It’s always satisfying to challenge new soundbites, especially when they mask the inadequacies that preceded them.
Similarly, it would be easy to believe that big database companies like Oracle are leading the renaissance of their own industry. It just wouldn’t be true.
Before we delve into the expected “our technology is better than their technology” debate, let’s remember why the argument itself is worth having. The short answer is that the entire global economy is built on technology. Every consumer served, job created, and resource managed relies on the intersection of business and technology.
The slightly longer answer is that we are still playing catch up from the 1990s and early 2000s. During this period, despite obvious signs that data was about to explode in relevance, database technology was void of meaningful innovation. It was a “sticky,” high-margin business for a few established market leaders in the segment. Businesses were paying a lot for the technology, which was based on 1970s architecture and was ultimately the slowest part of the technology infrastructure in most companies.
Why did that matter?
Because businesses were moving too slowly as a result. CEOs were waiting weeks to get binders about their own business performance. Chief financial officers were taking weeks to do routine accounting procedures. Marketers and sales leaders, despite increasing online activity from their consumers, were systemically unable to use that information in a secure fashion to better serve their consumers.
Data was becoming the new steel (or oil, or fuel, or gold or whatever your preference), but nobody in the database business was doing anything to solve for the serious shortcomings of their legacy products.
That’s why market forces took the database through something of a renaissance. This was initiated not in California conference rooms, but on a small academic campus called the Hasso Plattner Institute in Potsdam, Germany. With the faster pace of innovation, we’re already at another inflection point. Two forces have emerged to guide this new future: the champions of the old status quo and the intellectual forces that disrupted it.
It is clear where we should look for a credible perspective.
Size and speed matter: the lessons of the in-memory revolution
The argument was simple enough: if we aspired to a digital economy, we weren’t going to build one without putting data into main memory. From the perspective of business customers who were exhausted by high-cost, low-value databases, this was a welcome argument as it ultimately enabled a massive reduction in their total cost of operations.
From a computing perspective, the in-memory architecture sparked a fundamental reinvention of data management. It enabled translytical processing, fundamentally simplifying the database design, eliminating data duplication between OLTP and OLAP systems, and minimizing unnecessary data movement and processes.
As Hasso Plattner himself has said, if you want to understand this transformation, look at the business outcomes more than just the technology arguments.
CFOs accelerated: Waberer’s International cut month-end close by 60%. CIOs accelerated: Farys reduced database size by 7x while increasing live reporting by 50%. COOs accelerated: Kolon reduced data storage by 75% and increased supply chain transactions and reports by up to 205x. There are countless other examples like VMware, which simplified IT infrastructure and centralized seven legacy data warehouses and 15 terabytes of data into one, 6TB database.
This explains why analyst firms, like Forrester, now acknowledge the shifting center of gravity in data management. Just this year, it released the first-ever competitive analysis of translytical databases.
Legacy is still legacy
The impressive track record of the in-memory era has caused a flurry of new marketing activity around “autonomous databases.” Let’s talk about so-called self-running systems.
It’s critical to remember that database industry legacy players are still stuck rationalizing or compensating for the layers of old technology that plague their offerings. To wit, while the idea of “autonomous database” has been in progressive data management roadmaps for years, for some the term itself is just a bandage to cover up for an aging product.
This isn’t really innovation; certainly not on the scale of in-memory. Just using the term “autonomous database” doesn’t get rid of the 50% of the legacy database, which even in 2018 is still indexes and aggregates, the cruft of an architecture from the late 1970s. “Autonomous database” doesn’t allow business objects to be simplified from 27 tables to two for thousands of enterprise customers at scale, while increasing throughput 8X on the same hardware.
Want to learn about SAP’s approach to in-memory computing and data management systems and what this mean for the world of business? Then check out this episode of InsideAnalysis to find out. Host Eric Kavanagh (@eric_kavanagh) will interview me on Monday May 14 at 3:00 p.m. ET.
Please be sure to sign up for the complimentary IDC maturity model, which shows the stage of your company’s innovation relative to its database. I welcome any comments or feedback as well. Leave a note in the comments section, or tweet me at @McStravickGreg