There was a time when AWS re: Invent, the annual customer extravaganza hosted by Amazon's cloud arm, was rife with ads. The innovation that came out of the company was so mind-blowing that it was hard to keep up with the onslaught of news.
But this year was different. If last year was incremental, this year was downright slow when it came to significant news.
For example, last year 28 news about the event were published. This year, they have dropped to 18. There was less relevant news to present.
The second day keynote on artificial intelligence and machine learning focused on incremental improvements to existing products. There were so few significant announcements that journalist Frederic Lardinois wrote a post with images mocking the lack of news.
It seems like it's gotten to the point where the ecosystem has grown so large and there are so many products that the company has decided to focus on making it easier to work with and between those products (or with third-party partner products) than building things from scratch.
From a news perspective, that means there really is less to write about. Eight new SageMaker features o five new database and analytics capabilities which sure are important to people who needed those features, and need to amass an already feature-rich product set.
It's not unlike Microsoft Word over the years - it's a perfectly good word processor, so the only way to really make it better was to release one new feature after another to make it relevant to an ever-wider or wider audience. granular.
To be clear, making products work together or better with other products is not a bad thing. In fact, it is necessary. People who work within the system every day probably appreciated the ability to move between services more easily, but it's not like announcing SageMaker or the Aurora database for the first time. Those were huge.
To be fair, that's not to say the conference was completely devoid of new product announcements. In fact, the company introduced several new tools, including Amazon Supply Chain, a step toward supply chain data management, and Amazon Clean Room (which is one of the weirdest product names in a long time), a way to manage customer data from multiple disparate data sources. The latter looks a lot like Salesforce Genie, introduced at Dreamforce in September.
The company also introduced a couple of Zero ETL products. ETL stands for Extract, Transform, and Load, and it's a time-consuming process that every data scientist must go through to get their data into a format that they can use in their model. Amazon came up with a way around ETL, and it could be huge as it incorporates this concept into its various products.
While these were certainly interesting, they were not of AWS's level and ability to continue to produce amazing products year after year.
Constellation analyst Holger Mueller said the company is reaching a period of maturity, but from a customer standpoint, this is probably a good thing. "AWS is entering a period of consolidation, where integrating its products is more important than adding new products and services."
“For customers, this is good news as it increases the value of their AWS services, allowing them to move faster and be more agile, in short, increasing business acceleration,” he said.
But even Mueller wonders if incremental change is enough. “On the other hand, if the trend continues, the question is: Has AWS run out of innovative ideas?”
Anshu Sharma, CEO of startup Skyflow, believes the era of cloud innovation may be over for all cloud providers, not just AWS, because the building blocks have been built.
“I think it's almost done: all the basic primitives like database, storage, compute,… Now, these companies are trying to compete in a couple of remaining areas: market power, partnerships, etc. So it has gone from product wars to distribution wars,” he added.
If he's right, the kinds of ads we've been seeing from AWS (and other cloud providers) won't surprise us anymore. It's a similar dynamic to what we saw with smartphones a decade ago when product launches stopped being must-see TV. Distribution simply isn't as sexy as new product announcements, and if that's the case, the two-hour presentations packed with filler and incremental feature enhancements will become increasingly difficult to stomach.
