Cloud computing is clearly one of the hottest technology trends. What started out as simple social media solutions (such as Facebook, YouTube, and Gmail) grew into departmental solutions (such as Customer Relationship Management) and then into full operational ERP system solutions. There’s an obvious appeal to hearing that you no longer need to worry about hardware or upgrades and that approved users can access the solution from any web enabled device. But where does a company’s BI strategy fit into the Cloud? As with anything in technology, it depends a lot on your business needs and existing technology landscape. This article discusses key points to consider when assessing your company’s BI roadmap and its potential migration towards the Cloud.
There are now several Cloud BI solutions offered on the market. The marketing materials for all of the options show how easy it is to create nice looking dashboard and report content, describe the speed to implement, and discuss the benefits of not having to buy or maintain the required infrastructure. While the nice visuals and promises of reduced costs will get upper management excited, having been in the data warehouse and BI industry for over 20 years, it is my experience that understanding details about the source data, and the required extract, transform, and load (ETL) to support the business reporting and analytical needs are critical in evaluating where Cloud BI fits into your BI roadmap. Here are some points to consider:
- Do you have an existing “on premise” data warehouse? If yes, you have to assess the reasonableness of migrating all of this content into the Cloud and assess the extract, transform, and load (ETL) strategies of keeping it refreshed in a timely manner.
- Is the majority of your source data in traditional on premise systems, or in Cloud solutions?
- If the majority of source data is on premise, then an ETL strategy will be required to upload data over the web into the Cloud; this data upload process will not have all of the options available that on premise solutions have for improved load performance.
- If the majority of source data is already in other Cloud solutions, then creating ETL processes between Cloud solutions can be a streamlined process that will work well.
- How high are the incremental data volumes? There will be extra time required to move data from on premise into the Cloud. The volume of data to move determines if this is a major factor or not. Besides the speed component related to loading the data, there are still some considerations as to how much data some Cloud BI solutions can hold.
- How close to real-time do the users require the data? While on premise BI solutions can be used to tactically point to operational systems, Cloud BI solutions need the data loaded into the Cloud. This is a factor when determining what reporting strategies fit best in the Cloud.
Clearly, understanding the data requirements is critical to any BI strategy; there is now a bit more emphasis on this area when looking at Cloud BI. Continued improvements to high speed internet access will eventually minimize or negate the data ETL considerations presented in this article. Likewise, as more of a company’s technology solutions are moved to Cloud based solutions, the ETL considerations presented here will be lessened. Cloud BI offers a lot of benefits, and vendors continue to invest heavily into this area. Assessing the underlying data strategy is the hidden component that all IT teams need to consider when creating or updating their BI roadmap.