Of truffles and truffle pigs – Seven steps to data-based value creation

Much like the search for truffles in a forest, the pursuit of data-based value creation is a particularly challenging undertaking without the necessary knowledge and appropriate skill set. The following seven steps illustrate how the hunt for “truffles” in the area of data-based value creation can be successfully launched.

 1. Establishing the data project under direct supervision of top management

Data accumulates in many areas of a company and is thus also in located in different corporate silos with often only limited data exchange between these silos. However, in order to make the most of the significant potential data project offer, it is often necessary to access relevant data from several subareas of an enterprise. With top management acting as «connecting force» between these corporate silos, it also possesses unique access to information from various different areas of the enterprise. Given this constellation, it is highly advisable to establish data projects under direct supervision of top management.

 2. Finding the «truffle pig»

The detection of “truffles”, in this case the identification of potentials for data-based value creation within an organization, often poses significant challenges for companies. Whenever possible, the responsibility for identifying such potentials should be assigned to an individual that possesses both sound expertise in the field of data science as well as business aptitude. Since data-based value creation is strongly driven by data technology on the one hand but also involves important business factors on the other hand, the most promising path to identifying these potentials comprises an integrative understanding of these two respective fields.

3. Planning a route through the company

The next step consists in identifying the specific potential for corporate data projects. This is to be achieved best by holding brainstorming sessions in various company units. In order to be able to identify potential use cases as well as possible, the exchange with employees who perform specific functional tasks and possess corresponding expert knowledge is indispensable, as those who are close to specific functional tasks can best assess potential use cases of data-based value creation in their area of responsibility.

4. Conducting brainstorming sessions

Once the relevant functional experts within the company have been identified, the brainstorming sessions should be conducted in the respective departments. The aim is to determine in which corporate areas AI could find a concrete application and in which environment data-based value creation can be generated. Does this preferably take place in the context of text, image or audio processing? Or rather in the monitoring of processes or in the generation of forecasts?

5. Specifying specific data and technical requirements

Once potentials for data-based value creation in the company have been identified, further specific data and technical requirements regarding the data project must be clarified: Is the quality of the existing data sufficient? What are the relevant metrics? Is a new data infrastructure necessary? During this step, close cooperation with the functional experts from the respective company units is once again essential. Since these individuals work with the relevant data on a daily basis, their expertise enables them to quickly give the project a specific focus and thus significantly accelerate the project progress.

6. Identifying «quick wins»

The next step’s goal is to consolidate and prioritize the use cases generated during the brainstorming sessions. At this stage, a decision over which projects corporate resources should first be invested for further planning and subsequent implementation should be made. Typically, those data projects which provide a high value added whilst simultaneously offering relatively fast and easy implementation are tackled first and are therefore considered as so-called “quick wins”.

7. Initiating the project implementation

Consideration should be given to the question whether the data project are to be carried out in-house or with the support of an external partner. Companies often use external partners for initial data projects. If these initial data projects prove successful, later data projects are often launched and implemented by the company itself.

These seven steps are intended to provide a first orientation for the identification of potential “truffles” in the area of corporate data-based value creation. In order to promote a data-oriented corporate culture as well as a deeper insight into the topic of data projects, the Executive School of the University of St. Gallen is offering the course «KI-Projekte erfolgreich managen» in late summer 2020. Further information relating hereto can be found on the Executive School’s homepage: https://es.unisg.ch/programme/ki-kuenstliche-intelligenz-projekte-erfolgreich-managen.

The content of this article is based on the webinar «KI: In sieben Schritten erfolgreich Datenprojekte umsetzen», held by Prof. Johannes Binswanger, Professor of Business Administration and Economic Policy at the University of St. Gallen & Marc Schöni, Cloud Solution Architect at Microsoft Switzerland. The webinar is available under this link.

Illustrated by Julio Prina