Software ZKivision

For the technology life cycle stage conceptualisation, literature by Sommerlatte and Deschamps (1985), Specht and Möhrle (2002), Gerpott (2005) as well as Burgelman, Christensen and Wheelwright (2009) serves as the foundation. As mentioned, the expert interviews conducted in R1 provide the foundation for the assessment below, and are displayed in a condensed way here: Figure 5 below shows the resulting classification for each indicator of the technology life cycle stage, based on the condensed expert assessments (red boxes). Additionally, the blue box shows the resulting overall determination of the technology life cycle stage: RIMDB are in-between the growth and maturity stage. Therefore, RIMDB have already reached a medium level of exhaustion of their competitive potential. Below, each indicator assessment is described briefly.

Figure 5: Technology Life Cycle Stage, adapted from the lecture slides of Technology Management at TU Berlin and University of Potsdam, summer 2014 (based on: Sommerlatte & Deschamps, 1985; Specht &Möhrle, 2002; Gerpott, 2005;

Burgelman et al., 2009) Uncertainty of the technological performance

There are few problems with the technology RIMDB. Some occur with the software-stacks building on RIMDB though, such as enterprise-applications, which are not yet ready to use the technology to its full extent. The speed that RIMDB can provide is therefore often nullified, e.g. when additional legacy- software is used or data is transferred into an Excel-table due to lack of better gateways to the new database. There are areas where more research is necessary to ensure smooth performance of RIMDB, e.g. its high energy consumption of main memory. Overall, performance issues can be

considered rather minor regarding the overall technological performance of RIMDB, which is why the uncertainty of technological performance can be rated as close to low.

Degree of technological diffusion

The diffusion of RIMDB is still lower in comparison to other (mostly hard-disk-based) DB systems. According to the experts, RIMDB are rather expensive, hampering diffusion in more cost-sensitive application areas. Many companies have to carefully assess which data is administered in-memory and which is kept on a disk that is cheaper than memory storage. However, the trend towards price declination for main memory storage appears to continue, which would support future diffusion of RIMDB. Many application fields for RIMDB appear to be untapped at the moment, offering further diffusion potential. Based on this assessment, the degree of technological diffusion as of now can be considered low to medium.

Number and breadth of potential application fields

Practically any field where an entire set of stored data is used as the input for a query can be a potential application field for RIMDB. They include: reporting and business analytics, data-mining and machine-learning issues such as forecasting, text-mining, evaluation of medical study data, genome analysis, geo-data-analysis, targeting and tracking of users and advertisements, recommender systems, search engines, as well as traffic planning and monitoring. Although several of these application fields are in themselves established, the main advantage of RIMDB lies in the speed with which queries can be answered. Especially when it comes to more unstructured IoT and sensor data, there are also other technologies and players aiming to grasp market share. According to the experts, RIMDB may be able to gain more ground in application fields relating to ERP and CRM (strongly relating to the aforementioned field of business operations), and other business intelligence applications, as opposed to the field of IoT data. Availability of RIMDB through cloud solutions opens application fields for more cost-sensitive players (e.g. SMEs). In sum, the number and breadth of potential application fields can be considered as increasing/big, slightly tending towards established. Research and Development (R&D) intensity and development type

R&D is being conducted to further improve RIMDB performance and mitigate the performance issues mentioned above. Moreover, there are R&D activities concerning the application fields (especially when it comes to the implementation and best practices) and partly cost reduction (mostly regarding main memory storage). Overall, the intensity of R&D is medium (neither high nor low) as it concerns both applications and costs.

Development requirements to product maturity

The experts consider the RIMDB technology itself as relatively mature, as it was already implemented by several major vendors and the technological components are generally well understood. Further development requirements mainly concern further cost reduction and availability of compatible applications for the various fields where RIMDB can be used. The literature confirms this assessment,

since research has started in the 1980s and by 2011, several RIMDB were on the market (Loos et al., 2011: solidDB by IBM, TimesTen by Oracle, SAP HANA). Development requirements to product maturity are therefore medium to short-term (less than a year to three years).

Availability of the technology

As a software (albeit based on hardware, unless used in conjunction with a cloud system), RIMDB are available to a large number of potential clients worldwide and not subject to the availability restrictions of more hardware-based technologies. The main impediment to availability of the technology is the cost of implementation, which is why full mass market availability cannot be inferred quite yet. This issue may be mitigated in the future due to further cost decline in main memory and increased application of cloud-technology (reducing customer costs). Another issue could be the high energy consumption of main memory in some areas. Hence, the technology is not yet fully available to a mass market, but certainly in a market oriented fashion (moving close to mass market). Aside of the understanding of the technology RIMDB itself and its life cycle stage, an overview over the active players and the competitive field of RIMDB is also important for being able to prepare the scenario analysis and determine driving forces. The following section provides this overview.