Hacker News new | past | comments | ask | show | jobs | submit login
Techstars Graduates’ Survival Rates: What the Numbers Show (wsj.com)
106 points by markmassie on Nov 22, 2014 | hide | past | favorite | 39 comments



I know a lot of zombie startups that make little to no money, yet they raised seed rounds that will last them 2 years.

I know a handful of other startups that were "acqui-hired" but the founders did not see any returns (and the VCs didn't even get their initial investment back). Even in the case that the returns were 1.2x the initial investment, these are not the winners that investors are looking for.

I would love to see a more in-depth analysis of these situations, but unfortunately, many of them are not publicized.


I loved TechStars but the WSJ analysis seems flawed. Primarily because I think that data is stale (there are definitely zombie/dead companies counted as "active") and a category like "acquired" it too vague. The way it reads it seems like that was a good outcome (and for companies like Gradcad it was!) but many others it is basically acquihire.

There was a good conversation between Dave McClure and Sam Altman about value of the grand slams vs singles/doubles: http://www.youtube.com/watch?v=489JA4ERzUY


I don't think its flawed as much as limited by the availability of hard data about acquisitions, or the profitability of existing companies, and they addressed those issues to some extent.

So basically I see it as interesting, and better than nothing at all.


This data isn't as hard to collect as it seems. Contact each company, and ask then. There are less than 1000 data points here.


Companies are under no obligation to co-operate. I also think you're mis-under-estimating how news stories are written. Maybe this is an opportunity for people who specialize in telephone surveys? As boss or an employee question is about good use of company time. Not to mention the p/r and confidentiality considerations.


Even if companies were willing to be forthcoming about profitability and the terms of acquisitions, there are probably fiduciary and other legal requirements that would preclude that.


> I know a lot of zombie startups that make little to no money, yet they raised seed rounds that will last them 2 years.

Were these zombie startups in YC or Techstars?


You see this in the startup world in general. Worth noting there's often a lot going on behind the scenes at zombie startups – off-label experiments, etc., so they can be less dead than they appear.


Both.


Well, on a long enough time span, all companies die. Picking the number of years afterward to measure by is tough and may not be the best way to do so. Perhaps measuring the success by a more personal and human method describes the elephant better: Deferred income. Take an average (yes, also fraught with trouble) of yearly income for people of that skill set (entrepuners, coders, widget makers, inventors, etc ) and then see if the founders and employees made less, within a standard deviation, or far above that average, on a year to year basis. This is really troublesome to get the data at all, as it is very personal and emotional to a lot of folk. Also, the average is pretty wonky here as well. However, it should be a starting point to see how companies are doing a little bit better than a plain up/down number.


This is entirely reasonable. Return on human capital is a valid investment metric for people who contribute time rather than money.


the problem is that those companies took investor money and the purpose of them is to in the very mininum, turn a profit. Most startups don't focus enough on making money from the start and go out of business.

Looking at most of the startups in my local area, I can almost immediately tell which ones will be out of business.

hoping for a buyout isn't a business model.


I sincerely hope you are going into the VC game then! Skills like that are very hard to come by and worth a lot.


If you have a look at the Y Combinator startups (at http://yclist.com), quite a large proportion of the 2008 and earlier startups appear to be dead. Although it only lists "Dead" for a some of them, if you click on the links for the other sites (apart from the ones that say "Exited") you'll see a lot of their websites don't exist any more. It looks like about 80-90% are dead.


It would be an interesting exercise for someone to go down the list and try and contact each of the seemly dead companies to confirm.

Not sure if there is a way for people to contribute updates to yclist? For example just clicking through from the oldest first I see that the company snipshot was acquired at some point but can't see any tech coverage on it apart from a page on the acquirers website.


I'd be happy to contact and otherwise help out updating that Yclist; hadn't seen it before but there are a lot of older URLs that aren't working. With the corp name (if it's possible to find) for say, California, you can look up the organization and see if it's still active, etc.


That's about normal for VC-funded firms. About 10% are big winners, about 10% go bust, and the rest become "zombies", able to meet their operating expenses but not return their investment. Zombies are a headache for VCs; they require attention but generate no revenue.

YCombinator has been reducing the amount of startup capital each startup gets, so that the zombies die faster.

http://www.quora.com/Whats-the-real-reason-for-the-drop-in-c...


As of April they now give a flat $120,000 for 7%:

http://blog.ycombinator.com/the-new-deal


I'd like to see data on "no accelerator". This of course would be tougher to put together, since by definition there exists no coherent list of such companies in one place.

Other related data breakdowns I'd love to see:

- Number of founders

- Amount of seed capital taken

- Type of funding vehicle (equity, convertible note, SAFE or similar)

- Location


I've read one draft paper that does exactly that... Try to compare companies that have gone through accelerators to counterparts that haven't. I would hope it gets published sometime next year. I can't share the content (I agreed to keep it confidential), but I found it intriguing. (I'm the founder of http://www.Seed-DB.com so that's how I'm in touch with some researchers in the area)


@sama said: "there are now 27 YC companies worth $100 million or more!" - I wonder how many there are for TS?

@Grabcad was sold for ~100m, I wonder what the other major successes have been?


They haven't exited, but prominent companies are: DigitalOcean, SendGrid, Orbotix/Sphero, DataRobot. (They've ask raised pretty significant rounds)


I'd put Localytics on that list as well.


Anyone have a good idea of what are the 27 companies that are worth 100 million+?


These charts aren't great. A single survival curve could display all of that data, and account for the different starting times of the startups[1]. But it looks like Datawrapper can't do survival curves. Also, a 3 section pie chart is a waste of space.

[1] http://en.wikipedia.org/wiki/Kaplan%E2%80%93Meier_estimator


My PhD adviser's rule of thumb with pie charts: don't use pie charts.


Florence Nightingale invented the pie chart to be able to convey medical data visually in a time before other visual devices were invented. They still serve a role, but an argument against them today is that representing two dimensional data in three dimensions can be misleading (such as w/surface area).


Sorry, but did you mean one-dimensional data represented in two dimensions, as in scalars represented through areas?


Pie charts can be three-dimensional, but thanks for clearing up what I muddied.


Did someone say... survival analysis‽ Here's some R code to handle the censoring of the data mentioned in OP ("The program’s growth skews the overall stats toward the most recent program entrants–those that have had little chance to either fail, be acquired, or prove their independent staying power."):

---

  # http://blogs.wsj.com/venturecapital/2014/11/20/techstars-graduates-success-rates-what-the-numbers-show/
  # data:application/octet-stream;charset=utf-8,Year%2C2007%2C2008%2C2009%2C2010%2C2011%2C2012%2C2013%2C2014%0AActive%2C2%2C2%2C7%2C15%2C36%2C68%2C121%2C120%0AFailed%2C3%2C4%2C5%2C11%2C9%2C10%2C5%2C0%0AAcquired%2C5%2C4%2C7%2C5%2C14%2C15%2C4%2C1
  rates <- read.csv(stdin(),header=TRUE)
  Year,2007,2008,2009,2010,2011,2012,2013,2014
  Active,2,2,7,15,36,68,121,120
  Failed,3,4,5,11,9,10,5,0
  Acquired,5,4,7,5,14,15,4,1

  library(reshape2)
  rates2 <- melt(rates)
  colnames(rates2) <- c("Status", "Year", "Count")
  rates2$Year <- as.integer(substring(as.character(rates2$Year), 2))
  startups <- NULL
  for (i in 1:nrow(rates2)) { startups <- rbind(data.frame(Status = rep(rates2[i,]$Status, rates2[i,]$Count), Year=rep(rates2[i,]$Year, rates2[i,]$Count)), startups) }
  library(survival)
  # define startups which have been acquired or failed as dead
  startups$Alive <- startups$Status == "Active"
  # define startups in 2014 as 0 years old, etc
  startups$Age <- 2014 - startups$Year
  sf <- survfit(Surv(Age, Alive, type="right") ~ 1, data=startups); summary(sf)
  #  time n.risk n.event survival  std.err lower 95% CI upper 95% CI
  #     0    473     120  0.74630 0.020007     0.708099      0.78656
  #     1    352     121  0.48976 0.023007     0.446680      0.53699
  #     2    222      68  0.33974 0.022007     0.299236      0.38573
  #     3    129      36  0.24493 0.020778     0.207412      0.28924
  #     4     70      15  0.19245 0.020269     0.156552      0.23657
  #     5     39       7  0.15790 0.020407     0.122571      0.20342
  #     6     20       2  0.14211 0.021202     0.106083      0.19038
  #     7     10       2  0.11369 0.024715     0.074248      0.17409
  plot(sf)
  # https://i.imgur.com/76B7AxN.png
---

Since there are no covariates or anything in the provided data, we just get a curve. It looks like a pretty steady decline per year, with half of them 'dying' in the first year. The curve flattens out towards the end, which suggests that there might be some sort of time-varying hazard going on (possibly the accelerator has gotten less picky and the earliest startups were best?).

Of course, there's a bigger problem: one might argue that treating 'failed' & 'acquired' the same is painting a misleadingly negative picture - surely acquisitions represent successes? But we can't mark acquired as 'alive' because then they'll never die and then the graph is just of explicit failure... So let's switch to a form of survival analysis which has multiple kinds of deaths, 'competing risks survival analysis' (https://en.wikipedia.org/wiki/Relative_survival); we'll treat acquisition as one form of death, failure another kind, and any startups which are 'active' are considered censored (we don't yet know what their fate will be):

--- library(cmprsk) sc <- cuminc(startups$Age, startups$Status, cencode="Active"); sc # Estimates and Variances: # $est # 1 2 3 4 5 6 # 1 Acquired 0.0134537767 0.0791545678 0.172799416 0.223444079 0.321616811 0.397350061 # 1 Failed 0.0141745147 0.0579750421 0.118175302 0.229593560 0.299716940 0.375450190 # # $var # 1 2 3 4 5 6 # 1 Acquired 3.62406271e-05 0.000301240943 0.000805565675 0.00118663857 0.00205862260 0.00275961348 # 1 Failed 3.97263451e-05 0.000220342717 0.000570181024 0.00140171794 0.00200661959 0.00272841167 plot(sc, lty=1, color=c(3,2)) # https://i.imgur.com/Tafk9B0.png ---

If we only consider 'failure' as a bad outcome, then this is more helpful than the first survival curve, as we can read off the risk with time easily from the graph or table.

This graph is the opposite of before, we're now seeing the cumulative risk over time for each kind of death - eg by 7 years after founding, a startup has roughly 40% chance of having died at some point, roughly 40% chance of having been bought at some point, and just 20% to still be active; given another few years, I think actives would drop to ~0% and it'd be roughly 50/50 - and a half-failure-rate is close to the summary in OP:

> Techstars failure rates, at least so far, are a little lower than the industry average, according to estimates from the National Venture Capital Association, which says that overall about 40% of venture-backed companies fail, 40% produce moderate returns, and 20% produce high returns.


I wonder if most of the people behind these startups are still working under the "work hard, huge payout" mantra, given the evidence against that highly unlikely outcome.


Does this include startups that were incubated by another accelerator which was bought by TechStars?

Excellerate Labs in Chicago comes to mind..


What makes you think Excelerate was 'bought'?



So 0 IPOs so far?


Can't tell if this is a joke. Typical time to IPO is like 8+ years aka longer than TechStars has been around.


What is always deceiving in statistics like that? Authors rarely realise the fact that even if just a single startup survives with a good product, it is much better for humanity than survival of a single journalist or statistician (except R programmers, those are precious).


"We also think these breakouts aren’t necessarily a good way to judge accelerators."

Is Y Combinator a bit arrogant to say that the success of its startups is a bad a metric for evaluating their success? If it is a bad metric, what's a good metric?


Profits? The distribution of returns on investments is not flat.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: