Call us at 650-400-3029 (PST)

How To Fix The Broken Links In The Analytics Value Chain

The folks over at ZS Associates sponsored a study by the Economist Information Unit on analytics titled “Broken Links: Why analytics investments have yet to pay off”.

70% of respondents say analytics is already “very” or “extremely important” to their own business’s competitive advantage. In just two years, 89% of respondents expect this to be the case.

Just 2% believe transformations in capabilities at their companies have had a “broad, positive impact.”

Here we see the classic challenge of analytics – 70% think it is very or extremely important but only 2% say their analytics efforts have a broad, positive impact. And like every other assessment of this problem, including ones we have conducted here at Decision Management Solutions, the problem is not in the data or in the analytics themselves! The problems are in framing the problem and operationalizing the results. Everyone seems to have the analytic technology they need but they just aren’t getting it to work for them.

The EIU study did a really nice job of showing where the problem is. It outlined a set of steps and identified the area where most challenges were identified:


And this shows clearly what they mean by broken links – the links at the beginning and end are where the most problems occur.

The first broken link is in the framing and setting up of the problem to be solved. It turns out to be really easy to build great analytic solutions to problems you don’t have. This comes from a failure to really describe the business problem. Most analytic teams will identify the business metrics they want to improved and then dive straight into the data. Unfortunately, analytics cannot improve metrics, at least not directly. You need something to connect the one to the other and that something is a set of decisions that, if improved, will improve results.

The other broken link is the deployment of these analytics. It’s remarkable how many analytic models get developed and then never used, or at least not used in anything approaching a reasonable time. As the survey results on a recent International Institute for Analytics webinar showed, 60% of projects failed to act on their analytics in “months”. If you can’t prescribe an action to take based on your analytics, and can’t change your organization to make these analytic actions real, then the analytics you build won’t help.

Companies in the survey see taking action and managing change as the best opportunity to improve closely followed by problem definition. In the follow up blog posts to these series, I’ll cover Framing Analytics with Decision Modeling to address business understanding and problem definition and Operationalizing Analytics with Decision Modeling to address deployment.

Read the Rest of the Analytics Value Chain Series