESEO is available in eight language models: English, French, Spanish, Portuguese, German, Finnish, Norwegian, and Swedish.
We asked Restivo to share his development process and advice for bringing big ideas to life.
Can you briefly walk us through how you came up with the concept of YESEO to its launch in March?
Restivo: The starting idea came to me when an editor asked why a story did not rank high on Google. Once we analyzed it, I thought, what if we could use a tool in Slack to perform analysis on any story and for as many newsrooms that might want to use it?
In my experience I have connected with building tools to help journalists do their jobs better. I have been lucky to be supported through a Reynolds Journalism Institute fellowship to deliver on their priorities of creating free, accessible, practical innovation to help newsrooms with their SEO best practices.
The goal of YESEO is to reduce the time to come up with that relevant headline, the right keywords to use, and give you the right information at the right time to help you get your work seen and read. It is looking to use natural language processing and other AI tools to help give journalists more data about their story, and use the power of computing to help journalists get their stories to the people who want them the most.
YESEO wants to be your story’s new best friend. Helping you unlock actionable insights and data for everything you put into it. We’re at over 170 workspace installs and counting!
What were some of the unexpected challenges you faced while building YESEO?
Restivo: YESEO is the first time I’ve built a product that can be installed by multiple workspaces, which brought its own unique set of challenges.
The eight-month fellowship was constant learning. First I got great advice to “fall in love with your problem, not your solution”, so I really dug into what people in the industry said their biggest pain points were. Then it was how do I create something that will work and test it constantly either in interviews or with test newsrooms and then distributing it has been a great challenge.
Building YESEO meant I had to learn new frameworks, databases, and data modeling just to store the data from the app. Then I had to figure out how to create metrics, so I can track user behavior and on top of that build the entire system so that a user can get their story analyzed in under two minutes and receive buttons with insights they can act on.
You mentioned in your introduction post that you plan to experiment with additional features. How do you prioritize what integrations to include?
Restivo: I was lucky to be a part of the News Product Alliance mentorship program that ran earlier this year and my mentor Aaron Williams gave me great advice as my project launched. He told me to not worry about adding any features right after launch and just watch how users interact with YESEO and they will show you what problems need to be solved. In that first month, users showed me issues I would have never seen and as a result I put together tutorials and introduced ways to remove friction for users when they ran into problems.
The app released in March with many of the features users see now, but the core code that runs the app has changed drastically since that time. As part of the Generative AI in the Newsroom project, I built a brand new system that I’m still tweaking on how the app can message its users directly with updates, tutorials, and get instant feedback. In that project I tested out headlines amongst GPT models and that test framework has driven me to continue to test new changes and get feedback on how helpful they are to YESEO’s users.
I have about five to seven ideas for future features that could make YESEO more valuable, but I am still iterating and most of all listening to the users of YESEO. I have been working hard to learn their stories, how they came across it, and then what they think useful changes would be. I have a notebook of feedback from users and I have tried to cross off as many improvements as I can make over time.
YESEO is a free tool for newsrooms, but I know that if I ever want to have a future as a paid service I will need to create more value for newsrooms. I have been actively exploring many different ways YESEO can help newsrooms beyond what it does now.
What are your top three tips for journalists to turn a big product idea into a reality?
Restivo: My first one would be to remember what it is about you that inspires your idea. Everyone has big ideas, but what makes them unique is who you are and how you think about your problem. I don’t have a “capital J Journalism” degree, but I know that my background and experiences have helped me when I think through problems that I want to solve to help journalists. I always try to remember that, when some people might see it as a weakness, I see it as a strength that my experience producing digital media, or testing, or working on code helps me see things from a different angle. When I try to propose solutions, they come from another angle, but they always have that common throughline on helping newsrooms.
Second, would be to remember to relentlessly focus on who the product idea would benefit. Think about what they want and their pain points and figure out ways to deliver to them what they need. This will probably mean you will be constantly adjusting what your idea is to make it fit the need, and that’s a good thing. Ideas I had over a year ago for YESEO were discarded throughout the process, because they didn’t fit with what I knew the users would value the most.
Last would be to always talk to people and look for advice. There are plenty of people in this industry and many of them are smarter than me, so the more often I chatted with someone smarter than me, the more I took away a new insight. Once you’ve talked to enough people, then sometimes you can activate an idea really fast, like when Richelle Gordon asked me could you break down what keywords from the story were in the headline, a lightbulb went off and that night I was able to build what is now a core part of the YESEO experience: using that natural language processing to instantly show a user if their headline and story strongly or loosely relate to each other.