While deploying data and data analytics is expected to yield new value, it is not without its challenges. The most important challenges (as counted in Table 4 and illustrated by the number of links between the concepts in Box 8) are related to privacy (which is related both to actual challenges regarding compliance with legislation and to perceptions of privacy), organisational readiness and awareness of the potential of data for new services, the actual process of successfully
providing new and relevant insights, and data quality.
Table 4: Challenges mentioned by the organisations that take part in this study
Category of challenges
Types of challenges Number of
organisations Privacy Privacy, data protection, use of personal data 18 Organisational
readiness
Awareness, organisational culture, and organisational skills
18
Data quality Data quality 13
Value mechanisms Link with value propositions, (new) business models 12 Technology Data integration, standardization, legacy systems 12 User involvement User participation, interaction with citizens 6
A lot of the datasets on liveability in the city contain personal data which means privacy is a first major challenge, especially when multiple datasets are integrated and used for new purposes. The Early Warning System in Almere, for instance, tries to collaborate with the Dutch Data Protection Authority (CBP) as it treads a fine line by combining several datasets to create profiles on a six digit postal code area level, in which it also looks at demographic characteristics. It illustrates that at its core one of the qualities of big data is that it can provide information with a high granularity (creating profiles that hoover somewhere between the level of individuals, demographics, social groups and street level).
BOX 8: Main challenges cited in the interviews
Another aspect of data-driven innovation and big data is that combinations of datasets may lead to new insights. However, there are legislative restrictions – designed to protect the privacy of citizens – regarding personal data or the combination of datasets that could lead to information that can be used identify citizens. One of the interviewees described how this makes it difficult to ask for consent from users to use their data to develop a new service based on the use of data, because nor the organisations aiming to process personal data, nor the users cannot properly determine the information that can be obtained by combining data sets. And privacy is not only a matter of legislation. Several interviewees
emphasised that organisations not only have to deal with the law, but also with
customers’ perceptions. One of the interviewees referred to the public backlash when Equens, an organisation handling payment data for banks, presented its plans to sell aggregated information based on the transaction data of consumers’
payments.44
A second challenge concerns the awareness of organisations of the (potential) value of their data and how this could be deployed to develop new insights, products or services. As described above, data analytics is often deployed at the core of the business model. New applications of data analytics require time, human resources and money that are not directed towards the primary process of an organisation. Therefore, awareness demands a certain belief because data
analytics requires an investment in staff, tools, infrastructure and possibly new data sources. Subsequently, awareness needs to be translated into commitment, especially by higher management.
A third challenge is the deployment of data analytics, which requires special skills.
Especially when organisations engage with analytics that extend their traditional business intelligence activities by asking new kinds of questions (for instance regarding profiling and predictions) and combining new kinds of datasets, new skills are required. The concept of ‘big data’ is explicitly mentioned in the interviews in this context. Furthermore, several interviewees indicate, as the complexity of analytics increases, the importance of a certain level of democratisation of the data analytics, actively engaging domain experts by making the analytics mode
accessible, for instance with visualisations and ‘playgrounds’ where they can use data and business intelligence tools to define new, more complex queries that can be executed by the actual data experts.
A fourth issue, the quality of the data, is directly linked to data integration and analytics. While data quality always represents a multi-faceted issue (e.g. regarding the integrity of data, frequency of updates), for example when multiple organisations collaborate, especially the integration of citizen and user participation in formal processes and databases can be difficult. These difficulties are both technical and organisational. Stadsbeheer, for example, uses data collected by citizens via the Beter Buiten app to detect mutations in the field, for instance due to vandalism.
However, this is not directly processed in the Stadsbeheer administrative back office, because they need to maintain quality. The police use data from Twitter and the more formal Police app, but this data is not part of the formal police reports that, for instance, will be used in court. The police works with a rigid chain of command and wants to keep control and maintain their existing protocols. This is also linked with authentication and a better understanding of the value of information from other sources, such as citizens/users. As described earlier, for Iens – where user
participation is core – reviewers need to log in before they can post something and Iens takes the experience of the reviewer into account to process the review in the overall rating of a restaurant to maintain the overall quality of its service.
44 http://www.volkskrant.nl/vk/nl/2680/Economie/article/detail/3446572/2013/05/24/Equens-ziet-voorlopig-af-van-verkoop-pingegevens.dhtml.
4 Towards a data ecosystem
The previous chapter introduced the outlines of the data landscape regarding liveability in Rotterdam, including the organisations that are active in this field, the value creation mechanisms of the data analytics behind their existing and new value propositions, their data applications and data strategies, and their ambitions and challenges regarding data-driven innovation. The third step of this study is the description of the emerging data ecosystem in the domain of liveability in the city of Rotterdam in order to investigate its impact on and implications for this domain.