Ian’s Fantasy VC Portfolio
Inspired by Turner Novak’s fantasy venture capital portfolios, I am putting together a fantasy VC portfolio.
Inspired by Turner Novak’s fantasy venture capital portfolios, I am putting together a fantasy VC portfolio. The goal simply is to learn — learn about startups, venture capital, trends, and everything in between.
The portfolio focuses primarily on Seed to Series B consumer and technology companies, but I do not want to necessarily limit myself from learning about cool companies outside of these constraints!
Here is the current portfolio with links to the investment write-ups, below:
Let me know your thoughts and feedback. What did you think of the companies? How can I make my analyses more robust? Reach out to @iandsilva4 on Twitter.
Stage: Series B // Total Raised to Date:$53M
Description:Genvid Technologies tools and services let you create immersive, interactive live streams.
Genvid Technologies plays at the intersection of video games and media to create interactive experiences, riding the tailwinds of the large and growing video game, e-sports, and live streaming markets. Their products solve important problems on both the provider and consumer sides and have the monetization capabilities to fuel future growth. Execution will be important, but the company is led by seasoned cloud gaming team.
Quantitatively, the video game industry offers ripe opportunity to capture value all across the value chain. The industry is worth $150B and growing. Tangential to the industry, the video game streaming market is a $40B+ market and is expected to grow at a 20% CAGR over the next five years. Twitch averages almost 2M concurrent viewers currently, of which ~350K are US based and has seen a 30% CAGR since 2014.
Genvid has an opportunity to take advantage of the rise of video game streaming by making the experience more interactive. This benefits both sides of the table as users get a more engaging experience and platforms are able to monetize at a better rate and magnitude due to their more engaged and larger audience.
Genvid currently offers four main products: (1) e-sports viewing experiences, (2) clickable live streams, (3) individual viewing experiences, and (4) embedded personalized ads.
The e-sports viewing experiences aims to tackle a large problem in the e-sports industry that could help unlock further growth. The main way to watch e-sports is through a live stream of a player’s screen. This is not the best way to watch game play — when we watch sports, we do not watch a screen flickering between different players’ body cams.
Clickable live streams and individual viewing experiences help to further engage the user. A large problem with watching a video game is that if you have not played it before, it can be difficult to understand what is going on. Individualized experiences can offer UIs showing anywhere from the basics of the game to advanced strategic maneuvers being made, depending on the viewer. Clickable live streams offer an option to “play” along with the streamer, increasing engagement and accessibility.
Lastly, ad products help provide direct financial incentive to incorporate better immersion to game play.
Genvid can monetize in several different ways. Games that offer better viewing experiences (e.g., awesome esports broadcasting) can drive better sales performance of the game. Accordingly, Genvid can offer their interactive streaming APIs to offer unique and creative ways to engage with the game. Their integrations with Unreal and Unity engines target just that.
Similarly, Genvid can provide their services to platforms like Twitch and Youtube whose business is based on providing outstanding content-viewing experiences.
Genvid can also look to monetize the interactivity by offering new advertising channels, such as personalized ads in the game.
Given the growth of video gaming, e-sports, and streaming, Genvid has a lot of potential value to capture.
The leadership team comes from the closed ‘supercomputer-powered’ cloud gaming business Shinra technologies, where they held roles of Senior Vice President, Director of Business Development, and Director of Technology. Shinra aimed to let you play games so visually intense or complex that they could only be played through the cloud. This experience gives them strong industry connections on the publishing side and deep understanding of the infrastructure that supports video games.
Though based in New York, the team also has strong connections in Japan, where video gaming is a large industry.
Genvid’s position as the go-to interactive streaming partner can help it grow into the provider of massively interactive live events (MILEs). One of the ways to think about them is that they are like Netflix’s Bandersnatch, except every decision is made by the collective viewers live in real time, changing what everyone sees and permanently impacting the story. I believe that the future of storytelling will be more personalized, interactive, and social, and owning the infrastructure for MILEs will be valuable.
For more on MILEs, check out my piece that touches on their importance to making storytelling a social experience and Matthew Ball’s perspective, co-written with the CEO of Genvid, as a way to grow cloud gaming.
Stage: Series A // Total Raised to Date:$15M
Description: Parrot Analytics measures and predicts global demand for content, across all platforms, around the world.
Parrot Analytics helps media companies better understand the performance of their content and forecast what content might perform well, to deliver better top-line results. Going beyond viewership counts that legacy competitors like Nielsen offers, Parrot Analytics is able to gauge more important metrics like viewer enthusiasm and its effect on customer subscription or cancellation decisions. By directly providing value to the profit engine of media companies, they should be paid handsomely for their work.
An important capability for media companies is to understand how their content is going to perform, and more generally, what types of content perform well. Currently, the data they use to evaluate and predict performance comes from companies like Nielsen, which offer viewership counts. However, there is a lot more to content success than simply how many people have viewed the content. Think about all those times you had The Office on in the background while you were scrolling through social media or eating dinner — that’s not an engaging experience.
The underlying data is what informs the decision making the determines how profitable media companies are going to be. Former VP of Engineering at Netflix, John Ciancutti, has said:
“Netflix seeks the most efficient content. Efficient here meaning content that will achieve the maximum happiness per dollar spent. There are various complicated metrics used, but what they are intended to measure is happiness among Netflix members. How much would it go up if Netflix licenses, say, Mad Men vs. Sons of Anarchy?”
A data provider that can provide an edge can capture significant value for its contribution to a media company’s profit engine.
Parrot Analytics offers a platform that quantifies global demand for content using the world’s largest audience behavior datasets. Parrot aims to measure the ability of a show or film to grab a viewer’s attention — an increasingly scarce commodity. In addition to collecting viewership counts, the company is also able to capture enthusiasm for a show and use that to predict impact on subscriptions or cancellations.
They use “demand expressions”, a metric that takes into account a host of signals across the internet, such as search engines, social media chatter, fan and critic sites, social video views, piracy downloads, streaming traffic, downloads, and even traffic to Wikipedia and micro blogging sites, all on a per title and country basis. Additionally, metrics to quantify a show’s “rate of decay,” or shelf life, as well as its “travelability,” its potential appeal in various regions help companies decide where to allocate resources. These are all much more advanced and informative than simple viewership counts.
Parrot monetizes through a monthly subscription service. For non-enterprise clients, it is $59 per month, while the enterprise offering is likely charged similarly to other enterprise software (e.g., per user, tiered offering, etc.).
As discussed in the market section, because the high-value service they are offering media companies in a mission-critical area, Parrot Analytics should have no problem monetizing, as long as its platform provides value to its customers.
CEO Wared Seger has a science background, with a Bachelors degree in neurobiology and a Masters degree in Bioscience Enterprise. Before founding Parrot Analytics in 2014, he served as a mentor and leader of a premier student-led organisation that seeks to foster innovation and entrepreneurship in science while creating links between the academic community and the wider New Zealand high-tech industry.
Stage: Series B // Total Raised to Date: $50M+
Description: Hive offers data labeling, ML model development, and enterprise application development to deliver "AI-as-a-Service".
Hive helps companies build AI/ML solutions for enterprises from data accumulation to enterprise application development. With a full-stack solution, Hive can truly build proprietary, custom solutions that solve the unique needs of customers and grant a competitive edge that is hard to replicate. As the competitive edge could be a differentiating feature, Hive should be able to capture high-value margins. The founding team brings strong technical and business skill sets that should enable them to balance ambitious vision, technological constraints, and financial discipline.
The potential of artificial intelligence and machine learning is well known. However, implementing them can be challenging. Ideal solutions are often highly custom with either a vision or unique circumstances that an out-of-the-box solution can fully deliver on. This is being mitigated by the maturation of the AI industry as offerings become more niche, however finding the right partner and trusting them to deliver the product as needed can be difficult.
One solution to this issue is having a “product-less” company that can come in and build custom solutions tailored to your business. Companies typically look to implement AI in order to (1) optimize efficiency of a workflow (i.e., nice conveniences) or (2) gain a competitive edge via the capabilities of the proprietary technology. For media companies where how good the first three articles on the page are determine whether you obtain a customer or hedge funds looking to beat the market, you cannot use generic solutions that rely on commonly available data. You’re not going to get your edge and attempting to perform this in-house can be daunting and costly. This is where Hive has an opportunity to be a trusted AI partner that delivers that edge.
Across creating a proprietary data set, training and developing ML algorithms, and putting the front-end needed to complete and enterprise application, Hive offers the full-stack of AI engineering.
By working to build thousands of ML models, Hive will have deep competency that will be able to produce higher-value models than competitors can offer and cheaper models than companies can create in-house.
As they establish their position as the leading custom AI-as-a-service provider, the company can leverage its deep expertise and vast amounts of data to start to create a monolithic AI engine that it can leverage as the base for its future projects or even products. AI/ML algorithms have not improved dramatically over the past few decades (we just have the computing capability and sufficient quantities of data needed now). Rather than algorithmic genius, the most important part for developing a differentiated AI/ML product is the quantity and quality of the data. Managing the creation of the data and working with several customers bring quality and scale that will set it above and beyond competition.
Focusing on the area of AI/ML that provides competitive edge, the value that their services bring can be enormous, giving them pricing power over customers. In this regard, targeting the high-end of the market should bring in healthy margins for the company.
Growth might be a larger challenge for the company as not having a specific product lowers awareness of the product and customers might be tempted to go with a (potentially) cheaper, lower cost solution for the product. However, good marketing and sales can certainly offset these.
Hive was founded by Kevin Guo and Dmitriy Karpman. Kevin, CEO, brings a venture capital background, as well as a Bachelors and Masters in Computer Science from Stanford. Dmitry, CTO, has a Bachelors in Computer Science, Mathematics, and Statistics from the University of Missouri and Ph.D. in Computer Science from Stanford. Together they founded Hive in 2013.
They have strong technical backgrounds and are run by someone who bridges the gap between tech and business. This should set him up well to be able to balance ambitious vision, technological constraints, and financial discipline.
Stage: Seed // Total Raised to Date: Unknown
Description: Latitude creates limitless, self-guided entertainment game worlds, powered by AI.
Latitude is creating the future of storytelling by using artificial intelligence to produce dynamic and interactive narratives in real time. This is part of a broader trend of making more interactive storytelling media, but while others are focused on the video/graphical aspects of producing content, not much attention is being paid to autonomously develop story lines. Latitude is doing so through its game, AI Dungeon, and is seeking to serve as the underlying infrastructure for creators to use narratives created by AI in their stories. The team has already executed spectacularly on technological development, user acquisition, and monetization, setting them up for success.
As I have written previously, to date, media experiences have always been standardized and passive ones. The future of media is making the experience personalized and active. Allowing the user to engage with a narrative in a meaningful way will offer superior entertainment value and become the predominant mode of storytelling.
However, traditional media firms are not the best set up to lead us into the future. They will be disrupted by the video game industry as its technology is much more aligned to creating dynamic, real-time stories. For example, it takes Pixar about 24 hours to produce one frame; when you do something (e.g., move your character, shoot a gun, etc.) in a video game, you are creating a new frame, one that has never been created anywhere before in 1/60th of a second.
Video games are already invading the storytelling media space in their home turf. Epic Games’ Unreal Engine is being used to produce scenes for Disney’s The Mandalorian. Currently Epic is just creating backdrops, leaving two main capabilities of producing content: (1) script writing and (2) acting. Many companies are working on developing digital, artificial humans (among other video/graphical necessities) that will be able to one day produce Oscar-worthy performances. However, few are working on technology to autonomously develop narratives— this is where Latitude plays and will win.
Latitude’s current product is AI Dungeon, a text adventure game which uses artificial intelligence to generate unlimited content. The game uses GPT-2 (Generative Pretrained Transformer 2) natural-language-generating neural network, created by OpenAI, and is transitioning to GPT-3.
Users can endlessly respond to a story event with a response of their own that will influence the narrative. The game has its limitations, but is inspiring of the future of storytelling by AI. This is evidenced by the 1.5M users and $15K its Patreon campaign generates monthly.
Currently the game offers a “freemium” pricing model where the game is free-to-play with a premium version of the game that features the “Dragon Model”, a better model based on the new and improved GPT-3. The premium model is priced at $9.99 per month, more expensive than many AAA games and not much cheaper than a Netflix subscription. However, considering their income from Patreon, which is essentially a donation, perhaps they have enough of a user base that can support this price and fuel further development of the game. They are likely using this as a “nice-to-have” income that helps offset operating costs, but they will likely need to revisit pricing as they continue to grow.
The grander vision is not necessarily to use AI Dungeon as a tent-pole game that differentiates itself through its AI-based story line. Instead, the team plans to use this game to improve their AI-based storytelling capabilities to create an underlying infrastructure platform for other game creators to build their own AI-based games on. This platform route will best take advantage of their technology and best capture the market potential of autonomously developed narratives discussed above.
AI Dungeon was created at a hackathon around the first release of GPT-2. After seeing traction from the initial game he created, co-founder Nick Walton, decided to continue building on it. Along with his brother, co-founder, and CEO Alan Walton, they created Latitude.
Nick brings a background in computer vision, robotics, and AI, while Alan spent time in data science and analytics prior to creating the company. Thus far, they have been immensely successful across technological development, user acquisition, and monetization and have already expanded the team to six people.
Stage: Seed // Total Raised to Date: Unknown
Description: FourFront combines real-life actors and AI to create social and cinematic experiences.
FourFront Media tells stories where consumers are by leveraging artificial intelligence to interact with the audience at a personal level through social media platforms. This rides a larger tailwind of stories going trans-media to reach consumers anywhere, as seen with the launch of Star Wars’ The Mandalorian TV show and Netflix’s The Witcher TV show. The product to date has focused on influencers telling stories, but has the potential to expand to even larger media producers. The team is led by Ilan Benjamin and Anna Melamed who have storytelling backgrounds, but their willingness to be agile in finding product-market fit will drive their success.
In the world of storytelling, one of the most important consumer trends is a simple one — audiences want more. One manifestation of this not having to wait several months or years for the next opportunity to engage with a storytelling universe. They want to be able to interact with the world whenever they want.
One of Disney’s strokes of genius with the Marvel Universe was it’s innovation to turn it into a profit machine by increasing the number of releases per year. Thus, audiences could keep up with the Marvel Universe every couple months, rather than every couple years.
The next step of this transition is taking fictional universes trans-media. In this transition, stories don’t just have multiple different movie releases in the year, but they span different types of media altogether—movies, TV shows, books, video games, etc.
The most popular example of this is The Mandalorian, a Star Wars spinoff. The series has its main episodic trilogies and has recently introduced side stories like Rogue One: A Star Wars Story, Solo: A Star Wars Story and the TV series Star Wars: The Clone Wars, giving fans tremendous amounts of content. However you only interact with movies once and the gap between them is still large. As such, Disney launched The Mandalorian, which adds just not another story to follow, but one that takes place over several weeks, due to the different media format. Star Wars fans now have new content to consume year round.
Another example is The Witcher, based on a 1990s book. The book was turned into a video game in 2007 and is now a TV show produced by Netflix. The Witcher started streaming, the video game saw its player count grow 3-4x and the book series returned to the Times Best Seller list. The media experiences complement each other.
Owners of storytelling IP have plenty of partners to work with for the alternative media types discussed above (book publishing houses, video game developers, TV/movie producers, etc.). However, one area to even more deeply allows consumers to interact with stories wherever they are is to bring the stories to the platforms that they are already on regardless — Instagram, Snapchat, TikTok, etc. FourFront Media offers storytellers the ability to bring their stories to directly to consumers.
FourFront offers the ability of for people to tell stories through social media at a personal level. By combining artificial intelligence with content produced by the influencer, the audience is able to interact with stories through their own social media profiles.
To date, FourFront Media has conducted two stories with influencers. The first is with an LA-based actress and influencer who has a stalker who kidnaps and kills her. Users are sent on a chase through Instagram, Facebook, Reddit, and Medium to figure out who the killer is. In the end, the killer is you — you are sent an image with a deep-faked image of you on the killer’s body. The second story is around an actress/influencer who gets blocked by all of her friends on social media and has no idea why. Over three days, she sends text, audio, and video messages directly to users to figure out together why she got blocked.
The product offers a unique way for influencers to interact with their audience in a world where unique content is a big differentiator. However, it remains to be seen how genuine or interesting it is to interact with an AI representation of an influencer in a story. As discussed above, FourFront might see better success in catering their product to storytellers trying to bring their universe to their users, no matter where they are. Their willingness to be agile in finding product-market fit will be the key to their success.
As it offers a similar service to that of Twilio, FourFront should price its products to developers based on usage. This would be the primary source of revenue that drives FourFront’s success.
FourFront could earn money through other ancillary revenue opportunities like micro-transactions to deepen the story and brand partnerships within the story. However, these are just icing on the cake.
FourFront was founded by Ilan Benjamin and Anna Melamed, who founded the company in 2019. Ilan comes from Inkitt, a Kleiner-Perkins backed crowdsourced publishing platform that offers a “bite-sized”, immersive reading app, and brings expertise in interactive screenwriting, particularly for the “Snapchat” generation. Anna brings a journalism and marketing background.
A challenge might be developing out the technological capabilities, given their backgrounds, but they have been able to produce a well-performing product, thus far, so this may be less of an issue than it seems.