FEM inc Technology and How it Converts Viewers | Rachel Payne & Natasha Mohanty

How do you make technology out of this? Yeah, that's a hard question. And we're actually continually trying to answer that question because it isn't just gender that we're looking at. We're looking at diversity. We're looking at creating empowering experiences that can resonate with a wide variety of people of all ages and races and cultures and locales. So, you know, when you think about how do you build technology that can solve for impact That might be one of the hardest things to do, right? And I think probably you're all maybe grappling with some element of this. One of the things I found in the work that I've done in thinking about social change and large-scale systemic issues is that sometimes it's not neat and clean. There are cases where it can align perfectly with what you build, and then there are cases where there's some complementarity But you have to be able to allow yourself some flexibility to explore. Because when you're talking about change, large-scale change, there's a lot of forces at play. And there's a lot of ways in which you need to measure to calibrate your performance. So when we started out with our technology, we got down to the basic core of we want to create empowering experiences through media and entertainment. And how might we do that? Well, we can serve up these experiences that are really driven by that deeper motivation of the user, reflect that user, who they are, and whatever that means, and leads to deeper engagement because there's something that's resonating with them emotionally. So our first product in the market is a video recommendations platform that any website can embed that pulls from their library and can supplement with other video content to create a playlist that's customized for each partner and personalized for each user. And then as they interact with it, we learn what they like, and then it becomes more personalized over time. And in the background, continuously, we're thinking about all these other things. So it's in the algorithms. It's in every aspect of how we're thinking about it. But the user doesn't have to make a choice. Do I want empowering content right now, or do I just want to watch some reality TV? They want all of it. We're all multidimensional human beings. You want some things sometimes, you want other things other times. And so we're just trying to tap into what they need in that moment from wherever their starting point is. Did you want to talk about that at all?

Yeah, I mean, I think— I feel like if you haven't, you must have seen recommendation systems before. When you watch Netflix, when you watch whatever, YouTube, they give you recommendations. So frequently we get asked from a technology standpoint, how are you different? And I think it is really the core feature space that we look at that is fundamentally different. And the thing that does excite me is, you know, when Rachel was talking about, you know, Harvard did all this study and they understood long-term impact, one of the things I am actually surprised now that I know that research is that very little of that is actually done in the digital space. So I think what we're really trying to bring in is how can you do all that on websites, on apps versus, you know, bringing people together into a room and doing things. So I think one big way we're doing that is if you represent content in the feature space that really is human-centered, then it's much easier to understand the impact, really deliver on quality there, and fundamentally engage the user more. So I think on the surface it looks like other recommendation systems. In the background, it's really the kinds of data we're collecting is pretty unique because of the feature space we are representing content on.

And it's all contextualized. So you would imagine that you're reading an article and then this would appear at the bottom of the article, on the right rail, or in the middle of the article. It's wherever they put it. It's very adaptable. But the idea is that we are— there's a lot going on in the background that is a little bit different from the way that other recommendation systems optimize. So they optimize for clicks. And sometimes, by the way, those thumbnails represent women in really negative ways too. by the way. I'm just gonna say that. So what we think about are what are the ways that we can create an experience that takes the work out of discovery, appeals to a deeper driver behind why someone's watching or reading, and at the end of the day is an empowering experience. So we're still pretty new. So yes?

Oh, sorry to jump in. Also doing video recommendations based on text?

Oh yeah, there's a lot of— yes, yes, absolutely. But we're not doing— We are doing text-to-video recommendations, but we're only working in the space of video. Yes?

So I was wondering, have you checked how much percent of the readers reading the article will actually click down into the video and watch it?

Yeah, so actually I have that data.

Oh, here.

So on average, we have about a 7% click-through rate. Through rate and a 6% view rate. That drop-off's generally due to latency and other factors. But we instantly, pretty instantly, doubled time on site. And what we found, which was really remarkable, is that 25% of people who watch a video through our platform come back in less than 7 days and watch more videos. And we also have a plug-in that comes in after the video completes. It's a post-roll related video plug-in. And with that plug-in, we found that For every person that finishes the video, they watch an additional 4 videos in the same session. So it's actually a pretty immediate spike.

I think your question is pretty important, because that's actually the challenge that publishers are facing themselves, which is how do they convert their readers into viewers? Because video is actually way more monetizable than text content. So that's kind of the problem they are grappling with, too. One of the reasons they actually like us is we have—

Higher CPMs.

Yeah, I mean, we help educate readers that videos exist. Because frequently, people have been reading text on the internet for so long, they forget that you can also get videos too. And one of the things we've been doing a really good job is, because they are contextualized and interesting, once you learn that the video is there and you like it, you're more likely to remember it the next time you come.

Yeah, the behavior changes.

And so there's almost like an exponential factor happening as we're just educating people, like, hey, there are cool videos here. They're relevant. you might want to watch it. And that's actually changed behavior on the site itself, which I thought was a pretty powerful thing. Because to be honest, one of my biggest fears when I started this was, what if no one clicks on video? But that's no longer the case, which is great. Mobile video is growing like crazy.

We found that a lot of our partners completely got rid of their video galleries, actually, after using us. Because if you think about a video gallery, you have your nav bar. Someone clicks on videos. And then you go into this gallery, and it's like, Thousands of videos that you'll never have time to find, and you don't even know what you're looking for, so how do you browse? And by the way, they're all chaptered and serialized, and it's just a huge amount of work. So they just abandoned the entire video gallery, and now they're just using our platform to showcase exactly the video that the person would want in that moment. This is an example of a branded implementation where Nike wanted to get a message of empowerment around, you know, just go out there and go play ball. And it's a story It's actually a really amazing video about female athletes and what they had overcome to become a professional athlete. And it's a story around— basically, she's saying, I just wanna play ball. And so it's a way of speaking to the path and how challenging it often is. But this is— anyway, it's a branded experience that also pulls in content from YouTube, from the source, and from the brand, all in one, seamlessly unified. I think that's a great point.