Many organizations think that, simply because they generate a lot of reports or have a lot of dashboards, they are data-driven. While those activities are part of what an organization does, they are usually backward looking. That is, they are often a statement of past or present events without much context, without a causal explanation of why something did or did not happen, and without recommendations on what to do next. In short, they state what happened but are not prescriptive. As such, they have limited upside.
On the other hand, prescriptive analytics, such as predictive models, involve answering questions related to what, who, when, where and why, making predictions, recommendations and telling a story around the findings. They are often a key factor in a data-driven organization. Those insights and recommendations, if acted upon, have great potential impact on the organization.
However, such insights require the right data to be collected, the data to be reliable, the analysis to be good, the insights to be considered in the decision and to drive concrete actions so that the potential can be realized. This process is known as the analytics value chain.