Whether you call it business intelligence, analytics, or reporting, BI has become an indispensable tool for modern businesses. It’s safe to assume that BI, in some form, is in use at every major company in the world. What’s even more amazing is the fact that BI as we know it has only been around since the 90s. Here’s a rundown of the history of BI, along with a look at the future.
1.0 Full Stack, On-Premise Solutions (90s – Early 2000s)
In the beginning, business intelligence was expensive, hard to maintain, and even harder to use. The earliest examples of BI as we know it had their heyday in the 90s and early 2000s. Companies like Cognos, Hyperion, and Netezza created all-in-one systems that gave organizations the ability to run basic analytics on large amounts of data.
These systems were completely on-premise. An army of engineers, technicians, and IT staff maintained all the hardware, software, and data on-site. Like the earliest computers, early BI systems were large, finicky, and somewhat impractical. Another challenge to these early BI systems was the amount of work it took to find answers. One couldn’t simply ask a question, they had to engineer the question. It could take a team of engineers months before they had everything set up. At that point, it could deliver quick answers, but the amount of preparation and sheer labor that went into each question was prohibitive. Companies had no choice but to focus on core KPIs.
The challenges and costs of these systems meant they were only available to the world’s biggest companies. However, they did something nothing else could do. If you wanted BI, they were the only game in town.
2.0 Open-source, DIY, On-Premise Solutions (2000s)
The next generation of business intelligence was much like the first, except open-source. One could run it on their own hardware using open-source software. This meant maintenance required fewer people with fewer skills. You still had to maintain your hardware, but you didn’t need to pay for a special box or an army of technicians. For most of the 2000s, technologies like Pentaho and Microsoft SQL Server Reporting Service were the most common approach to BI.
These technologies were more flexible and easier to maintain, but their results weren’t a vast improvement on older forms of BI. They were simply cheaper and more accessible than traditional full-stack solutions. They still didn’t make data discoverable. In other words, you still needed an engineer to figure out a way to return the results you wanted. There was a limit to what you could ask. Still, the open-source approach lowered costs significantly, and more companies could use BI than ever before.
3.0 Virtualized Servers (2011 – Present)
The next evolution of BI moved hardware into the cloud. Since around 2011, the virtualized approach to hardware has dominated the world of BI. Virtualization made BI accessible to more businesses, because it lowered costs. For one, you didn’t need to purchase and maintain hardware anymore. Instead, you could rent cloud servers from a company like Amazon. At this point, BI started to go mainstream. A tool once reserved for elite businesses became accessible to companies that didn’t even own servers. BI became cheaper than ever, but it still required a team to make use of it. They didn’t have to maintain servers, but they still needed special skills to helm the software.
As BI became more accessible on the financial end, a few companies popped up to help make it accessible from a usability perspective. A company called Hortonworks had some success packaging BI components, which reduced implementation times and requirements. The demand for a user-friendly approach to analytics was clear. People wanted the power of a full-fledged BI tool, but with the usability of tools like Google Analytics. During this era of BI, that gap began to close and methodologies began to merge. BI became friendlier and easier to use.
4.0 BI-as-a-Service, No Servers (2016 – ???)
The next era of business intelligence is already under way, but because it’s so new, companies are battling over its definition. Everyone has a different idea of what the next phase of BI looks like, but there’s an overarching trend towards friendlier, more accessible tools.
There’s a good chance the next era of BI won’t use servers at all, not even virtual servers. Technologies like Amazon Athena have proven that serverless data querying is possible, even at a large scale. Serverless technologies are still new, but their potential is clear. iCharts, for example, essentially analyzes data where it lies. There’s no need to move the data or duplicate it onto a dedicated analytics environment. This way, data from your business tools stays in your business tools.
Another trend is a move to an “as-a-service” model. Just about every other business tool is available as a service, and BI is no different. The benefits of owning and maintaining your own hardware are disappearing. Unless you run a business that needs the absolute fastest in analytics, there’s little need for an on-prem solution. Even virtualized servers may be on their way out. They simply get rid of physical maintenance costs in exchange for subscription costs. You still need to pay someone to maintain your servers on the software side. Now that serverless queries are a viable option, they just might change the way analytics works.
Past, Present, and Future
When you think about it, the entire history of BI has been a shift towards something that’s less expensive, easier to use, and more powerful. When BI began, only certain companies could afford it. Before long, it became more affordable, but only certain people had the skills to use it. Today, the wall between data analysts and regular employees has started to erode.
Not long ago, computers themselves were limited to the biggest companies and the brightest people. At the time, no one imagined everyone would have a computer at their desk one day. Today, business intelligence technology finds itself in a similar place. More companies have realized that even the most advanced technology can only go so far when you keep it inside a boardroom. An advanced insight in the hands of a CEO can be powerful, but basic insights in the hands of every employee can be game changing.
For more information on the next generation of business intelligence, check out our resources page.