Five Maturity Levels in Data-Driven Organizations
Data is a critical competitive factor and is becoming more and more crucial for achieving success. In theory, hardly anyone disputes this to any extent. But in practice there are some very different levels of maturity when it comes to handling data. Based on practical experience gained in medium-sized companies in particular, I would like to try to describe five typical levels of maturity. This is intended to provide a helping hand on the way to becoming a data-driven organization.
The focus of the following maturity levels is on how data is handled. This is related to the topic of Business Intelligence (BI), but it is not identical. In my opinion it emphasizes something different, something more modern. Various maturity models have been put forward in the context of BI (including those by Gartner and TDWI), but these frequently focus on developing from a report-oriented organization to advanced analytics (taking an expert look into the future).
There is a wide range of systems, each of which administers data, but they are not extensively interlinked with each other. Typically, there is one, or several ERP systems as well as a variety of “tools”, which increasingly include solutions in the cloud. Each of the systems usually provides isolated reporting, data is often interlinked by hand using Excel and evaluations demand a large amount of manual work. Data silos have emerged and knowledge is fenced off in “territories”. Evaluations are primarily retrospective. The company finds it hard to use data to help them take a competent look into the future.
The data landscape comprises data islands – bridges have to be built by hand.
Based on the islands described above, individual bridges have been built which connect the data in various systems. This means there are several comprehensive reports, but answers to questions based on data are only available to a limited extent (and mostly only for just a few users). Excel is still considered to be significant for evaluations. There are redundancies in the data, inconsistencies and frequently it is necessary to rework the data retrospectively due to insufficient data quality. Reports can be conflicting depending on who creates them. Any bridges are usually programmed by hand and are fragile.
The data landscape comprises islands which are networked via fragile data transportation.
A central data warehouse has been established which data from all relevant systems flows into (including data in the cloud), in addition to the ERP, thus providing a secure up-to-date data inventory for the most important aspects of the company. On the whole, access is strictly regulated, some standardized reports are available on the basis of the data warehouse. For certain questions, users are provided with selected parts of the data inventory, for example via a data mart. Work is in progress on improving data quality. This maturity level correlates to the traditional understanding of business intelligence.
In this data landscape all the islands are connected with a central data warehouse which the data is transferred to.
