“Ever tried. Ever failed. No matter. Try again. Fail again. Fail better.”
Irish playwright Samuel Beckett’s famous quote has become Corporate America’s new mantra.
From Okta to Salesforce to Blackstone, executives told BI’s Sarah E. Needleman and Ana Altchek getting it wrong is ok. Just do it quickly and learn from it.
The tech industry has always been a big proponent of failing fast, but Corporate America is now catching on thanks to AI. The tech allows companies to quickly launch new tools to a wide group.
And, perhaps more importantly, it’s a way for executives to prove the big money they are spending on AI isn’t going to waste. (Hence why they want you to fail fast.)
It’s not just for launching products either.
Even the ideation phase can be quickly sped up with chatbots that can talk through ideas. What previously might have taken executives (or, more likely, their underlings) hours of research can get figured out a lot faster.
There are still some hurdles with the fail-fast approach.
What does failing even look like? There isn’t a big, red alarm that goes off every time an AI project fails. (Although that sure would be fun.) Executives will say they set clear guidelines for a prototype beforehand, but a tool’s benefits can be nuanced. And maybe you just need a little bit more time to really make it sing. All that makes it a lot harder to decide when to pull the plug.
Building airplanes in the sky. The rush to get things out can lead to the belief you’ll just fix things on the fly. But that’s a dangerous precedent. Just ask the video game industry. When games came in physical copies you had to blow on to get working, companies made sure their product was bulletproof. Nowadays, they can digitally ship a game knowing an update for a glitch wont be far behind. It’s a dangerous game that can cause major headaches.
Opening the floodgates. A formalized product launch slows things down, but it also makes sure everyone is on the same page. A massive greenlight risks a lack of standardization. That might not seem like a big deal for a prototype. It becomes a bigger problem down the road if the product needs to be reconfigured to fit into the company’s wider tech stack.
