Predictions are a tricky sport to play. Should you get all of them proper, you performed it secure recapping current developments or gave wishy-washy statements which might be arduous to confirm. Should you acquired all of them unsuitable, you probably went for the headlines with little accountability on these claims. Yearly, we attempt to discover that steadiness to assist information enterprises within the yr forward. Nevertheless, considered one of my favourite pastimes, is wanting backwards. The objective isn’t self-validation — it’s accountability. We ask ourselves the identical questions yearly: Had been we proper? Had been we early? Did we miss vital alerts? Or did the market transfer in a course we merely didn’t anticipate?
We simply revealed our newest report card. And as at all times, we don’t grade ourselves generously. Some would possibly name me a “powerful grader”. Learn the complete report: Scoring Our 2025 Cloud Predictions for extra.
Now let’s discuss 2025. Though a world pandemic didn’t strike, it was a yr of utmost volatility as AI acceleration collides with the cloud market amid a time of chip shortages, new (and someday vanishing) tariffs, rising sovereignty issues, and quickly approaching sustainability deadlines. Total, we scored higher than earlier years. We had been spot on or partially proper for a couple of and confronted exterior challenges for different predictions.
Our greatest hit: Built-in RAG providers develop into the most popular new cloud providers.
We predicted that hyperscalers would quickly combine vector databases, embeddings, and vector search into first-party RAG providers, leaning on each homegrown and open-source frameworks. This prediction was spot on. AWS, Microsoft, and Google set the tone globally with making RAG a core platform functionality, in Bedrock Data Bases, Azure AI search, and Vertex AI RAG Engine. Chinese language cloud suppliers adopted in sort by embedding RAG deeply into native information administration choices. At the same time as business consideration shifted to brokers and multicloud container platforms, RAG by no means light. As an alternative, it turned the foundational spine of context engineering for agentic workflows.
Our greatest miss: Hyperscalers miss CO2 emissions objectives — once more — however with enhancements YOY.
We predicted that hyperscaler emissions would rise from 2019 ranges however enhance yr over yr in 2024, signaling a rebound from AI‑pushed will increase. That solely occurred for Microsoft; AWS and Google emissions went up. There are some early indicators for rebound like Google reported decreases in scope 1 and scope 2 emissions however AI continues to be hitting sustainability objectives considerably.
In order for you the complete breakdown of how every 2025 prediction scored and why, learn the complete report. And in the event you’ve adopted this collection earlier than, you recognize what comes subsequent: we’ll do that once more subsequent yr with our 2026 predictions.
If you wish to dive into this report or have any questions round FinOps, cloud price administration tooling (incoming Wave in This fall 2026!), cloud platform groups, cloud abilities, cloud governance, or business clouds you will discover me at [email protected].
