By Dale Kim
The need for immediate access to data is a popular theme in business today. The most successful companies will be those that react quickly to changing environments. One of the biggest challenges organizations face today is ensuring that many different types of users get access to granular data in a timely manner in order to perform analytics on that data.
Gone are the days when day-old, aggregated data was sufficient for making critical business decisions. This is why the notion of agile business intelligence (BI) is becoming top of mind. Agile BI allows data consumers to analyze all levels of data without the typical delays found in a traditional environment. Users don’t want to, or can’t, endure delays due to slow processes or technology limitations. They also don’t want to have to spell out exactly what they need up front prior to obtaining access to the data. They want the flexibility to explore a variety of data sets including ones they did not anticipate a need for–to get the insights to let them be successful in their job.
Can You Perform Self-Serve Analysis? You Can Enable Agile BI
A key enabler of agile BI is the availability of interactive, self-service analytical capabilities on enterprise-wide data. End users should be able to ask any query on the data without continually turning to IT for changes to the data sets. Agile BI is not only about what users can do once they have data access, it is also about how quickly users can access all the data.
In the past, “self-service” was a misnomer. Business analysts only attained freedom in bursts. While they enjoyed self-service once they were given access to data, the route to that point was far from fast. IT teams spent hours, days, and even weeks behind the scenes to prepare data for use by analysts. That problem today is exacerbated as companies generate more data. The data management effort grows with the data, but companies often don’t or can’t increase headcount to keep up with the greater IT requirements of mounting data. This means that IT teams are further burdened by big data challenges, resulting in significant delays for business users.
Traditional BI Tools Aren’t Flexible
Traditional BI tools simply can’t deliver the modern expectations of agility. These tools can’t provide BI for all types of users and typically require significant IT intervention to make data ready for analysis. This is largely due to the scale limitations, which force the use of extracts that only summarize data. When analysts need more details, they need more modeling and data movement. Typically, if end users require new analytics reports, they’ll need to request changes to the data from their swamped IT department, leaving them waiting days or weeks to run analytics. Users also find it difficult to collaborate using these traditional BI solutions. With the limited data found in extracts, there is less opportunity for exploratory collaboration that leads to new insights. Their interactions are typically limited to email, phone, or chat without the benefit of visually based data sharing.
To Achieve Agile BI, Plan Ahead With These Four Infrastructure Suggestions
If you expect business agility in your BI practices, plan ahead for an architecture using an agile BI platform that provides minimal data movement, direct access to granular data, scalability, and collaboration.
Minimal data movement. In traditional BI environments, analysts could only ask new types of questions once the IT team prepared the data for them. This entailed moving data to another repository with intensive data transformations, which added significant delays and risk of error. If instead, your BI platform did not need significant up-front modeling that limited the types of queries you could run, then you could reduce IT dependencies and thus speed up the time to insight.
Direct access to granular data. Related to the above, users have historically been given an extracted summary of data sets rather than the entire data set. This was done to boost query speeds on specific types/sets of queries. The problem is that if users needed to drill down into details or run different types of queries, they would have to send a new request to the IT team. If you can avoid the extraction steps and otherwise boost performance on the direct access to granular data, users can run analysis on aggregated data as well as the underlying detailed data.
As more data and users are brought into the system, scalability is critical. Modern BI platforms need to be able to scale to handle a new load in an efficient way. With the massive increases in data volumes, scaling requires a scale-out architecture on commodity hardware, especially when cost-effectiveness is a factor.
Business analysts often run analytics on a standalone desktop tool like Microsoft Excel, which makes collaboration and sharing more difficult. Instead, browser-based applications built on your BI platform enable interaction and ongoing insights among your user community. The ability to build and deploy analytical applications on your BI platform is an important capability for finding insights as a team.
Modern BI Tools
Traditional BI technologies are not well suited for modern environments requiring faster analytic access to detailed data. That’s why a host of new technologies are emerging as faster, more scalable, and more flexible options to the traditional technologies. SW
Dale Kim is the senior director of products/solutions at Arcadia Data. His background includes a variety of technical and management roles at information technology companies. While Kim’s experience includes work with relational databases, much of his career pertains to nonrelational data in the areas of search, content management, NoSQL, and Hadoop/Spark, and includes senior roles in technical marketing, sales engineering, and support engineering. Kim holds an MBA from Santa Clara University, and a BA in computer science from Berkeley.