Social media is not just a platform for sharing photos and videos, but also a treasure trove of information about the latest trends and consumer behavior.

Andi Govindia, co-founder and CEO of Riviter, recognized this early on and has been using cutting-edge technology to turn social media visuals into predictive insights. In this episode, we delve into Riviter's approach and explore Andi's journey from creating a visual scrapbook during an internship to the fouding of her company that uses computer vision and predictive algorithms to provide brands with social media insights. Join us in learning how brands can effectively communicate with their customers and bridge the gap between media execution and social insights.

You can reach out to Andi on LinkedIn.

Many thanks to Andi for being our guest. Thanks also to our producer, Natalie Pusch; and our editor, James Carlisle.

Hello, everybody, it’s Lenny Murphy. Welcome to another edition of the GreenBook Podcast. Thank you for taking the time to share some of your day with us as we share cool stuff with you. And on that note, this is going to be very cool. So, my guest today is Andi Govindia, the co-founder and CEO of Riviter. Andi, welcome.

Thank you. Thanks so much for having me.

It’s great to have you. So, you and I connected at the end of last year because I saw Riviter’s technology. I forget exactly how that happened, but I’m glad that he did. And I thought, “Okay, this is really cool.” Because it’s such a different approach, an evolutionary approach, to social media analytics in the kind of a broad category. But we’ll get into all that. So, why don’t we first give our audience a little bit of background on you and on Riviter, and then we’ll get into what got me excited about what you’re doing.

Of course, yes. I’m Andi. I co-founded Riviter about eight years ago with my co-founders, Kris Woodbeck and Rachel Chen. And what we focus on is giving brands predictive insight using the visuals that consumers and tastemakers share on social media. So, it’s as simple as that. We look at posts and social media, predict what trends are going to be big, how consumers preferences are changing, and in particular, how they’re demonstrating this through their photos.

Okay.j although I don’t think you did it justice, Andi.

So, prior to Riviter, I’d spent about seven years in marketing. And so, I think I had always had this desire to really see how my customers and consumers were using products in the real setting. So, I think that that itch was always there to scratch. But actually looking back, the formative experience was in my very first college internship.

So, long came Instagram, right?

You know, I got really lucky because I was introduced to my co-founders just with this idea. So, I have no technological background, I can just have ideas and talk about them. And it just so happened that Kris and Rachel had been working in computer vision in a way that, at the time, made them look kind of crazy, [laugh] right? So, the whole field was going into kind of the Googles of the world and the of the world. And it was about can computer vision label the world.

Yeah. So, that brings up an interesting point. And I think especially the obvious use case, for me for the basic idea of technology was logos, right? Something that it’s really easy to identify, you know, I’m holding a Coke can, right, or whatever. But it’s interesting that you focused on beauty, which I wouldn’t have intuitively thought was an easy application of the technology, right, and as an early trend-spotting solution versus just a counting solution. “Oh, yes. Look, we see, you know, a hundred people were, you know, showing Revlon,” or [laugh] whatever the case may be.

I think this is where insights really comes into play, right? So, we originally started with this incredible visual search technology. And as we were talking to brands and retailers more, we were talking to them about, you know, shop with a picture and kind of the most obvious application for the technology. And they said, “Well, what happens if a consumer comes to my website and is looking for something that I don’t have or that I’m out of stock in? Doesn’t that reveal more about the fact that I don’t understand my consumer, about the fact that they wanted something that they came to me expecting to have for them and I didn’t have that?”

So, but you went to step further than that because you also started identifying, effectively, communities, bringing communities within the solution. And can tell us a little bit about that intersection? And again, without making it too complicated, what impressed me about Riviter was it wasn’t just a counting tool. You combined it with solutions like a predictive algorithm, you know, with things that would be really helpful in early-stage ideation and product development as well as kind of tracking and identifying these populations and you know, what they look like. So, tell us a little bit about that. Because it’s just such a well-baked solution.

Thank you for saying that. Yeah, it has been hard-wrought and in the product of people’s requests. And that’s the beauty of social, right, is that it’s just another means for which people to organize themselves like they want to in real life. And so, just as in real life, people organize themselves into communities on social by virtue of the conversations that they participate in or the people that they tag or the people that they follow. And so, we’re actually able to see that network graph in social media and we can see, oh, when you Hailey Bieber dyed her hair turquoise, a month later, Billy Eilish, who’s in her network, dyed her hair turquoise. [laugh]. We can actually see that thing that binds them and follow this.

Very cool. So, what other categories? Obviously, there’s a training component to it to start looking at other things, but where else have you found traction outside of health? I’m sorry, outside of beauty. Sorry.

We’ve been kind of surprised that visual categories were really intuitive to us, and beauty and fashion were really clear. But we’ve gotten requests from travel, people are booking their travel destinations based on where they want photos [laugh] and what they see in photos. We’ve gotten requests, even from communication. How are people using their phones? And in what settings? And can we actually see whether they’re with their family or with friends?

Let’s go back to that brand community idea. So obviously, there’s the observational component and the predictive component. Is there an approach within Riviter to allow the brands to engage with those communities as well?

Yes and no. I think, you know, consumers really do want to see that brands are responding to the things that they’re saying and that’s why they’re tagging them and that’s why they’re sharing these conversations with them. And so, I think there’s a big opportunity for brands to say, “Hey, we’ve seen you and we’ve heard you, and this is how we’re responding to something that you wished existed in our product or something that you shared with us.” And having that be part of the dialog, and Riviter can serve to connect those, right, to say, here are the actual influencers and people who have shared these comments. And here’s how many of them they are, and here’s the circles that they operate. Here’s the channels that you can reach them through. Here’s the people that they follow or the publications that they follow. Here’s how to connect with them. But ultimately, that relationship should be on the brand and it should be them to directly make that connection with their consumers in their audiences.

[unintelligible 00:13:15]. Any plans to ever be that bridge directly? To help enable that, or let that be—

So, one of the things that we really like is learning from our partners and from our customers. So recently, we’ve been talking to a lot more of the agencies and talent managers and groups that work with creative talent, like influencers, and they talk about this idea of, can we take people who don’t necessarily have a huge, huge following, but they’re influential in the communities that they operate in, and make a really good match between them and the brands. And so, being equipped with here’s quantitative data that shows how good that matches and here’s information to show how influential they are in beauty versus fashion versus categories that previously, I think, we kind of painted with a broad brush and assigned creators to all categories, and there really is an aspect of fit to that. And so, being able to make that pairing in a way that is going to generate a good return for a brand is something that we’re working on in partnership with exactly those people.

So, it’s one of the things that I found really interesting about Riviter is you don’t think of yourself as it insights company. You play within insights, but obviously, based on discussion we just had, you see a broader role for what—I assume you do; I shouldn’t put words in your mouth; you tell me if I’m right with this—a broader role to help enable data-driven connections—if I can kind of summarize my take on what you’re doing—which is an area that the traditional market research industry has been, oh no, that that’s a third rail. [I don’t know if 00:14:53] we want to touch that, right? But increasingly, I think it is a needed component for companies in our space to be able to drive that more effectively. So first, am I right with that, that read?

Yes. Yeah, it’s happened really organically, I think—and this is probably something that’s true of all market researchers—is that you deliver an insight and then you want to know that the company has taken it on and that they’ve done something with it. And most of the time, you don’t get to see that. And so, I think for us, it was, well, what can we do to bridge that gap? How tactical can we make these recommendations and these insights, and started realizing that a big part of the gap is actually being able to reach these creators and reach these communities in that organic way, and the way that we’re recommending.

All right, so let’s tack into that because, you know, there was AI, and then here came November [laugh] of last year, and all of a sudden there was oh, crap AI.

We’re so excited about this because we love what it’s done for the conversation. Someone gave me this analogy which I thought was really great. They said, “Well, you don’t understand—can you explain to me how alternating current works?” [laugh]. “Can you talk to me about how electricity is made?” No. Most people don’t just know. They flip the switch and the light is on.

So, it’s interesting. I was having a conversation yesterday with a friend in the industry. And they brought up a topic that then came up again this morning with another conversation that I had not thought about. And I was aware of that technology being used already, from a survey standpoint to commit fraud, right? Fraudulent responses.

Yeah [laugh]. Yeah, it well, sorry to say, it’s already been happening for a while, right? So we, almost from the beginning of Riviter, we’ve had to build really sophisticated ways of detecting whether people on Instagram were real people, whether they were purchased accounts that, you know, kind of operate as click farms, whether they were bot-driven, whether even they were real people, but their likes had been purchased and giving inauthentic [measures 00:21:47]. And so, I think it is an obligation of the providers of these technologies, A, to think about how do we create those safeguards and put them in our technology. And then B, how are we transparent about what’s going into the training?

What an interesting world we’re living in. And I think particularly in the arena that you are in because I suspect this is true, is that in social media, that is the most likely place to see the application of these technologies in maybe a not-so-virtuous way, [laugh] right? So, you’ve got your work cut out for you.

Well, you know, we focus on making sure that we’re talking to real people and that those real people are given a visual voice in front of brands. And ultimately, I think that transcends what the technology can do. It’s really just meant to enable that connection and enable it at scale. And the rest smooths the edges and it makes it a little bit faster.

Alright, so if I think back to the advent of social media analytics, right, in its broadest sense, there were so many predictions back in 2010, that, oh, this is going to replace, you know, traditional market research. You won’t ever need to do a focus group again. You’ll never need to do a survey again. And that obviously has not occurred. I think we’ve seen the evolution of an augmentation of another tool that delivers complementary insights from a very different angle.

Yeah. That’s a great observation and it’s very true. I think over the last few years, we’ve gotten a lot better at meeting our customers where they are. And so, if they want to keep doing their surveys, they should keep doing their surveys. If they want to do traditional ethnography, they should keep doing that, and there’s a really good time and place for that.

I agree. I hope that that’s where we get to. I think we are. You know, from the [unintelligible 00:27:41] conversations that I have, I think they want an answer to the question—the business question—and the most effective way to do that, right? I think we, on the supplier side, we get really caught up in, you know, the methodological rigor and, you know, on our favorite tools, et cetera, et cetera.

I hope, and what we’re building toward, is kind of the same way I described our relationship to ChatGPT as our best intern [laugh]. Something that we can go to at any moment and ask something of it and get that quick response and then be able to take that back and apply our own knowledge and our own context to it, is that I hope that brands and insights teams will think, “I just want to a quick answer,” or, “I just want to see what’s going on with my tribe or my segment,” ideally, in the future, they are community-oriented around their brand's community and not just the general population because that’s just as important. But to say, “I have this question.” Let me just ask Riviter really quick and then I’ll take that back to my team and provide this. And I know it’s quantitatively sound.

Very cool. All right, I’m going to put you on the spot for a minute, but in a good way; you get to pitch a little bit. So, one of our listeners is a fast food chain, right, if you haven’t worked with him before. How long does it take for a new segment to get up to speed? How do you onboard new segments that you haven’t worked with?

[laugh]. It’s something I’m actually excited about is we’re releasing this year, out-of-the-box segments. And so, this includes things like fast food lovers, this includes travel aficionados, and tech aficionados, and luxury-oriented consumers. And so, in theory, it can be just as simple as, I’m in the sector and I want to look at Group A. What can you tell me about them? And then we’ll go back to them.

All right. I want to be conscious of time for you and our listeners, so will head towards wrapping up. But is there a trend that you identified—without breaking the client confidentiality—right, but you identified a trend that just surprised you? You were like, “Oh, wow. Okay, that’s really interesting.” That was just kind of revelatory for you, personally?

A couple. I’ll share the one that we kind of share the most because it’s such an example of a beautiful visual trend. So, this is the unicorn frappuccino was released, I want to say in April 2017. Did very well. It became very socially viral, did very well for Starbucks.

Very cool. Unicorn toast. Alright. That sounds more interested in avocado toast. I know my kids, “Dad, try avocado toast.” Like, “No. That just looks gross.”

I think the thing that I’m most excited about really is how mainstream the conversation around AI has become and how now it’s an expectation that we know how we want to use it, that we’re going to be thinking about how to use it in our day-to-day lives. And so, I hope that that brings people to us to say, “Okay. Now, I do want AI to help me and my insights, and now I do want that connection via social.” And that’s the other side is that we’re starting to think a lot more around communities and the communities around brands and how brands think about their audiences in that format. And so, bringing those two together, I think now’s the time. People are open to both things. They’re talking about both about those things and hopefully, we get to work with them.

Very cool. So, outside of work, what—if you’re like me, there really is—you hear that and go what?

[laugh]. I think I’ve been lucky enough that I’ve been an entrepreneur for eight-plus years, so there is actually a time in that area where the startup stops becoming your life, which is nice [laugh] and you start thinking about other things again. So, we just moved to New York a year ago and so we’re still exploring the city and I think that’s what I’m most excited about is there all these micro-cultures, all these little neighborhoods, and, you know, you can take the insights person out of the job, but [laugh] they’ll go out and observe people. So, I am excited about that. And I think summer is around the corner. That be really great to see the city in that way.

Summer in New York City. I’ll tell you my worst—some of my worst memories are July in New York. And this comes from somebody raised in the south, so I got to warn you. They’re like, “Oh, my God, it’s so humid in Atlanta.” Like, “Dude, have you walked down Madison Avenue in July?” So, if you haven’t experienced that yet, Andi, I got to warn you.

Now, I know why people go other places in the summer. Thanks for the warning.

[laugh]. Because it’s an oven, you have these buildings, and they’re reflecting the sun down on the black asphalt and it’s just hot. So, anyway. It’s still a pretty cool place. All right, anything that you would like to share that I didn’t ask?

Gosh, I mean, I think if we had a whole other hour, I would be asking you [laugh]. There’s so much that you see, and you know, questions that are maybe staying unanswered in the insight space that—are there areas where we can think about applying our technology? Are they’re, kind of, gaps that we can fill? So, let’s have a long conversation after this about this things [laugh].

All right. Well, you have my email, so I look forward to it. That’s great. Thank you so much, truly. Again, I can’t remember exactly how we connected. Do you remember exactly how we connected?

I think it was through an insights veteran who had also been on the disruptive technology side and said, “You know, you got to talk to Lenny because he’ll help contextualize [laugh] for everybody else.”

Okay. Well, yeah, it’s ancient history now, right? I mean, but I’m glad that we did. So. And I hope, audience, you’ll take from this that, you know, there are so many incredible existing applications of technology like what Riviter’s doing around imagery really should be part of the toolkit within the insights organization.

Thank you. And thanks for being an advocate for startups. And it was a pleasure.