Good morning. Today we have more reader feedback. This edition is focused on everything not related to the sexual harassment scandal: Startup failure rates, Blue Apron, and corporate innovation.

I’m trying something new by including the Twitter handles or email addresses of the responders that were interested in keeping the conversation going. Remember, all feedback you send me is anonymous unless you say otherwise. (PS. Find me on Twitter, too: @eringriffith.)

Lots of positive feedback on this column, but the most common criticism went something like, “Actually, you’re not including startups that failed before they could raise venture capital! If you count those, the number is much higher!”

Yep. I looked at a lot of data sources for this article and used Cambridge Associates because their data was robust, first-party, went back two decades, and included specific return thresholds. I found no data sources that specifically track the number of tech startups formed. Defining what constitutes company and what is two college kids with an idea is as difficult as defining what is a startup and what is a regular “lifestyle” business. The only data I found that includes pre-venture companies was from the Census Bureau, but includes all small businesses, and obviously does not track investor returns.

The second-most common feedback revolved around my definition of what constitutes a failure. “Actually, 1x is a bad return!” Or “Actually, many companies that return 1x or more to investors make no money for their founders or employees! That feels like a failure!” Indeed! I had to draw the line somewhere. That’s the threshold I chose. ¯\_(ツ)_/¯ Also, thank you to the reader who offered to nominate me for a Pulitzer for this article. (Ha!) Onto the rest of the feedback.

Simeon Simeonov writes: The big change that happened in the aftermath of the 2008 crash was that the shape of the investment funnel changed rapidly. The seed-to-A death rate shot up as seed money increased and Series A money (and firms) decreased. (Further, lean startup techniques have increased the total number of zombie companies that are barely subsisting: Not growing but not dying. If you consider those, the failure rate would be even higher.)

Zach writes: I was shocked at some of the practices of funded start-ups when I was exploring acquisitions of them. I saw their cavalier attitudes in the wake of thousands of customers negatively impacted, dozens upon dozens of employees let go and tens of millions of dollars burned through. They were unapologetic for their mismanagement (my word, not theirs), and seemed to indicate they were just following their marching orders as a start-up (i.e. change the world or fail fast).

Sean Browne writes: You will see that the 60%-70% failure rate matches what most VC’s are aiming for in their portfolio in way of a .300+ batting average on investments. My guess is that over time that number will continue to go down as VC’s get smarter in providing resources above and beyond capital.

Doug writes: If up to 6 or 7 out of every 10 ‘venture funded’ startups fail as the Cambridge Associates research suggests, that sure is news, and a crazy high number that founders should be really weary of.

Will writes: So true that Silicon Valley asks one to be skeptical of all establishments except the Valley establishment. The irony is lost on most.

Steve writes: If I were a large investor or board member of Uber, I would worry about the downside to the company and liability to the board if/when the government decides to come after the company. Every decade has its villains: Milliken in the 80s, Quattrone or Martha Stewart in the 90’s, Greg Reyes in the 2000s, or Martin Shkrelli of today. I would worry that the government needs to make an example out of someone or some company given all the greed and excesses that Silicon Valley has come to represent. And to me it’s likely more a someone than a some company since no one went to jail for the banking failure of 2008 and there continues to be public outcry.

I’d be scared that Travis is the perfect foil for the government to hold up. Who could find a better posterchild? The government could easily make it nearly impossible for Uber to succeed and for that reason I believe Travis had to go. And since adversity tends to lead to opportunity more than it leads to more adversity (Uber’s business continues to kill it even in light of all the bad actions and press), I believe it made sense for the board to clean house and then find a kinder. gentler CEO who the government would not want to go after.

Nate writes: The Uber business model is ready to be disrupted, probably by one of the auto companies, if they get their act together (a big if). A monthly subscription that’s less than a car payment (apparently average in the US is $489) offering Zipcar-style car access and on-demand rides with a lower per-mile rate is really compelling, and makes it much easier to insert the autonomous vehicles these companies are working on anyways. Uber is an incumbent in the quickly changing “mobility” marketplace.

On Corporations and startups (part three of this series is still coming, by the way):

Jake writes: Companies want to appear to be forward thinking by associating with startups through accelerators and innovation teams but in reality, nothing changes when it comes to empowerment and getting anything meaningful done. To your point about these companies and their business models being disrupted, I have now started to refer to this as “Innovation Dialysis”, because the companies are clearly not long for this world.

Michal Kauffman writes: By Stage 4, in addition to the panic the company may be feeling as a whole, all sorts of competing interests come out of the woodwork when it comes time to actually move forward with significant investments and real money: from the European tech team that is jazzed about the acquisition, to the U.S. tech team that’s threatened by it, to the corporate VC team that hates it because it will undermine a competing investment in their portfolio, to the Services Division as a whole worried about their jobs if the acquisition goes through and much of their work gets automated, etc….

And these competing interests help explain why even with a Stage 6 new CEO, failure rates are still pretty high. When we get asked to help post-acquisition with the integration/implementation phase, those competing low-level and mid-level interests are still there, still undermining the acquisition, even if the deal was and is the right move for the company overall. It’s a hard problem for incumbents solve (even with a new CEO), and almost always the problems are cultural and run much deeper than any single particular technology investment or pivot.

Andrew Olsen writes: I was struck by the contrast between how established biotech and pharmaceutical companies have dealt with this problem vs. what you described in your 6 step process. Big biotech and pharma companies are absolutely dependent on being able to successfully innovate by partnering with and acquiring disruptive startups because of the nature of the patent cliff in pharmaceuticals.

Jonathan Lehr writes: I disagree about portfolio approach and that more meetings and points-of-contact is better. I think that’s where classic “Innovation” teams go wrong and spend too much time with accelerators and demo days. Key is really starting from within, figuring out some business pain points — big or small, across front and back office — and then taking a curated approach to meeting relevant companies. These meetings end up being more focused, the startups can be properly evaluated against business needs, and when alignment is found, there is business support on the enterprise side to help push along the proper testing and contract signing when the startup successfully addresses the need.

On the failure of Hello, the sleep tracking hardware startup that shut down. (And newly relevant with the news of Jawbone’s liquidation.):

Idan Cohen writes: As the founder of two hardware companies, I agree that building hardware is hard. It introduces a layer of complexity, which any physical item brings. There’s firmware on top of software, there are manufacturing and supply chain and fulfillment challenges. And of course, it requires a longer time to market and much more financing to do it right.

But, how easy is it to build consumer software/apps these days? What’s the last company that made it? How much money is being poured into consumer software just because of the fear of losing the next Snapchat but eventually investing in another failing social app that ends up shutting down or getting “acquired.” How much money was poured into Path?

Building consumer products is hard, be it software or hardware. There is a unique difference in the pain of a defunct $200 object on your shelf vs. an app that you need to delete from your home screen.

Mark writes: It’s a morality tale be re-told hundreds of times in Silicon Valley right now. Here’s how it works:

  1. Founder creates something cool
  2. Founder raises money at inflated valuations
  3. Crazy valuations ensure lots of $$$ without the forfeiture of more than 50% of voting power
  4. Founders stay in control, regardless
  5. Company begins to fail – founder holds on harder
  6. Investors beg/plead founder to allow a real CEO to be hired
  7. Founder holds on harder
  8. Company tries to sell itself while failing
  9. Company fails
  10. Founder still in charge

Examples are legion – Zynga, Snap,…