Overview
In most companies, Business users can access data only through the intermediation of IT department. So IT department is more the gatekeeper of the data rather than the enabler. This does not stem from the desire to control information but from the necessity to make it easily accessible: complex BI tools, continuously- evolving technological landscape, disparate and disconnected data sources made the data management extremely complex and time consuming. Such complexity can only be managed by an IT-centric data management model. But the price was lack in speed and flexibility, heavy processes and a relatively-limited number of reports.
But, the market shift to digital services and the transformation into digital companies is not anymore compatible with this mode: the business models of a digital company is based on analytics and the need of such capability can’t be satisfied only the IT department:
- Today, Companies collect “all” data, not only the “valuable” ones; consequently the number of heterogeneous data sources is growing exponentially, making data integration extremely complex. A single, integrated, data model containing all internal (and external) data sources is an almost impossible mission.
- The massive use of analytics to create competitive advantage is multiplying the number of data use cases. DWHs are no longer the only big systems consuming data. In a digital company, data is consumed by a plethora of new systems. The centralized development of analytic solutions in a single department is not viable solution, especially due the fast time-to-market of digital services.
- The extended use of analytics to create competitive advantage has made the production of analytics even more complex and highly-skilled-resource consuming. Today, data is not only use to create decision-support reports but also to predict customers` and environment behaviors, to find hidden patterns, and other predictive analytics capabilities. Analytical skills are required to support these initiatives.
- Last but not least, digital services are characterized by an amazing fast time-to-market: new services are continuously launched in the market to test consumer’s reactions. If positive the services is engineered, if not if retired. This model implies a quick (but robust and secure) prototyping of the new services; a model not compatible with the traditional Business-IT application development cycle.
So any strategy where data is the competitive differentiator puts a tremendous pressure on IT BI departments, as data and insight on data must be quickly and easily available to practically any company users.
So company data management needs a new model, where IT become a data and service enabler. This new model – as it refers to the idea that anyone can access data without the inter mediation of IT teams – is often called “data democratization”; even if the data self-service is just a part of the whole new model.
The path to IT data-enabler model
In important to highlight that Digital transformation not only requires IT providing quick and easy access to data to the whole company. IT BI and analytical teams must also to support the development and the maintenance of the new analytics solutions and last but not least they have to provide flexible solutions for storage and analytic tools on demand. So IT BI department must move from a strategy based around data management (or data-keeper) to a new strategy where IT is the enabler of data and analytics services.
Said that, the trajectory of data and analytics democratization is clear: IT must automate traditional BI tasks as much as possible to free up resources that should be allocated on the new required capabilities such analytics and cloud. Here the main elements of the data democratization:
- Self-Service data provisioning platforms allow business users to easily access data without IT intervention. Users should be able to create the needed reports and analytics by themselves. IT intervention should be limited only to complex situations. Obviously data security and privacy are fundamental to enable such function.
- Automatic data catalog creation and update. In order to empower business user, stored data stored must be available and easily searchable by anyone. Without a complete and constantly updated data catalog, data is not accessible or usable. Data description and dictionary are mandatory to enable the data democratization. Source solution analyst intervention should be limited to complex cases; users should mostly be able to identify the data they need by themselves.
- Standardize cloud approach for analytics tools and data storage. To use the data and create their analytic use cases, Users need also storage and analytic tools, not only data. IT dept. must provide such capabilities in a flexible way: additional storage areas, new analytics tools, etc should be available to the users as easily as the data. Software as a service model is the right answer to these fluctuating request. Moreover, public cloud models should be evaluated as possible solution to host BI and analytic applications.
- Highly- skilled analytics resources to support users in complex analytics model creation and maintenance. IT analytical skills will be fundamental to support business users in creating complex Bi and analytics use cases.
- Automate and differentiate operations basing onBusiness importance. In the digital service world, some analytical applications are mission-critical, so operations must be differentiated based on revenue importance (i.e. interactive recommendation engine should be 24×7 mission-critical as they generate revenues; Financial BI reports are critical from quality point of view but they are not 24×7; marketing reports are important but not mission critical).
The picture below represents the shift of IT resources in next couple of years.
Conclusion: the advantages of data democratization
The road to the transformation into digital companies goes through the data democratization: without it, companies will be not able to efficiently leverage their data to provide digital services. Thanks to the data democratization, companies will achieve the following benefits:
- Agility: users will not need to wait IT specialists to provide them data, reports or insight. Users shall develop their own analytics systems by them-self.
- Resource bandwidth: self-service data systems will release IT resources which should be redirect to develop complex analytics solutions.
- Fast launch of new services: self-extracted data can be used for quick prototyping and launch of new services. Campaign’s feedback can be quickly reanalyzed to refining the marketing campaign.
- Data-set expansion. Self-service data not only allows to easy extract data, but it also promote data sharing and collaboration. New data set created by one user (i.e subscriber data profiling) can be used by other users too.
In conclusion, the self-service data and analytics is a key factor of the success of Digital companies