<img height="1" width="1" src="https://www.facebook.com/tr?id=901494157838465&amp;ev=PageView%20&amp;noscript=1">
Skip to content

Building Iconic Brands Through Predictive Social Intelligence

Stravito May 18, 2023


In this episode of the Consumer Insights Podcast, we speak with Andi Govindia, CEO and Co-founder of Riviter.

Strong brands build communities around their products; iconic brands build products for communities.
But how do you identify and understand those communities?

Social intelligence is a powerful tool to gain an understanding of your consumers that’s otherwise difficult to achieve through traditional research methods, and AI can help you make sense of it efficiently.

In this episode of The Consumer Insights Podcast, Thor is joined by Andi Govindia, CEO & Co-Founder of Riviter.

They cover: 

  • Why insights should be the foundation of every business and brand
  • What separates a strong brand from an iconic one
  • How technology can help to capture early signals of consumer attention
  • The role of social intelligence in your market research toolkit
  • Why it’s important to understand who you’re communicating with, not just broader macro trends
  • Opportunities and pitfalls of AI for insights
  • How insights democratization impacts the role of insights professionals
  • Key criteria to help you differentiate between AI providers
  • How pairing AI with human expertise can help to gain a deeper understanding of people

If you’re interested in learning more about how social intelligence and AI can help deepen your understanding of your consumers today, tomorrow, and 5 years from now, tune in to this episode of The Consumer Insights Podcast.

You can access all episodes of the Consumer Insights Podcast on AppleSpotifyGoogle, or use the RSS feed with your favorite player. Below, you'll find a lightly edited transcript of this episode.

Thor Olof Philogène:

Hello, everyone, and welcome to Consumer Insights Podcast. Today, I'm excited to have a brilliant Insights thought leader joining me for what I know will be a fascinating conversation.

I'm thrilled to introduce today's guest, Andi Govindia, the CEO and Co-Founder of Riviter, a company who uses social graphing, AI, and visual search-powered insights to show brands and retailers where to play and how to win. 

Since Riviter's founding in 2015, Andi has led Riviter to backing from numerous organizations, including South by Southwest, the National Science Foundation, and Pritzker Group Venture Partners. She has an MBA from the University of Chicago Booth School of Business, and she's on a mission to make cutting-edge innovation accessible to industries and audiences who might otherwise be overlooked by technology. Thank you so much for joining me, Andi.

Andi Govindia
- 00:01:24:

Thank you so much, Thor, for having me. 

- 00:01:26:

So, Andi, could you take a couple of minutes to tell us about yourself, your company, Riviter, and how you got to where you are today? How did it all begin?

- 00:01:35:

Yes, as you mentioned, we founded Riviter in 2015. I co-founded it with my co-founders, Chris Woodbeck, and Rachel Chen. And back when we started, we really just had this incredible visual search technology. So they had worked in computer vision for almost a decade before I met them. The technology could identify objects and images and actually describe them. So you could say the qualities of a product, the colors, the prints and patterns, the shapes, all about the product. And so immediately we thought, let's take this technology to brands and retailers and help people shop with a photo. That seemed pretty intuitive. But pretty soon, we realized when we were talking to brands and retailers that there was actually a bigger problem underneath. They said, “Okay, great, we can have this shiny feature on our website. But what happens when a consumer comes to our website and they look for something that we don't have, either because we never made it or because it's out of stock? Isn't that a bigger problem that this consumer has entrusted us to sell them, this thing that we don't have? Is there anything that we can do about that? How do we solve that problem? And is there actually anything in the photo that could tell us how we solve that problem?”And so we said, “We're probably going to need something with predictive power. We're probably going to need to see something earlier than the actual search on the interaction on the website. Is there another place where consumers are sharing photos of themselves and the things that they want in a really public forum?” And this was right around when Instagram started to really take off and so we started looking there and sure enough, people were sharing photos of products and the things that they want. And better yet, they were organizing into communities so we could actually see people and how they share these desires with one another.

So we built Riviter and found out that this really did have predictive power. Sometimes we can see things 24 months in advance of really becoming a big trend. And that's how Riviter was born, from insight itself.

The definition of an insight

Thor - 00:03:29:

Oh, that's such a fascinating story. And given when you sit in the world of consumer insights, how do you define an insight?

- 00:03:37:

Yeah, I appreciate this question as an entrepreneur because I think the two things are so inextricably linked, right? You're taking an observation and you're having to turn it into an action that you could take, a business that you can form, a problem that you can solve or an experiment that you can do. And so to me, an insight is something that takes an observation and turns it into something actionable. When you say, “Okay, we've talked to these brands and retailers, they've given us this observation, this data that there's a concern here. They've also entrusted us with it. So they think that there is something to do with images.” 

And the insight is exactly that is, let's now take this observation because we've been enlisted with this. Let's actually run that experiment and see if it's possible. And that's what we do as business builders. And I think that's what brands and companies do for their consumers as well.


On the importance of insights

- 00:04:29:

And building on that, from your perspective, why is it that these market insights, these market and consumer insights are so important? What is it that if you take kind of a broader perspective out from your entrepreneurial perspective, what do they, generally speaking, allow businesses to do. 

- 00:04:46:

Yeah, aside from being the foundation of our business, I think ideally it's the foundation of every business and every brand. We like to say that strong brands build communities around their products, but iconic brands build products for communities. And that is to say, there are people who love hiking in the mountains, there are people who love craft beer in the US. There are beauty lovers in the UK. And they expect brands to be able to see them and they expect to be able to share the things that they want and the things that they want out of brands. The process of insight is what builds that bridge to say, “Here's this group of consumers, we hear you, we can understand what you want and we understand how to fill those needs.”

- 00:05:25:

And if we talk about the work you're doing at Riviter, I mean, I think it's fascinating. And I also think that it's incredibly important for brands as the attention economy continues to evolve and become so competitive, especially on social media. Could you tell us a bit more about Riviter's secret sauce? Obviously, I'm not asking you to reveal it, but help us understand how that helps brands capture early signals of attention.

- 00:05:49:

I love that phrase, attention economy. That's not ours, that's yours. So we're going to borrow that. It really is about that. We built these tools to let us connect to one another and in doing so, we're able to reveal what we're passionate about and what we want to pay attention to. So through our signals of attention, we're sharing an indication of our current and future behavior. And so the three components of our technology really center around that. We said, “Fundamentally, people organize themselves”. So rather than just watching all the hashtags and all the language in the world, let's treat people as people. Let's look at groups in their communities and let's actually segment them. Whether it's our segments or whether we build lookalike segments based on our customers' existing personas, let's look at groups and identify people within them, how they shift over time, people who enter and exit the group, and then the tastemakers within those groups. So that's number one, is that we're actually looking at groups of people on social media. 

The second piece is, as I mentioned, our really powerful visual search technology that really allows us to see what are these things that show up in photos that aren't necessarily in the caption. And in fact, only one in five posts even has a caption, which is increasing even more with TikTok. Let's see what they're revealing in their photos and the context in which they're using these products and the context in which they live their lives and give them a visual voice to be able to bridge that gap. 

Finally, because we have so much data because we've been able to look at it over the course of many years, we do have a really powerful predictive capability. That is fundamentally what helps us to give our customers and brands an advanced start in building the products and marketing, and fulfilling the needs of these communities.


Thor - 00:07:32:

There are so many questions that come to mind, but as we spend a little bit more time on social intelligence, what advice would you give brands looking to strengthen their approach there? And what do you think are some of the most common misconceptions about this area?

- 00:07:47:

Yeah, it's really evolved over the last few years, I think in a good way. When social intelligence first started being used, it was treated as the fourth bullet point. And what I mean by that is, market research teams would present a finding or make an assertion and say, “And this report validates it. And this survey we did validates it, and the Google search validates it, and also social validates it,” and it was kind of treated as, “Here's just one more point to prove the assertion I'm making.” But really, fundamentally, what it is is a reflection of our society. And it does tend to be, in a lot of cases, a safe place for people to go to talk about their passions and their interests and to share parts of themselves that they wouldn't think to share in a focus group or survey or in a lot of cases, they wouldn't be reached by these traditional market research methods. To start to transform the way we think about social media, as this is not just a place for broadcast, this is actually a place for connection. And how can we make sure that we're honoring that connection as brands and looking at that on a day-to-day basis, that this is not a job for just our customer care team or just our PR team to be listening to what our consumers are saying to us? This is a job for the whole company, and let's evolve the way we think about social intelligence to be an everyday essential connection with our customers.

- 00:09:05:

I absolutely love that and it ties a lot to some of the discussions we've had on this podcast that have been around understanding the human being behind the consumer and it's clear to me that this is a great way to do just that. 

But if we dig a little deeper into that, do you have any stories you could share that would illustrate how insights sourced from Riviter have fueled foresights or innovation and that in turn has led to a client building a much better process or product or whatever it might be?

- 00:09:38:

Yeah, I'll share a story from one of our very first customers because I think it illustrates just how powerful respecting communities and knowing who you're looking at can be. So this was a media company that had a really powerful, really strong readership, very highly interested in fashion and they had been enlisted by a large mass US retailer to design a fashion line for them. And in doing so, they had this captive audience. They were pulling their readers, they were putting swatches in front of them, really highly engaged, really beautiful market research and they said, “Well Riviter, we have access to our consumers, can you give us access to the rest of the market and the rest of social media?” And interestingly our data said, “Well let's also look at this retailer's audience because it's a different audience and you're selling through them and let's make sure that there's a match there”.

It just goes to show to let's really understand who we're talking to and not just the trends that are happening in the macro environment or with our own research. Let's look with a wider lens and make sure that we're connecting to the people who are ultimately going to be served by this product.

In fact a lot of these very fashionable, very interesting kind of swatches that the media company was putting forth were not the same match with this retailer. The retailer's customers were looking more for everyday, accessible, essentials and basics and this is what our data said and that ended up being the types of products that sold through the best once this product line was launched. It just goes to show to let's really understand who we're talking to and not just the trends that are happening in the macro environment or with our own research. Let's look with a wider lens and make sure that we're connecting to the people who are ultimately going to be served by this product.


Opportunities for Ai in insights work

- 00:11:05:

I love that. And AI is obviously having a big moment right now. And it's something that's on our minds a lot as insights leaders. As a CEO and co-founder of a company that provides brands and retailers with AI-based intelligence, where do you see the opportunities for AI in insights work? And conversely, where do you see the potential pitfalls?

- 00:11:28:

It's huge and it's so exciting. We've been doing this for eight years, and the first seven of it, I kind of felt like I was a little bit crazy, like a crazy scientist talking about these terms. And now suddenly, even in the last six months, it's household terminology. Everybody knows what we're talking about when we say things like generative AI. I think it's going to become a lot like the internet, right? We're going to have to weave it into our day-to-day. We're going to have to build a lot of discernment around how we use it, how we build capabilities around it, and how to differentiate between providers of it. There are two kind of bits of advice that I would share in thinking about AI and that it is:  think of it almost like a brain, and everything that goes into it is going to be important. And everything that comes out of it needs to be discerned. 

The main question to ask when thinking about any AI is how has this AI been trained? Has it been trained with legal documents and scholarly papers? Has it been trained with people's blogs from the internet? What information has been given to it? And that's something that's really important to us at Riviter as well. That we take the input of subject matter experts within every single product domain, that we understand what's important, what the machine needs to be able to recognize and to build that in. 

I think it's going to become a lot like the internet, right? We're going to have to weave it into our day-to-day. We're going to have to build a lot of discernment around how we use it, how we build capabilities around it, and how to differentiate between providers of it.

And so there are two kind of bits of advice that I would share in thinking about AI and that it is:  think of it almost like a brain, and everything that goes into it is going to be important. And everything that comes out of it needs to be discerned. 

Then on the other side is, how do I then check? How do I make sure that what this AI is telling me is true and that I know that it's been the AI produced and that I can then provide feedback and input back. And again, that's something that we've done to be able to take that feedback in. But I think it is also the responsibility of technology providers to be open and to say, “This is something that we've done with AI. This is not human-generated, be discerning, be just as careful, if not more, as you would be with something human-produced.”

But with all of that said, it's something to be really embraced. I think it's going to accelerate the pace of our innovation. I like to say it's our third team member of third intern. It's really useful and really protective, and let's just honor and recognize the things that go into it come out of it.

- 00:13:28:

I love that. And I think it's a very positive outlook to how one could see this as a new member on the team. If we look ahead more broadly, beyond AI in particular, what opportunities do you think there are for insights professionals to make true business impact and to challenge the status quo?

- 00:13:48:

Yeah, I think it can be a scary thing, but also a really exciting thing, the fact that insights have become really democratized, right? So anybody can go and read a trend report, or pull some data from Google search and kind of triangulate and provide evidence for their assertion. And so insights professionals have this opportunity to really be stewards of - how do you do that well? How do you do that responsibly? And how do we do this process of insight, rather than just taking these observations and this information? How do we transform this information into action that we as a business can take, products that we can build, experiments that we can run? And so I think there's a lot of opportunity to be that translator, to say, “Okay, not only have I produced a lot of this observational data, not only have I produced some insights, but now I'm actually going to help steward this insight into action for you, maybe even help you test this too. So maybe testing on Instagram stories to put some swatches in front of your customers.” Let's actually get this process a full loop story and bring these insights to life in partnership rather than just being an information provider.


Challenges in the industry 

- 00:14:56:

And if on the flip side, when you think about the many challenges that you could say would face these Insights professionals and the wider industry in the near future, what are the main ones you would highlight?

- 00:15:09:

Yeah, I think there's a lot of temptation to really defend the academic rigor of something, and with good reason, right? To kind of defend the sample or defend the detail with which we're going into this research. And I think while that is so important, being able to see that with the realities of the business and to say, “Okay, well, I'm going to get this information to you, is it going to give you enough confidence to move forward? And if not, I've given you really rigorously tested assertions. How do I bridge the gap for you taking this and moving forward with gusto?” I think that is the challenge that all insights professionals have always faced. But even more now it's: how do I build actionability and actually start to recommend and build those bridges? To o be able to put your prototypes or your own plans in front of consumers in an accessible, scalable, quick way so that it doesn't have to be a multi-year, it doesn't have to be super involved. Can I use things like social media to quickly see how people react to something?


Thor - 00:16:11:

I wanted to spend more time, a bit on the AI component. And a lot of the models that are generative are based off of predictive technology, which is statistical in nature, probabilistic, if you will. But at the same time, I think you emphasize on multiple occasions how it really helps you understand the person. It helps you understand the person, helps you understand the groups. And I'd like to tie it a bit to how would you make the connection to understanding the human being, the humanness, and the empathy. How do these technologies allow us to get extra strength in achieving that?

- 00:16:46:

Wow, that's such a powerful question. What I found in my experience with things like the open AI tools is that they're really good at bringing hidden things to the surface. So as we converse with people, we all inevitably come with our biases. We hear things that we have anchors for that resonate with us, and then we tend to ignore accidentally things that we don't have a reference for. 

What I found in my experience with things like the open AI tools is that they're really good at bringing hidden things to the surface

AI is biased, but not in the same way. So it doesn't necessarily know what I think is important, it knows what it thinks is important. And ideally, you have a different partner in that AI, someone who has different biases, in a sense, to be able to see those things. So one example that I'll share is that historically, when we've run our data, we've really been looking for enough sample, right? So we really want to see signs where there's credible enough signal, enough photos, enough people sharing it for us to rely on the trajectory of a trend or a product which is statistically responsible. But at the same time, we've always known, okay, well, there are certain times when one or two really impactful. People will share one thing, and that's got to have some weight too. We have to be able to tell that story.

AI is biased, but not in the same way. So it doesn't necessarily know what I think is important, it knows what it thinks is important.

What the AI has gotten really good at is distinguishing those two and saying, okay, well, I can represent the statistical mass, but I can also tell you here's this one example and let's treat it as that. But I can bring both to the surface and I can automate how you talk about that and make sure that you do and make sure that that's represented as well.

- 00:18:19:

I think this is so interesting, do you have an example?


Predictive intelligence ft. Beyoncé 


Andi - 00:18:22:

Yeah, my favorite example of this is with tie dye. So this was a trend that was actually started pre pandemic, but really had another wave. And obviously it's been around forever, but it's had a resurgence in the last couple of years. And the perfect example of this was July of, I believe, 2019. Beyonce wore a tied eye swimsuit cover up and it was posted on page six in New York and the New York publication. And this one post really kind of triggered that trend. And so to be able to say, okay, I'm looking at the markers of influence. And I can tell you that when Beyonce wears something, it's going to become a trend. Being able to come at that as a human is useful, but being able to automate that in an AI fashion allows you to see this every time it happens. And not only that, there are lots of Beyonce who we don't know about, who have just as much influence in their community and we need to be able to see those trends as well. And so about a year before its main peak, before a lot of the kind of mass search trend reports were coming out, we were able to see the signal of tiedye is going to have a big couple of years. And sure enough, it did.

- 00:19:28:

I just love the sentence, there are a lot of Beyonce we don't know about.

- 00:19:32:

Not to anger the hive, there's only one Beyonce.


Advice for a career in predictive intelligence 

- 00:19:36:

It really makes me curious about the predictive intelligence piece, right? Because it's something I think we've talked a lot about, but also is being discussed a lot in the media, which is how should we think about that? And because ultimately it's such a new component in today's, I would say discussion in the insights industry of having not the predictive component, but we talk about generative AI and the predictive element of generative AI. So what I'd like to understand is if you could go back in time and give yourself advice, obviously early on in your journey, you had a completely different perspective of what you would do because you thought you would have this incredible visual search technology, you could describe items and then you ended up doing something different. So if you could give yourself advice, if you go back in time and talk to yourself and say, hey, when you embark on this journey, working with predictive intelligence, this is my guidance, what advice would you give yourself?

- 00:20:32:

Oh, wow, that is a powerful question. I think if that were the case, I would know that these last six months for the world of AI really would have kicked it into super high gear and I would have said, stay the course, Andi, and double down on the education that you're giving to brands and retailers and your customers. So when you are telling them what AI is and how to think about it and how to train it and how to get input and output out of it, double down on that. Because there will come a day when that will make sense to them and then all of that training will come back and they'll be glad that they have that background. And then go as fast as you can when 2023 hits. Because now there's so much out there, there's so much conversation about, there's so much technology, which is great for the space, but now the trick is discerning.

Rather than the conversation just being about this really powerful tool let's talk about what we can do with it and how we can change our business with it. Not just how to use it, but how to speed things up in the day-to-day. 

How do you know a good AI from a bad AI? How do you treat that responsibly and now we've got to catch up with that education and say make sure your provider shows you how they trained the AI and is willing to take your input in informing the AI back. And let's do that quickly together. Rather than the conversation just being about this really powerful tool let's talk about what we can do with it and how we can change our business with it. Not just how to use it, but how to speed things up in the day-to-day. 


Thor - 00:21:52:

Thank you so much. As you know a large part of our audience are people in different levels of their career and in their journey. But I think for many of us it's new to apply AI in our day to day no matter how senior we are and staying on the topic of career advice, what has been the best career advice you've ever received?

- 00:22:11:

It was actually the year between my first and second year of getting my MBA. I had a choice to either pursue Riviter through an accelerator at that time or to take an internship with a major retailer which in a lot of ways could have helped the business. And so it was really a conflict. I thought, okay here's this great opportunity with this major company, great brand to have on my resume a lot of stability or so I thought at the time. A lot of guarantees and learning or I'll do this thing that I've never done before that I don't know how to do. It seemed like a big risk. And at the time, I was able to talk to one of my professors who was a successful entrepreneur, and he said, Andi, every six months I'll give myself a milestone. And I'll say, if the business can do this, then we'll go forward. And if it can't, then we'll review and wind down, because the last thing you want is a zombie company that's just kind of crawling along. And so he said you need to give yourself this chance to test that. Now you know that that big retailer will always be there.

You take a lot of that, the successes and the failures both responsibly and personally and to treat it as an experiment to say we're doing something totally new.

You know that that opportunity will be there in the future but you don't know the next experiment. You don't know the answer to that question yet. Run the next experiment, give it a little bit of time and then based on that conclusion be done with it. And that was helpful in a couple of ways, right? It helped us keep pace. It also helped to depersonalize the danger of entrepreneurship. You take a lot of that, the successes and the failures both responsibly and personally and to treat it as an experiment to say we're doing something totally new. We're going to see how this goes. We've got clear measures of whether it succeeds or fails and then we're going to go from there. And that's been helpful advice to me as an entrepreneur but I think it goes to the world of insights as well of we're going to try a new approach, we're going to introduce something new to the company. Let's give it a shot and we'll have some measures in place and we'll see if it passes those and we'll move forward if they do.


Who Andi would love to have lunch with

- 00:24:01:

That's great advice. Unfortunately, we've gotten to the end of this recording and there's one more question I have for you, which is tied to the world of insights and to a person. So the question is, who in the world of insights would you love to have lunch with?

- 00:24:18:

I would have to say anyone in behavioral economics is by obligation. But I would really like to meet Bill James, founder of Sabre Metrics behind Moneyball, just because in a very parallel way, right? Bringing data in this way to an industry that in a lot of ways didn't want it or and or didn't understand it if they did, and being kind of the pioneer to disrupting that, I'd love to just commiserate with it and hear his stories, hear how he did it. It'd be really cool.



- 00:24:48:

I think a lot of the listeners would love to be part of that conversation too. Wow, this has been such an interesting conversation, Andi. You have truly a unique view of the industry and I think we can all learn from it. But before we wrap up, I'd like to play back some of the parts of our conversation today that has really resonated with me. When we talked about an insight, you defined it as something that takes something data point and turns it into something actionable. And when you talked about AI, you said that when we think about AI and how it's going to evolve, we need to think a bit about the way the Internet evolved. So it's going to become part of our day to day. Think of it as a brain. Ask yourself, how has this AI been trained? And then how do I then check that whatever is AI is what the AI is saying is actually true? And then lastly, ensure that whoever your provider is tells you if the output is AI or human produce, but more generally, embrace it and welcome it as a new team member. So I know that I've learned a lot of things from talking to you today and I'm sure our audience has as well. So thank you so much for joining me today! 

- 00:26:08:

Thank you so much, Thor. It's been a pleasure.