4 ms·

(Full disclosure: I was interviewed for this article and work with several of the companies mentioned.) Software startups should be worried about AWS (and "Big

by gk1 7y ago

(Full disclosure: I was interviewed for this article and work with several of the companies mentioned.) Software startups should be worried about AWS (and "Big Cloud" in general), especially those in the AI/ML, serverless, stream processing, containerization, IoT, data warehousing, and batch processing space. Those are the fastest growing cloud services and AWS knows it. I wrote about this at length and suggest ways to punch back: https://www.gkogan.co/blog/big-cloud/ https://www.gkogan.co/blog/big-cloud/ One retort I've seen from AWS defenders is: If AWS is so bad for startups, name one startup that was killed by AWS?![0] Two problems with this argument: 1. Most startups die in a whisper, not a bang. They fail before ever making a name for themselves in the market. And I happen to know that this journalist had no issues finding startups that failed or struggled, but did have issues getting them to speak on the record. Because who wants to be the loser in the story? (And yes, I do know several companies that were doomed because of the challenges mentioned in my essay.) 2. The startups that do come to mind easily--the same ones the AWS supporters show as proof that there's a level playing field--only do so because they managed to reach significant market- and mind-share before AWS swooped in. By now they're large enough to be able to chug along a while longer despite the intense (and arguably unfair) competition. Their survival, however, is far from certain. [0] Example: https://www.lastweekinaws.com/blog/amazon-isnt-killing-your-business-you-just-suck-at-it/ https://www.lastweekinaws.com/blog/amazon-isnt-killing-your-...

taurath 7y ago
There is no possibility to ramp up a new technology from slow to fast. You must go fast and huge out of the gate, or else AWS will drop in after 6 months and take most of your addressable market before you've even got off the ground. You're too slow when Amazon will put a team of 40 against your 5 person startup. And unlike Google/Hooli, they have a good chance of not screwing it up. Almost every one of the vendors at the AWS reinvent expo this year that have an "advantage" in AI, data insights, etc, will be competing with an AWS service by next year. Snowflake, Datadog, Splunk, etc. All the innovation in the room gets sucked up by the big company and offered at scale. The best you can hope for is to have a better UX and a brand that people will demand. The ones who will need to worry less are the ones who already have contracts with enterprise businesses. Relationship rather than technology companies.
- deanCommie 7y ago
  > You're too slow when Amazon will put a team of 40 against your 5 person startup. That part is not true. Amazon/AWS bootstraps new products much as a startup would - with teams of 5-10 max. I acknowledge the difference in runway availability and radically different risk of failure. But the actual dev teams are small.

by gk1 7y ago

taurath 7y ago

There is no possibility to ramp up a new technology from slow to fast. You must go fast and huge out of the gate, or else AWS will drop in after 6 months and take most of your addressable market before you've even got off the ground. You're too slow when Amazon will put a team of 40 against your 5 person startup. And unlike Google/Hooli, they have a good chance of not screwing it up. Almost every one of the vendors at the AWS reinvent expo this year that have an "advantage" in AI, data insights, etc, will be competing with an AWS service by next year. Snowflake, Datadog, Splunk, etc. All the innovation in the room gets sucked up by the big company and offered at scale. The best you can hope for is to have a better UX and a brand that people will demand. The ones who will need to worry less are the ones who already have contracts with enterprise businesses. Relationship rather than technology companies.

deanCommie 7y ago
> You're too slow when Amazon will put a team of 40 against your 5 person startup. That part is not true. Amazon/AWS bootstraps new products much as a startup would - with teams of 5-10 max. I acknowledge the difference in runway availability and radically different risk of failure. But the actual dev teams are small.