AcquiredThe Under-Discussed Database Market Gives SO MUCH POWER to AWS
CHAPTERS
Why databases are an enormous, fast-growing market
The hosts frame the database category as both huge and structurally important: nearly every computing workload depends on storing data. They cite the database software market at roughly $100B and growing ~10% annually, setting up why this space matters so much for AWS.
Database software as “the stickiest software of all time”
They argue databases create exceptionally strong lock-in because data accumulates over time and sits at the center of business operations. Once a company builds around a database system, switching costs become uniquely high compared to many other software categories.
Data creation is exploding—and it’s hard to intuit exponential change
They contextualize database stickiness by pointing to the rapid acceleration of data produced and stored. Common “more data last year than the prior decade” style stats are invoked to emphasize the compounding nature of data growth.
The internet didn’t speed up as fast as data volumes grew
A key dynamic: bandwidth improvements haven’t kept pace with the growth in stored data. This widening gap makes moving large datasets increasingly impractical, intensifying lock-in to wherever the data already lives.
AWS Snowball: shipping data to the cloud instead of uploading it
They cite AWS re:Invent examples where customers can’t realistically upload petabytes/exabytes to AWS. Snowball is introduced as a secure, shipped storage device that customers load on-prem and send back to Amazon for ingestion.
Snowball evolves: more generations and even edge compute use cases
As data needs keep expanding, AWS iterates on Snowball with new generations and variants. Some models add compute capabilities for field/edge scenarios, showing how AWS extends the platform around data movement and processing.
AWS Snowmobile: “never underestimate the bandwidth of a semi-truck”
When even Snowball-sized transfers aren’t enough, AWS escalates to Snowmobile—effectively a truck-scale data migration solution. The segment underscores how extreme enterprise data sizes have become and why physical logistics can outperform the network.
Migration timelines reveal lock-in: even the best solutions can take months
They note that even with Snowmobile-like approaches, moving data can take on the order of months, whereas uploading over the internet could take far longer. This reinforces why hosting enterprise databases inside a specific cloud creates deep practical lock-in.
Amazon’s own Oracle migration: a case study in how hard switching is
Ben shares that Amazon started on Oracle databases and didn’t complete its migration off Oracle until 2019—long after AWS launched. The anecdote serves as proof that even a company with massive internal expertise struggles to unwind database dependencies.
AWS built many database options—yet migration remained difficult
They observe that by the time Amazon finished migrating, AWS already offered numerous database products, including both managed open-source options and internally invented systems. Despite this breadth, the internal migration still took years, underscoring stickiness.
What this means for AWS: massive revenue shift and durable power
They conclude that the size and stickiness of the database market implies substantial future revenue that can move to AWS—and once it moves, it tends to stay. The chapter ties the logistical realities of data transfer directly to AWS’s long-term strategic leverage.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome