The ROI of BI

Something that has been bothering me for a while is how - as a data enthusiast - I communicate the ROI (Return on Investment) of BI (Business Intelligence). People struggle to justify why they should invest in a BI project because it's very hard to draw a line between a report and a financial benefit. Gut instinct tells us it's useful to have this information as we can understand our environment better with data. After all, nobody would drive a car without at least a windscreen to allow them to see the outside world.

I'm going to explain below, using the Gartner Information Management Maturity Model as a guideline as to how the ROI builds up as you elevate yourself to the next level:

Level 1: Unaware

In level one, everything is ad-hoc. There is no consistency in reporting approaches. Effort is duplicated as people re-run reports in their own style. It's an inefficient and inconsistent approach. The cost to the business here is that:

• They can't confidently see how they are doing.
• They are open to risk due to reporting errors.
• Too much time is spent building reports.

At this stage ROI is achieved primarily by time savings. If you have an analyst building reports in an ad-hoc basis, you're just burning human resources. By standardising some of this in a BI system, your ROI is easily expressed in terms of time saved. It's the easiest stage to work out the ROI of BI in. In motor vehicle terms, this is washing your windshield so you can actually see where you are going.

Level 2: Opportunistic

By level 2 there is some BI in place. Issues faced in level one are reduced, but not eliminated. Some processes are standardised but there's not much joined up thinking across teams. Finance do their thing, Sales do theirs.... and so on. Business costs here come from:

• Conflicting views on how the business is doing.
• Duplication of effort in data processing.
• Low report usage outside of experts.

At this stage, uplifting your BI to a more business wide approach delivers ROI through reduction of duplication of effort and alignment of business metrics. So there is a tangible ROI from time savings, and an intangible one from being able to have a 360 view of the business. In car terms, this intangible benefit is akin to having some dials on your dashboard so you can confidently identify that you are obeying the speed limit and not redlining your engine.

Level 3: Enterprise

By level 3 there is unified BI in place. As a business you now have a clearer idea of where you are going, at least at an exec level. The gaps in your maturity drive costs in:

• Frontline teams don't have the data they need
• Low data usage outside of managers and experts

At this stage, uplifting your BI to an enterprise wide approach delivers ROI through broader use of data. So the line to benefit gets less clear. The intangible benefit is from everyone having a view of their part of the business. In car terms, this intangible benefit is the parking sensors or cameras. At a very detailed level you can see where your car is and where you need to move it to avoid issues.

Level 4: Standards

By level 4 there is BI everywhere. As a business you now are well informed about where you are going at all levels. Any further investment in BI and Analytics is now more leaning towards the predictive (what's going to happen) and the deeper analysis of why events happened. The cost to the business now lies in opportunity cost:

• The business doesn't know why some events occurred.
• The business doesn't have guides as to what might happen next.

At this stage, uplifting your BI to cover more analytical use cases starts becoming much more intangible. The benefit is from having a deeper understanding of your business. In car terms, this is your driver assist aids, such as adaptive cruise control. You could live without these, but your driving experience is an intangible amount better because you have them.

Summary & Advice

Part of the challenge with getting BI projects off the ground is determining the ROI. As the business matures it becomes harder to draw a straight line between your investment and the benefits.

This problem can be addressed in part by tying your Data Strategy to your Corporate Strategy and metrics. Instead of building a report because "it feels useful" you can focus. Build a report that shows how product sales grow in a region to support how you invest in another region. Create a data asset that helps you understand your asset health as part of a strategic initiative to reduce maintenance costs.

I'd love to hear your thoughts,

James Beresford

CEO - Talos - Efficient Decision Making through Automation and Analytics