A:360 Podcast #15 – A Phased Approach to Analytics

Today’s A:360 discusses why it is highly recommended that most organizations take an iterative, phased approach to developing their analytics solution.

Watch and Listen

Click to Watch on YouTube.

Listen to the Podcast

Click to Listen on SoundCloud
Click to Listen on iTunes

Read the Transcribed Audio

Hey everyone. Welcome to today’s A:360. My name is Brewster Knowlton, and today we’re going to be talking about the importance of taking an iterative approach to building out your data and analytics program.

By an iterative approach to data and analytics, I really mean your organization should make a focused effort not to try to tackle and bite off everything all at once. Essentially, approach business intelligence or analytics in phases.

I hear a lot of talk about big data and getting to this point of unstructured data where technologies like Hadoop, Hive, and MongoDB are thrown around. Before worrying about those more advanced technologies, take it one step at a time. Crawl, walk, run.

Focus first on building out a data warehouse and identifying those high priority data sources (like a core or other important third-party data sources) that you need to integrate first. In this podcast we’re going to talk about the importance of taking that iterative approach and some of the dangers of going all in and trying to do too much at once as well.

Identifying ROI is something that a lot of executives are rightfully concerned about when it comes to analytics. Let’s assume you start by building out a data warehouse with a core and a few major third-party applications (these could be your loan origination system or a CRM system, for example).

Without investing all of the resources building out your analytics program up front, you can show incremental value to the more skeptical individuals within your organization with a phased approach. That’s not to say that those organizations that are going all in are taking a bad approach necessarily, but there are some organizations that have to take a step back and prove that ROI at each step of the development process. By taking an iterative, step by step approach to this, you can actually start to build up that scale, build up the ROI and start to incrementally show value without having to rely on one big initial burst after months of development and a larger up-front cost.

Taking a phased approach allows you to build out the skills of not only your BI team but the rest of the organization as well.

If you are going for a spoked-wheel model [of analytics] – where you have subject matter experts or power users within each department that are going to be responsible for some of the analytics in that area – you’re going to need time to build up their skills and train them on various things such as:

Analytics is certainly a learning process. By taking a phased approach to analytics, we can learn from our mistakes during each phase.

If we try to boil everything down to a single phase – especially for those organizations who are trying to build their data warehouse/analytics platform in-house (I would probably urge you to take a step back and reconsider if that’s the best approach) – you’ll likely wish you adopted an iterative development approach. If you make a critical mistake in the first phase and continue to make that same mistake (because you haven’t broken the project down into phases where you’ve learned from your mistake), you’re going to create a very, very difficult rats nest to unravel when you realize the mistake later on. As a result, you’ll have to go back through potentially every single phase and make changes.

By taking an iterative approach, you might make a mistake in phase one, but you can correct that mistake for only that phase’s work. Then, when you go onto the next phase, you’ll have learned from the prior phase and will be able to avoid making the same mistake(s).

Analytics is best handled with an iterative, phased approach. Break it down into phases and don’t try to bite off too much at once. This approach allows you to show incremental value, allows you to properly develop and cultivate the necessary skills, and it allows you to correct mistakes that may arise with only minimal issues or rework.

That’s it for today. Thanks again for listening to today’s A:360.


Subscribe to have new content sent directly to your email!

Photo Credit

Posted in Analytics, Banks, Credit Unions, DW/BI.

Leave a Reply

Your email address will not be published. Required fields are marked *