Every fall, leading startups, companies, and organizations in NYC pose business challenges to Cornell Tech. In Product Studio, teams develop and present new products, services, and strategies that respond to those challenges.
My team finished responding to The New York Times (Community) challenge.
The challenge is as follows:
How might we create a safe place for people of all backgrounds to discuss important local, national, and global issues?
Check out our prototype at
Video of Prototype
- GitHub Hosting
- Best Final Sprint Presentation in Grizzly Room
- Recognized as having one of the best team presentations during the final sprint in the Product Studio class at Cornell Tech out of 10+ teams inside the Grizzly presentation room.
- Artful Artifacts
- Recognized as a great team example that have produced meaningful artifacts that not only showcase an interesting way of representing their solution, but also act as useful milestones that can be referenced to in the future of the product development process.
|Delia Casa||Frances Coronel||Chumeng Xu|
Hi, we’re the New York Times (Community) Team, Delia, Chumeng, and Frances.
Our challenge was How might we create a safe place for people of all backgrounds to discuss important local, national, and global issues?
The people who comment on the New York Times platform are fairly homogeneous: they tend to be older, they tend to be white, and they’re more likely to be men. As a result, Times readers are missing out on different perspectives, and the Times is missing out on an opportunity to grow its commenter base and build a stronger relationship with its audience.
So we thought, how can the comments section do a better job of inviting people of all backgrounds in? Through our user research, we found that people are more motivated to comment when they feel they have specific knowledge to share because of their experiences—and people get more value out of reading comments that are informative.
Coded Out Solution
Let’s walk you through our coded out solution. Meet Ana. She doesn’t subscribe to the Times, and she has never commented.
One day, a friend shares an article on Facebook. She clicks through to read it, and she sees a pop-up. When she clicks the pop-up, she’s shown a question that resonates with her. Based on Ana’s reading history on the Times, our system has made a guess that she is Latina, and made the default question and comment view one that pertains to her background. She can click through to see the other questions, and she can switch tabs to view comments from people of other backgrounds, including children of immigrants and language experts.
She stays on the site longer than usual, reading through the responses and looking at the data visualization about the comments. Thanks to the question prompt, she can see that she has valuable information to contribute. So she composes and posts her comment. This makes the commenter base more diverse, and we can keep Ana coming back to read and comment by sending her email digests about other relevant questions and comments.
For user validation, we performed A/B testing and found that 90% of our test subjects were more likely to read the comments with our Q&A platform, and 80% were more likely to actually post a comment themselves.
But where do these questions come from? The Q&A platform uses the IBM Alchemy API to automatically generate smart entities from the text of an article. Each of these entities then acts as a potential group of readers who can provide valuable insights for the article. The reporter can then compose questions targeting each of these reader groups.
And when we talked to a Times reporter, we got feedback that our platform could not only integrate well into her work-flow but also into the content management system that she uses to publish articles.
But the real value of Q&A, aside from the fact it diversifies the commenter base, comes from how we are creating a platform that takes the highest quality journalism and adds on the most informative user-generated content.
Since a commenter self-identifies as belonging to one or more groups, we are able to subtly collect reader background information which leads to smarter targeting.
In terms of product development, we focused more on fleshing out the front-end since we feel that’s where our product value lies and with the back end, we got as far as being able to extract smart entities from article text.
Our business objective is to convert occasional readers, like Ana, into engaged digital subscribers. We estimate that we can convert 30,000 non-subscribing readers into digital subscribers by keeping them coming back to the site directly, pushing them over their monthly limit of ten free articles.
Go To Market Strategy
We recommend taking the feature to market via pop-ups on the site, a feature explaining Q&A, social media, and email notifications.
We floated a lot of ideas before settling on Q&A, and we pivoted after doing user research with Times readers. So we definitely learned to know our users first. Also, we saw that active engagement is really important to the future of news, and traditional media outlets like the Times recognize this and are adapting.
Lastly, to hand off our product, we plan to provide all of our assets to the NYT team. Thank you.
- Recognized as one of the best presentations during Sprint 3
- presentation that showcased a narrative-complete product in a great way
- Recognized as a team for having artful artifacts during Sprint 3
- produced meaningful artifacts that not only showcase an interesting way of representing their solution, but also act as useful milestones that can be referenced to in the future of the product development process
- GitHub Repo | Assets
- GitHub Repo | Q&A Prototype
- About Product Studio