The media’s emphasis on big data over the past two years has resulted in something of a spending frenzy. Today, data analytics dominates enterprise technology spending. But interestingly, satisfaction with the results is less than stellar.
The CFO.com article “CFOs Frustrated with Return on FP&A Investments”cites a 2015 CEB survey of CFOs that reveals widespread dissatisfaction. According to the article, the primary factors behind the discontent fall into two areas:
- The quality of analysis (its timeliness, accuracy, comprehensiveness, volume of detail, and actionability)
- Access to data.
When embarking on a selection process, it’s easy to get caught up with the eye-catching graphs and dashboards in some demos. However, I advise to focus first on a tool’s architecture. Basically, all tool options fit into one of two architectural categories:
- Those that are cube-based (OLAP)
- Those built on relational databases (ROLAP)
There are pros and cons to each category. But, it’s important to understand the tradeoffs before you begin an in-depth selection process.
The following paragraphs provide a “Cliff Notes” summary of the differences between the two categories. Understanding the basics will help you and your team select the analytics tool that best meets your business needs and expectations.
CUBE-BASED TOOLS: FAST, BUT NOT SO FLEXIBLE
Tools based on multi-cube technology (also known as OLAP, which stands for online analytic processing) are designed specifically for multi-dimensional analysis and reporting.
Because cube-based tools pre-aggregate data, they are very fast at churning out results to queries. However, this advantage is also a potential drawback because it limits a tool’s flexibility for handling changes to data structure or complex financial logic. Essentially, every time a change is needed – such as adding a new hierarchy or dimension – you have to rebuild the cube or create a new one. This can be a time-consuming process and one that often requires IT support. For system performance reasons, OLAP solutions often require multiple linked cubes, creating a complex application that can be resource-intensive to manage and maintain.
Cube-based tools are excellent options for those companies with relatively simple business models. They are not as well-suited to companies with complex financial processes and finance rules or for those that must manage ongoing business change.
RELATIONAL DATABASE TOOLS: EXCELLENT FOR BUSINESS COMPLEXITY, BUT YOU’LL HAVE TO WAIT A LITTLE LONGER
A relational database tool organizes items as a set of tables from which data can be accessed and combined in many different ways without having to reorganize the database tables. These solutions are easy to modify or extend without rebuilding. For instance, a new data category (such as a line of business) can be added without requiring other system modifications.
Analytic solutions built on relational databases provide greater flexibility so they’re a good fit for those companies, especially mid-to-very-large enterprises, that must cope with ongoing business change, complicated financial logic, or very large volumes of data.
However, these tools are not designed to provide “sub-second” response times. In fact, the larger the database and the more complex its data relationships, the longer it will likely take to process and compile results from queries.
HOW TO GET THE BEST OF BOTH WORLDS
Many of our customers are large, global organizations with diverse subsidiaries and volatile business environments. Their analytics requirements are extremely complex, but they also expect timely performance. This is why Tagetik, which is built on a relational database, has connectors to cube-based tools such as the Qlik Analytics Platform, andMicrosoft Power BI.
These connectors create analytic cubes automatically so that customers can take advantage of the Tagetik’s relational database with its powerful, built-in financial intelligence while also getting the processing speed of cube-based analytic tools as needed. The combination provides the best of both worlds without the user ever knowing what is relational or OLAP.
THINK CROSSOVER
Selecting an analytics tool is a bit like shopping for a car. For those who put a high priority on driving agility and speed rather than running in-town errands, there are many performance cars to choose from. And for those who mainly use a vehicle for carpooling children to and from school or soccer practice, minivan models abound. But many consumers want something in the middle - a car that offers both everyday practicality along with driving performance. Hence the development of the crossover vehicle.
Think of financial applications and analytics in the same way. You want to provide your users with a “crossover” tool that can address complex analytics and business change while, at the same time, delivering processing speed.
Efficient and effective analytics are essential for the modern finance organization. When making an analytics tool selection, it is essential to make sure its architecture is flexible enough to meet all business needs – short-term and long-term – and doesn’t impose unnecessary constraints or limitations on the analytics needed to run a successful business.
By Marco Pierallini, from: http://www.tagetik.com/blog/authors/marco-pierallini/2016-09-finance-analytics#.V9xKzigrLIU