I'm skeptical of this on a very plain basis: If you had technology that could do this for companies, why would you focus on the startup market and not instead on the potentially much more lucrative public markets?
If the technology is as good as is stated in the TC article couldn't they pretty rapidly build up a listing of undervalued companies on Wall Street and buy into them?
There's not really a technology here. They are a research firm. They collected a dataset from many web startups and can sell that data or make a nice ad based website out of it(in the TechCrunch/RWW space). It could be a profitable niche. Or it could turn into little more than a thesis paper. Of course, if it turns into a profitable niche it could always grow into something different and bigger.
But the technology, at least in my understanding at this point, is nothing more than running your standard regression on a proprietary dataset.
This is a social-based startup, not a technology-based startup.
It sounds like technology, because cool-sounding mathematical prediction is the selling point... But it isn't that hard to do (though with data, and time, you can surely incrementally improve it). The story here is a community-known credential. Its value comes from other people relying on it as a baseline. Like any other credential, it doesn't guarantee you'll be any good - but it is a better-than-random predictor.
Because there isn't anything like this available already, and it would be a useful thing to have, they have a good chance. For it to be used as a reference point, it doesn't need to be great - it just needs to be betterer- enough -than random (and get adopted).
A "good predictor" just means "5% better than totally random". If you have a strong team, and good connections... your chances are surely better than if you had a weak team and no connections (though exceptions occur in both ways). If you can formalize a bunch of these factors, and find what they weigh... you have a baseline predictor that's better than totally random.
In the public markets, there's already a ton of competition, mostly because of the availability and quantity of publicly-available data. Instead of competing with thousands of other established companies and investors in analyzing public companies, we've managed to solve the problem of collecting large amounts of data on startup companies instead, into which more than $67 billion was invested last year in cash in the US alone.
Maybe because startups are small enough and therefore simple enough (in terms of variables affecting success) for something like this to work, but large public companies wouldn't be.
But really I'd be skeptical about data in this first batch of tests. I'd bet that YouNoodle used these widely known startups as a training set for its algorithm, or at least as a test set...I imagine if one of their unit tests came back and said "Facebook is gonna be worthless", someone would tinker with the algorithms until it didn't say that anymore. Usually you deal with this problem by setting aside some data points until the very end for you to evaluate your accuracy, but without knowledge of where they got their training set, you can't really do this yourself
An analytics engine / machine learning tool has to begin somewhere. Why not start on a small(er) data set you are passionate about (and that has immediate benefit to your business model and user base), that you can record and measure more easily, and that is not generally obtainable by many others, before unleashing it on the public markets where people will be less tolerant of inaccuracies.
Actually, we're pretty serious about this. Connecting a large number of people in the startup industry -is- a big opportunity in itself (and we've already gotten a high level of traction on that front). However, we've found enough evidence to suggest that data can be used to significantly improve decision-making within it also, which is even more valuable.
There are already a lot of firms researching the public company market, like Renaissance Technologies.
To make a counterpart for startups seems interesting, and worthy. With the growing market of successful startups, the industry that supplies them with services such as this one should become viable.
Techcrunch should have also tested whether predictions for startups that failed were correct too; otherwise this is way too incomplete (YouNoodle might be designed to make you happy by making you believe your company is always worth a lot).
I would prefer a tool for entrepreneurs that would show them directly what factors they can focus on to get a higher valuation and/or higher chances of success.
The page would be interactive and users would drag bars or change data such as cost of infrastructure per user, cost of marketing per user, data about competition etc, and then the probability of success would change in real-time.
Startup Idea: Take the same concept and apply it to indie bands. It's such a point of pride among hipsters to predict the next big band that I bet it'd be a niche hit.
I forget the name but there is such a company out of Spain and CA who have an algorithm that analyze the song and compare it to others that were hits to give you a number.
This makes me want to make a YouNoodleNoodle that predicts whether or not YouNoodle will succeed.
Done! Here's the code:
<html><body><h1>NO!</h1></body></html>
But seriously, there are so many subjective factors (including luck) in judging the viability of a business that I doubt their ability to do so. If they produce a better-than-random track record, they can change my mind. Until then, I am highly skeptical.
I wouldn't be so sure. While I doubt YouNoodle will be able to make valuable predictions, people are gullible and this tool will produce great linkbait.
The predictions from YouNoodle might be closer than simply picking a random number from 0-$1B, but I think a regression with a few regressors(founders previous startup success, current income, market size, etc) would be just as useful.
But most people don't understand statistics, and if you overestimate the valuations of new startups, that particular startup will likely link to YouNoodle. They could turn into a successful player in the TechCrunch/Valleywag space.
I wonder if the prediction will be more than just a single number but in fact a distribution of probabilities with predictions. That will make it less trashy.
Quit hating; this is awesome. Statistical analysis can tell upstarts what's important and what isn't in a way a blog post can't. It might reveal statistical surprises a la Freakonomics. I'm just saying, I'm happy someone is bothering to do the math on what works and what doesn't. This is something I want. If my startup were taking off, I'd pay to access this information.
I like the idea, but there will always be anomalies that crop up. So just as Google is always fine-tuning their algorithm, so must YouNoodle. Only God knows what will really happen and it sounds like there is an over-emphasis on the Social Proof of those associated with a startup. Opportunities exist that can present themselves to anyone.
It says: "YouNoodle aims to make the prediction right before the first round of funding for a company."
What happens if a company can't or doesn't raise capital? If there were a tool on the site like the just launched http://webequity.com.com.auhttp://bit.ly/3lZTZa, these startups could still launch whereby the team is paid with equity in the startup instead of cash.
Also, with the valuations in place, tools could be provided to raise money by way of small payments from investors who would also use the site. Alternatively, tools may exist to simply locate such investors with the financial exchange to take place off-site. Such investors may co-invest alongside a high-profile investor who leads the round and who takes an active role in monitoring and working with the company. In this way, a company can uphold its valuation by enabling many diversely sourced investors to access a round who are prepared to pay a premium to access such an investment that they would otherwise not be able to find or participate in, and are likewise also willing to forgoe any board seat claim.
How is it impressive that a company in 2008 predicted data that is accurate as of 2008? I'm sure they legitimately used 2005 data in doing so, but they probably tweaked their algorithms until the 2005 data was accurate given what we now know. Anyone could do that.
This is posed as being equivalent to making accurate predictions of the near future, but it's not.
There are very standard ways of validating a predictor. Using test & training sets, and tuning the output to be optimized even on unseen data.
If they did this, they should be open about the numbers. Hell, I'd be open about the data too, and make a challenge to make a better predictor. If the algorithms still run on their machines, they can only gain by being open.
But in order to be a reliable predictor, I think it would be necessary to predict the economy as well. The same startup that would be successful during the bubble would not be so successful after it popped. In other words, I don't think they have enough inputs to be reliable predictors.
On the other hand, perhaps they don't need to be reliable absolute predictors. The VC's question is basically which should be funded. This might be able to give a rough ordering, which is plenty valuable.
The life of a company from start to big exit or large acquisition is pretty long. I'm not sure the economy matters so much -- just delays things sometimes.
I guess I was making the assumption that the 7 or 8 startups mentioned there were all of them. If they had some sort of large data set I'd be more impressed.
Still not so much as if their algorithm holds up to scrutiny over the next 3.
I really believe in this to change how investors validate their gut assumptions. All YN is trying to do is take the social, psychological and environment factors to generate an accurate estimate.
Historically successful entrepreneurial teams have exhibited drive, ambition, a great network and ruthless determination. What makes you say its going to change.
I really don't see the point of it. Its like horoscopes, and its basing all startups as a neutral idea with the same success rate - Facebook is the same as Techcrunch which is the same as Twitter, even though they are completely different in the size of the target audience.
It looks as if its mainly basing the predicted investments off the founders and what they have worked on in the past, but is that good enough? We all know startups can fail no matter who is behind them - Cuil, its founders had first hand experience with the leader of search, Google, but has still pretty much failed so far.
Plus, apart from a bit of fun, who would use it? I doubt investors would use it over their years of experience in the industry, and it seems like they are in fact the target audience!
I don't really get why this is "highly controversial" unless it's just more TC link bait.
Sounds like a good idea to me; if they're able to do what what Moneyball did for baseball and what PER did (to a lesser extent) did for basketball all the power to them. What's the downside of having more accurate tools and indicators> Since the tool will be public and founders will know what to focus on.
However, I highly doubt too many VC's will take them seriously until they start producing verifiable results instead of form fitting historical data. Very few took sabermetrics and some of the other sports' quant stuff seriously to start either.
If you read Arrington's first post about YouNoodle http://bit.ly/326Nx6, he was very scathing to the point of being unfair. Along with Crunchbase, my guess is that he wants TC to be the Goto point for startups, not YouNoodle, and YouNoodle is positioned to establish many relationships with many entrepreneurs to which he may be jealous.
"YouNoodle is also basing predictions on historical data, and in a rapidly changing world that is consistently disrupted by new technologies, those predictions are very hard to make. Humans who are on top of recent developments can make subjective decisions that are far more likely to be accurate than an algorithm."
No startup predictor can take in consideration how well team members execute their ideas.
Sure, it could take in account information about past startup founders, but what if you asked it about Larry and Sergey in 1998? No past information. How are you going to predict how well they executed Google?
Even though Google is an anomaly (black swan:), as you can see in the Techcrunch article, YouNoodle would've still predicted an $80M+ valuation for Google, in part because of the high quality team and advisors they managed to assemble from the start, including Ram Shriram and Andy Bechtolsheim.
I'd be interested see how YC companies are valued by this predictor.
YouNoodle may or may not be an accurate predictor, but I bet they're good at predicting which teams/startups are likely to get funding from investors/VCs.
Surely they don't have enough data on the startups that didn't make it, that didn't have a high profile. Will the algorithm not tend to be very optimistic?
This makes sense, now that I've thought about it some more. It is logical that they would use the predictor to decide where to invest. But, they probably don't actually have any money to invest. By hyping themselves on TechCrunch or whatever, they can get some seed funding :)
If the technology is as good as is stated in the TC article couldn't they pretty rapidly build up a listing of undervalued companies on Wall Street and buy into them?