Open up your Instagram feed, and you’ll encounter a sponsored post from your favorite celebrity or a paid advertisement from a company eager to reach active social media users.
Vexed by this inundation of paid, brand-related content on social media, a team of four Stanford students are innovating a solution in an age where digital advertising is omnipresent. They think there’s a more effective way for content creators to collaborate with brands — without “spamming their followers.”
The team’s brainchild is Viralspace, a twist on a social media platform that acts as a marketplace for user-generated content. Viralspace helps brands crowdsource authentic, trendy and brand-aligned content from everyday social media content creators, which brands can then acquire and use for their own advertising purposes.
How it works
Currently, Viralspace — founded last September by second-year master’s students Hiro Tien, Apoorva Dornadula and Ashwini Pokle — is a web application where users create accounts to post branded content. Users enter #challenges, which are brand-sponsored competitions soliciting submissions related to the challenge description.
For instance, Viralspace recently released the #whyilovecoupa challenge, asking users to “show us a photo or video of you with your favorite Coupa Café products, and tell us why you love Coupa in the caption.” Once the challenge ends, the creators of the top three submissions will earn Coupa Cafe gift cards. Currently, those submissions are chosen by the Viralspace team, but the team hopes to eventually develop an algorithm to pick the winning submissions.
Companies and brands can use every submission for commercial purposes, essentially crowdsourcing advertising material at a very low cost.
On the other hand, the value of Viralspace for the content creator — a person the team loosely defines as “anyone who posts content on social media” — is allowing them to monetize photos and videos that they already create.
Dornadula said that she differentiates Viralspace from potential competitors because Viralspace is not “selling audience or taking advantage of audience.” Rather, its focus is on content creation and the virality of social media posts.
User-generated content fetches an engagement rate that is seven times as high as that of brand-generated content, Tien said, citing the 2018 Kleiner Perkins Internet Trends Report.
“Brands like user-generated content and they like to see their customers with their products,” he added.
Viralspace launched their closed beta on Jan. 28 and their open beta on Feb. 4, which will run until the end of winter quarter.
The team and its spark
Viralspace was founded in Sept. 2018 by Tien, Dornadula and Pokle. Tien, the co-founder and CEO, is a second-year master’s student at the Graduate School of Business and School of Earth, Energy and Environmental Sciences. Dornadula, Viralspace’s technical co-founder, is a second-year computer science master’s student. Tien and Dornadula met as interns at Google last summer. Their conversations sparked an interest in creating a startup that helped consumers and brands predict the virality — how likely posts are to go viral, or to become popular online — of their social media content.
Tien and Dornadula had already applied to Stanford’s Startup Garage, an intensive, project-based course at the Graduate School of Business that allows student teams to foster new business concepts, that May. When the class started in September, second-year computer science master’s student Ashwini Pokle joined the team as technical co-founder to lend Tien and Dornadula the technical expertise she gained as a former software engineer at Amazon. After iterating through different versions of their idea and conducting interviews with over 40 consumer brands, they stumbled upon the market of content creation.
Michelle Lu ’19, a product design major with a background in marketing, management and consulting, joined the team as the director of marketing and content as the students continued into their second quarter of the Startup Garage course.
The pain points
The team conducted extensive market research throughout its first quarter of the Startup Garage course to identify pertinent pain points for brands. Tien found that companies “have spent a lot of money and time” on creating promotional content to reach their Generation Z customers.
“It was pretty surprising to us, because we thought that people would want to know about the virality of their posts, but it seemed like just getting that content in the first place was a really big pain point,” Dornadula added. “Ad agencies … are way too expensive for small consumer brands.”
The Viralspace team pivoted from predicting virality towards creating it in an attempt to fulfill small consumer brands’ need for low-cost social media advertising.
Lu recalls hearing about the inconvenience and inefficiency of advertising from a woman running a small condiment store in Palo Alto.
“It’s just her and a very small team, and she spends thousands of dollars just to hire a photographer just to do one photo session of her products,” Lu said.
“They hire photographers, videographers and even actors every month to come in and do a video and photo shoot,” Tien added. “It takes about 10 to 12 hours over two days. It’s a lot of work. They pay probably $2500 per month for that, and they get around 50 pieces of content. So it’s roughly 50 bucks per piece of content, plus all the time and energy they spent.”
The team’s takeaway from the interview was the idea to automate the content-generation process: to supply brands with, say, 2500 pieces of usable branded content instead of 50 in a low-cost, time-efficient manner by outsourcing the task to everyday content creators.
From the social media user’s perspective, Lu said that content creators are also attempting to seek new ways to promote brands without flooding their own feeds or those of their followers.
“A lot of influencers who get a lot of brand offers don’t always want to post branded content and flood their feed with that, so they’re often more willing to upload to a separate website,” Lu said. “Viralspace allows content creators to post brand-related content — without spamming their followers.”
“We think we’re getting a head start in the user-generated content marketplace by delivering authenticity,” Tien said, contending that there is a paucity of authentic connections between creators, brands and customers.
Power of the algorithm
Tien believes that Viralspace offers more functionality beyond beyond simply crowdsourcing content. Despite the continual evolution of their product, the team never lost sight of their initial vision of predicting the virality of posts on social media.
According to Tien and Dornadula, a computer-vision driven algorithm — which analyzes images and real-world information to make decisions — that will help brands predict the virality of social media posts is in the works.
“Ultimately, I think the DNA of the company — building some kind of virality prediction tool — is still there,” Tien said. “We are planning on building a suite of tools that the brands can use to not only collect all these submissions but also to quickly curate the best three without spending time looking through 2500 pieces of content.”
Tien envisions programming an algorithm which will automatically rate and rank content to predict how well the post will perform in comparison to the brand’s existing social media posts. This algorithm will be powered by computer vision and take advantage of a brand’s public social media metrics, such as hashtags, likes and comments.
“It’ll probably be some sort of system powered by deep learning that will have some sort of score for each image,” Dornadula said. Images, for instance, might be ranked based on their portrayal of brand, people and storytelling ability.
“This kind of functionality … would be useful for any brand out there,” Dornadula added. “Who wouldn’t want to know how much engagement a post is going to get you before you post something?”
However, Dornadula said she is also acutely aware of machine learning ethics and fairness, something she thinks many companies today have overlooked. She hopes to probe the algorithm and question why it surfaces certain content.
“Would we rank an image of an African-American person lower than that of a white person? That’s not something we want to do at all, and that’s something we need to be really careful about,” Dornadula said.
Funding and business model
In early January, Viralspace joined an accelerator program and received $150,000 in funding from Village Global, an early-stage venture capital firm backed by luminaries like Reid Hoffman, Bill Gates, Mark Zuckerberg and Jeff Bezos.
The team is working with its Village Global advisor Chris Yeh to construct a business model. They are considering charging brands for a Viralspace subscription while keeping user accounts free of charge.
The Startup Garage experiment fund also provided Viralspace with $1000.
When the Village Global accelerator program ends in March, the team will begin their seed fundraising round.
“All the features that we’ll be building will be geared towards creating network effects so that every additional brand and every additional user [that joins the platform] adds value to the first user and the first brand,” Tien said. He mentioned the launch of an in-app voting feature, where existing Viralspace users will have the ability to vote for each other’s submissions to earn a “People’s Choice award.”
Viralspace also intends to improve the user experience for content creators, such as by launching iPhone and Android apps and delivering prizes within the app platform rather than their current method of Venmo or Paypal.
Viralspace is currently reaching out to brands like Barry’s Bootcamp, Lululemon, The Pill Club and Nike LA to offer more #challenges for users.
Ultimately, Tien envisions Viralspace being a commercial social network and a “meritocratic marketplace.”
“Our vision is to fix the social networking experience between brands, creators and consumers because we think it’s fundamentally broken right now,” Tien said.
Pokle said that in growing Viralspace, she realized “having amazing tech is not enough.”
“People should be able to understand the need in the first place,” Pokle said. “If it’s not what people need, it’s probably not useful.”
Contact Alex Tsai at aotsai ‘at’ stanford.edu.