4 ms·
Here it is: https://youtu.be/gKNrOIR0euY?t=2209 https://youtu.be/gKNrOIR0euY?t=2209 Transcript: "I made many, many mistakes as the CEO and I'm glad that the go
by zmb_ 9y ago
Here it is: https://youtu.be/gKNrOIR0euY?t=2209 https://youtu.be/gKNrOIR0euY?t=2209
Transcript: "I made many, many mistakes as the CEO and I'm glad that the good things we did overcame the mistakes. A typical example is... Did anyone ever use a product called "Wave"? Some of you may have. We launched it and from day one its usage went straight down. And the question is: If you're the CEO, how many months of going down before you cancel it? ... Look at the numbers, three months, six months, nine months? Eighteen months. Not good. There are other examples of that category, and again, just bad decisions."
- duncanawoods 9y agoThanks. Sounds like management by metrics instead of vision.
- throw_away 9y agoAnd if that's your metric, then why slow-roll the launch by making it invite only? If the answer is performance, like I suspect it may be, then why launch if you cannot achieve success by the sole metric you appear to be looking at?
- wpietri 9y agoI don't think so. If you'd asked Wave's designers before the launch, their vision surely would have included increasing usage. It turns out few people cared about Wave. Schmidt's decision has been borne out by events. Did Wave take off when open sourced? No. Did anything Wave-like developed subsequently turn into a big market? Also no.
- duncanawoods 9y agoI agree to a point but I want to see numbers leading to reasoning. If you identify why the numbers are bad and it is not something you can or want to turn around then sure, kill it. Killing off projects based only numbers alone is tyranny of the bean-counters. > Did Wave take off when open sourced? Eh, how many "dump and run" open-source projects take off? Bringing an opensource community to life is its own brutally hard problem.
- wpietri 9y agoThe numbers were bad, Duncan, because not many people wanted to use it. Your theory that they are "killing off projects based only numbers" is something you appear to have made up. I think Google's real error was not killing it too soon, but allowing it to be developed in a bubble and then springing it on the world with a giant, "Ta da!" It let them ignore a lot of real-world problems and questions, investing heavily in untested design hypotheses. Once you get to the point of a huge global launch, you've baked in problems in a way that is very hard to fix.
- useranme 9y ago> Did anything Wave-like developed subsequently turn into a big market? Slack. Trello.
- wpietri 9y agoSlack is basically a spiffed-up IRC. Trello is a virtual index card board. Neither is, "designed to merge key features of communications media such as email, instant messaging, wikis, and social networking."
- pbhjpbhj 9y agoIsn't Slack Wave-like?
- wpietri 9y agoPreviously asked and answered. No, slack is IRC-like. The Wave approach also mixed in notions of emails, wikis, and group editing.
- alecthomas 9y agoI know it's easy for me to say in hindsight, but I think that as the CEO Eric Schmidt's big failure with Wave wasn't "I should have cancelled it sooner" but "I should have encouraged a better, more focused product".
- ryandrake 9y agoI don't know.. I'm trying to think of a product that was fundamentally broken in v1 that was eventually "fixed" by exec encouragement/buyin/investment and ended up succeeding. Maybe Internet Explorer. Yea, I suppose Microsoft is the master of this "v1=shit, v2=shit, v3=barely usable, v4=world dominating" approach.
- whatusername 9y agoHow about the iPhone? (No copy/paste, no apps, no 3G, etc) Apple usually mail the core experience with v1, but all the ancillary roles have to wait till v3 or v4.