Written by Sebastian Frandsen, Lars Thomsen og Nikolaj Nielsen, students of the Digital Innovation and Management Programme at ITU. This blogpost details the use of digital methods by students doing their MA thesis at the ITU. We were very happy to be able to supply Sebastian and his group access to the ANTA tool and sparring about how to describe the tool in their work. Please contact the lab at ethos@itu.dk if you are interested in making use of the lab’s resources in a similar way.
The Financial Services industry is undergoing rapid changes in these years with new Financial Technology (FinTech) companies disrupting the market by provisioning financial services exclusively through online software solutions. FinTechs operate with much lower costs than the traditional banks that also have expenditures for brick-and-mortar branches and large workforces. Common for the FinTechs and the traditional banks is that they are expected to deliver transparent services, and their customers have increasingly high expectations to get more value for less money. A big difference though, lies in the fact that many traditional banks are haunted by complex legacy-filled application landscapes, consisting of hyper complex cohesive systems, which constrain their ability to react rapidly, and adapt to the new digital “playground”, as opposed to start-ups that rarely have prior systems to take into account. Our thesis investigate how a large financial service company can balance balance its inherent exploitation of rather rigid structures, with the need for agility and exploration when transforming their front-end solutions. This includes both the development of new systems, and the reuse of existing ones.
Digital tools: A Quali-Quantitative Exploration of Field Interconnectivity
Our dive into the realm of digital tools, started a cold and grey February day, fueled by a combination of too much coffee and frustrations. To set a solid theoretical foundation for our investigation of our case organization, we had chosen to conduct literature reviews for the fields of Information System Development, Enterprise Architecture, and Organizational Ambidexterity. The source of our frustrations stemmed from the challenge of tying together the findings from the three literature reviews. Throughout the literature reviews, we had a hunch about the fields’ interconnectivity, but lacked a systematic way to explore and explain them. The solution turned out to be a combination of ANTA (Actor Network Text Analyser) and the visualization tool Gephi.
The first step was to upload all our literature to ANTA, tag them according to their field origin and let ANTA run a semantic analysis. This resulted in “measly” 28.000 keywords, which had to be either filtered, combined, edited, or removed. As ANTA unfortunately lacks the cognitive capabilities to automatically merge keywords such as IT projects and ITprojects, this step required some legwork. However, the time spend on refining keywords turned out to be a worthwhile investment.
Next, the refined keywords from ANTA were exported into Gephi, where we spend quite a lot of time exploring the various visualizations, which could be created through different filters, rankings, etc. Eventually, we settled on two visualizations. The first visualization was explored from a macro perspective, to investigate the compositions of the fields and their intra and inter relations. The blue nodes represent the keywords extracted by ANTA and their respective size is determined by the amount of unique articles related to the given keyword. 
The composition of the fields is from an isolated viewpoint not very interesting. However, when combined with our knowledge from the literature review, we saw several key findings echoed in the visualization. An example is the field of Enterprise Architecture, where the literature often emphasized the need to establish a more common terminology. In the visualization the orange nodes are centered around the keyword Enterprise Architecture, with little interconnectivity amongst each other. We had similar findings for both Organizational Ambidexterity and Information System Development, where the network composition echoed literature review findings.
In the second visualization, we zoomed in from a macro to a micro perspective, to explore the intersection between the three fields and find keywords tying the fields together.
These keywords were not necessarily the most prominent in terms of size, but represented crucial commonality between the fields. Each keyword was explored individually to ensure it was connected to all three fields. The spawned a long list of concepts, which was later on grouped based on similarities. The keyword groupings were used as a facilitator for further analysis, where we combined the groupings with our findings from the literature reviews and an initial data collection.
Overall, the digital tools helped us structure and explore our existing knowledge in a different way to bridge the gap between the fields of Information System Development, Enterprise Architecture, and Organizational Ambidexterity. While, we only scratched the surface of the possibilities brought by digital tools, these two simple visualization provided a structured way of moving our research forward.