We meet people everyday who are dealing with the challenge of setting up “Custom Analytics” for their company. Many of those professionals are very capable programmers, and when confronted with the data integration task, they very often ask themselves:

Should I write the code on my own and just “throw” all the data someplace, or should I use an ETL tool to do that?

In this article we will explain the reasons that companies use a tool for ETL rather than coding it on their own.

5 top reasons to use an ETL tool rather than “script your own”:

  1. Data Flow Management

    Using an ETL tool allows you to manage the different data flows using visual representation, and this is a huge help because:

  1. In many cases the data “travels” through several processes, getting enriched, aggregated and sometimes cleansed out. Having a graphical overview of the entire flow makes it much easier to understand the ‘big picture’ and ‘zoom in and out’ as needed.
  2. When seeing the flow and its components it is much easier to identify components of the ETL that are performing a very similar or identical action. In such cases using an ETL tool makes it possible to reuse the same logic (and update it once) instead of writing something similar again and again.
  3. Using a tool makes it easy to schedule the ETL jobs in a visual environment (as opposed to scripting in open source applications such as Crontab) to control the dependencies between the different components in the data flow.
  4. Using a tool makes it much easier to work as a team and transfer knowledge. Team members can easily understand the set up just by looking at the graphical representation of the data flow.
  5. Your top developer doesn't have to be in charge of making it ‘production ready’. It is possible to sign off to a lower tier professional - “NO BLACK BOXES”.
  1. Connectors

    The ETL tool has connectors optimized to data sources and targets, making the data transfer immediate. When using a custom connection, handling batch sizes, loops, file distributions and file compression is required every time there’s a new source or one of the sources is upgraded. All of these are handled automatically when using connectors.

  2. Transformations

    Modern ETL tools offer many transformations that can be used through a visual interface (drag and drop) instead of implementing them from scratch. For example: Implementing a complete Geo-Identification mechanism from scratch will take up quite a lot of time. Using a tool is likely to be quick and will limit potential bugs in the process.

  3. Investigation

    Often when a strange figure bubbles up in the visual layer, it is necessary to investigate where it originated from and how it was calculated. This is much easier and faster to do when the data flow is shown with an ETL tool.

  4. Professional Grade

    An ETL tool is designed as a ‘professional grade’ solution, which means that it is designed to scale up and handle large amounts of data without risking errors and performance hits due to its size. Reaching the same level of resilience with a custom made solution requires a very high level of expertise, significant development and QA effort, so it is likely to become more expensive than to “hire” a tool.

Data integration is an essential part of any BI project. Using a data integration tool, rather than writing the code yourself, allows you to take advantage of the development work done by the team of the ETL vendor, who focuses purely on ETL. This means that you can leverage their work and save time and money by having your top notch developers focus efforts on other things, and still not need to compromise the much needed flexibility and scalability that data integration cries out for.

To see how easy it is to integrate your data with Xplenty click here.