Failcon SF 2012 – Notes

I dropped by Failcon today which is FailCon is advertised as a “one-day conference for technology entrepreneurs, investors, developers and designers to study their own and others’ failures and prepare for success.” It was the kind of conference that I normally would really enjoy. A small enough crowd that you’re not overwhelmed and a venue (Julia Morgan Ballroom in downtown San Francisco) that was downright clubby.

Unfortunately, there was quite a lot going on at work today so I was not able to devote as much attention the the talks as I would have liked. I missed the morning talks so I only have notes from a few talks, what follows are my notes.


There is no “single” point of failure

– We only get to the truth after the 5th “why?” – keep asking “why?” to get out of the weeds and uncover systemic and procedural failures.
– For example, it was not o-ring on Space Shuttle that failed, nor the representation of the data from the o-ring temperatures. If you keep asking “why?” you will start to find the process which led to failures to produce poor reports that hid early, non-catastrophic failures of the o-rings that would have raised a flag to prevent the catastrophic failure that caused the Shuttle disaster.

Complex truth loses to Simple Lies

– Simple stories are fun to tell but they also are lies that hide the more complex truth
– There is a narrative bias. The teller of the story shades the truth to augment their perspective.
– Titanic, Hindenburg, New Coke – these are not just trivial failures that can be summed up in a simple story. There are multiple perspectives that tell the full picture.

Our own egos get in our way

-Heroes are inspiring, yet stupid. To be heroic, you need to ignore common sense.
– There is little that can be learned from these inspirational heroes because they will not give you the full truth, the real lesson.

We tend to fight the last war, not the war we’re in.

– Our most recent victory (or defeat) colors the challenges before us.

Scott’s blog post on his talk.


– “In order to get to MVP, take what the founder thinks is a good first product and cut it in half. Then cut it in half, again.”
– “Lean startup is a methodology made for businesses going through times of extreme uncertainty. Not to be applied if you’re buying a franchise, or making Madden 2012.”
– In front of VCs, we play “success theater”: using vanity metrics to avoid the smell of failure. The unit of progress is learning, not number of users, not revenue.
– The most important thing to do is start. The start will always fall short of the vision. Don’t think, “I’m not wasting my time with seeds, I only want trees.”  Where do the tress come from? Seeds.


– design blindness, ideas that you come up with make sense to you but are not obvious to others
– in order to get others to understand, you need to get critique early and often.
– hard to do when you’re a designer because you want to wait until it’s done.
– if you wait until you’re done, you’ll be too emotionally invested and will resist objections.

Three stories about failure
Piano Fail
– You can’t learn the piano by reading a book about it. There is no “Piano for Dummies” book.
– To learn a craft, you must practice that craft. Learn by doing.
– Helps to find a tutor. Someone who can show you the way.

– Design something, show it to people, they don’t get it, you fix it, they still don’t get it. It’s a vicious cycle because with each turn, you lose the fresh perspective of people who have never used your product.
– Evolution of GChat on-screen instructions. Didn’t know to hit to send chat. Didn’t want to add [Send] button on a product that was embedded in Gmail. Confusion with pop-up advertisements. Whack-a-mole.
– Need to listen to people, look in their eyes to get to the bottom of the secondary fail.
– Sometimes a fix is just a very minor adjustment.
– Like the game of Battleship, even if you’re good, you’ll miss a lot. You only get better incrementally.
– People are a moving target. What works today will not work tomorrow. Design in the present.
– Just a few years ago, “Pull to Refresh” on iOS was not known even to Apple. It was developed by a third party. It’s expected now.

– Like running. It’s good to have a buddy to help show your stuff. No one wants to run alone in the rain.
– Schedule the user study ahead of time, it’s a forcing function.
– One day design sprint lowers expectations. No one expects perfection after a few hours. Best way to break out of a rut, to try out a new approach.

For more, check out this excellent series on on how to run a 5-day Design Sprint

HOW NOT TO MANAGE A PRODUCT – Mike Arsenault (@mikearsenault)

– Grasshopper built a “Refer an entrepreneur” program to allow their users to share a discount. These refers converted at 22%, orders of magnitude over their other marketing efforts which were around 1.5%. Because these refers were so successful, it became the vision for a new type of product which became Spreadable, a customized “tell a friend” widget to put on your site to help your customers, “spread the word.”

– What happened? They blew thru $500k in 14 months and ended up having to shut down the site because of lack of interest.
– Mike shows a clip from their explainer video and then notes that the full video was 3 mins long, pointing out that most people lose interest after 60 seconds.
– 12 people joined to work on the project. Many were seasoned from the previous business which gave them (over)-confidence. Bigger Team does not equal Higher Velocity
– The new product was benchmarked against the metrics that made Grasshopper (7 years old, 40K customers) a success. There is no way a startup could measure up to an established business. Churn was too high, they didn’t talk to customers that were leaving. If they did, they would have learned the widget was not performing for those that managed to install it.
– The marketing channels that were successful for Grasshopper did not work for Spreadable. Lower search volume for organic search keywords for this product as well.
– Product required much more hand-holding than Grasshopper. Advertised to radio audience that was not technically proficient to install javascript.
– Wasted our Beta List. 2000 people on the list were not invited to get involved until after the launch. Beta people wanted to get involved early. Need to keep the list “warm” with regular updates.
– Didn’t charge on Day 1. It was in the market for 6 months for free. Putting a price on a product is a quick way to get feedback on the product’s usefulness. Price helps you measure. If people pay, feedback is more pointed.
– Too hung up on marketing, not on building the right product.

SOCIAL PROOF IS NOT PROOF – Michael Wolfe, Pipewise (@michaelrwolfe)

Lessons from ccLoop, finalist at TechCrunch NYC Disrupt

– Compelling team, great investors, Big Problem (enterprise inbox 2.0), Social Proof (press) = failed after 4 months
– Why?
– Basecamp, Yammer, Chatter, Sharepoint, MS Project – all useful and compelling but ultimately fail b/c they compete with the inbox. Michael surveyed the room on Basecamp and Yammer. Most everyone has heard or used these but very few have used them in the last 24 hours.

– Fail #1 – Team Fail.
CEO was in customer validation along with many other things. Need a dedicated Hustler. Every engineering team needs a hustler. Engineers just ask obvious questions. They validate the solution, not the problem. “Do you have an email overload problem?” is not a useful question. Validation is not “selling the product” – done right, customer validation is about making future customers pay you to build a product.

– Fail #2 – Product Fail.
Lots of Likes is not as a good as a handful of “loves” (even if that handful comes with some haters). Better to have passionate users.

– Fail #3 – Social Proof Fail.
The echo chamber for friends and family prior to launch reinforces bad decisions. Seek out fresh perspectives. Do not be swayed by the tech media.






Leave a comment