How Much “AI” is there in the Veritone IPO?
Table of contents
It was only last week that we saw a new AI unicorn birthed alongside speculation that this AI unicorn had also confidentially filed for an IPO. This led us to pontificate a bit about how AI is not the next dot-com because this time it’s different. This time, the technology that everyone is raving about is showing superior value causing it to be recognized by investors, unlike the dot-com era where the promise of superior value was enough to get it recognized.
Just today, we rubbed our eyes and saw that another “AI IPO” appears to have been filed. If we’re going to see AI IPOs going off left and right now, we need a voice of reason to make sure that people aren’t just adding “powered by AI” to their marketing brochures. The company we’re talking about is called Veritone, and we can confirm that their marketing team has successfully littered their homepage with references to AI:
According to several of our on-staff MBAs who have been researching this thing non-stop, we can also confirm that they have put in place that really cool animated graphic that shows the earth and then lots of lines with dots connecting them in the shapes of triangles imposed over the earth (this is another way you can tell someone is using AI).
Everything looks good so far, however instead of getting all excited and asking them to talk about “using AI”, we’re going to just dig right into the Veritone IPO S-1 filing to see what’s under the hood.
About Veritone
Founded in 2014, Newport Beach California startup Veritone has taken in $65 million in funding so far to become “the leading cloud-based artificial intelligence media technology company pioneering the world’s first Cognitive Media Platform (CMP). We can’t help but think about another media company called Rocket Fuel (NASDAQ:FUEL) which uses AI and big data to determine which ad is best to serve at any given moment. FUEL hasn’t done so well having lost -91% since their IPO leading lots of people to believe there wasn’t really any “there” there.
Veritone is different from FUEL in how they plan to use AI and big data. Here is what they say they do taken from their website (we made all the buzzwords bold for maximum impact):
We are a leading cloud-based artificial intelligence (AI) technology company with an open platform that combines powerful applications with orchestrated cognitive computing to transform unstructured data, extracting actionable intelligence from public and private content globally.
As we start digging into the Veritone S-1 filing, we find lots of information about the potential of AI and big data accompanied by cool diagrams like this one:
The only problem is that the Veritone AI platform isn’t generating much revenue – yet. In 2016, the majority of their revenues came from “media placement services performed under advertising contracts with our media clients” and they are in the “early stages of developing their AI platform“.
In 2015, Veritone took in $13.9 million of revenues from which less than $100,000 was from their AI platform. In 2016 though, they took in $500,000 in revenues from their AI platform which shows some real growth. The only problem is that during the same year, two of their key clients (LifeLock and DraftKings) which accounted for 43% of their revenues decided to spend less and consequently Veritone’s 2016 revenues fell -36% to $8.9 million. Not good.
Veritone is pinning their growth hopes on the promise of their new AI platform which integrates 45 different “cognitive engines” from vendors that include Google, IBM, Nuance, and even some cognitive engines they developed on their own. Their goal is to offer a single cloud-based platform that their clients can send data to for processing which supports all major categories of AI like transcription, face detection, object detection, sentiment analysis, geospatial data, etc. They have already licensed their platform to several leading radio broadcasts, a national news and sports network and a major financial news network.
While these filings are always prone to be overly cautious when it comes to warnings, the following points really struck home:
- Substantially all of our revenues are currently generated from our Media Agency Business, and our efforts to expand our AI platform business may not succeed
- We rely on third parties to develop cognitive engines for our platform and to create the related APIs
- We may not be able to expand the capabilities of our proprietary Conductor technology to optimize our AI platform
Essentially they’re integrating AI technology from other firms onto a platform that hasn’t generated meaningful revenues yet and they’re not sure it’s going to work as they expect it to. While there’s no reason why it couldn’t work and create loads of value, it’s still an unknown at this point in time. On paper this sounds like a great plan. With the $23 million they’re looking to raise from the IPO, they may even look to do some acquiring. After all, there are at least 1,500 AI startups out there that could provide bolt-on functionality for the open AI platform they are developing.
We’re not here to tell anyone what they should or shouldn’t invest in (except for OTC stocks… NEVER invest in that garbage). What we are here to do is provide insights into the extent to which various companies are exposed to various technologies like AI. Our conclusion is that not many publicly traded stocks have pure-play AI exposure at the moment. When we look at a company like Afiniti, we see an AI platform that is so powerful that it makes money by taking a small cut of all the value it creates. That’s the way to do it. Tell your clients that everything is free up front and you just want a piece of all that money they’re going to save, when they actually save it.
When someone asks us, “how do you know it’s AI?“, our answer is “you know it’s AI because of what it can do that traditional programming methods can’t do“. We used the example of Elon Musk’s hair (we swear that makes sense when you read this article) to show that the proof is in the pudding. We can recognize AI because of the results, not the promise of results in the future. With Veritone, you have to ask yourself – are they offering up results, or the promise of results in the future?
If successful, the Veritone IPO will trade under the ticker VERI.
Share
