HiFi Prototype Design Complete
Rationale
Freelancers are vital part of the Creative Industries, comprising over 30% of the total workforce, and c.70% of the workforce in certain sectors (Easton and Beckett, 2021). Freelance workers frequently have less access to training than their employed counterparts, and as a result they tend to rely on non-formal and informal routes to build knowledge and develop skills (Henry et al, 2021). Such routes are likely to be reliant on existing connections and as the Creative Industries has a high levels of social closure (Brook et al, 2020), which is likely to lead to inequalities in access to training opportunities.
Our research shows that there is a prominent knowledge gap in the freelance community, with 56.5% of freelancers recently surveyed responding that they were not well informed of their rights, 68.5% having issues with late or non-payment, and 57.1% relying on peer advice to solve issues such as contract negotiations, rates setting, payments, boundary setting, and soft skills around communicating clients (Connell et al, 2022). Online forums play an important role in enabling knowledge sharing and informal training in creative occupations (Patrick- Thomson and Kranert, 2020). Members of the forum studied in this research would regularly pose questions involving similar contractual situations (e.g. pricing work, contractual issues, and late payments), and forum users offered advice. However, there was some frustration at the lack of a better way to address these recurring, highly similar questions. This recurrence of similar questions suggests that there is a high demand for peer advice, and that forums are a relatively inefficient way of providing this.
To look at how this peer advice system could be developed using advanced technologies, a project team was established involving academics working in the Business and Computing areas of Edinburgh Napier University. In particular, we sought to investigate the potential of an AI powered advisor to provide a system-based solution to the recurring questions and answers of freelancers working in the Creative Industries (aka creative freelancers). The team comprised Dr Holly Patrick-Thomson (Lecturer, Business School), Dr Paul Lapok (Research Fellow, SCEBE) and Associate Professor Alistair Lawson (SCEBE).
The Project
Figure 1 outlines the research process followed in developing a Hi-Fi prototype of the Freelance Advisor system. The 8 initial stakeholder analysis interviews showed that setting rates was the most common contractual issue encountered by our sample of creative freelancers (other, less common, issues included protecting IP, securing a contract, and under/late payment). The interviews also uncovered that although freelancers do use the Facebook groups mentioned earlier for advice, they more commonly turn to their peers because of the trusted and reliable nature of peer advice. However, a desire for anonymity often impedes their ability to access advice from colleagues who are not direct friends. There was also a recognition that the advice required was often advice around the soft skill of negotiating rates, rather than the concrete knowledge of how much to charge. As such, we derived certain parameters for the system:
· A recommendations system that not only derived rates advice from the internet but also actively procured peer advice, using;
· The ability to ask questions and make recommendations anonymously, however;
· Profiles would need to include sufficient information to contextualise the questions asked and recommendations given (without identifying the peer), and;
· The ability to rate (i.e. give a review) on the advice given would be key to allowing the system to improve via understanding which rates and recommendations are more/less useful to users, and;
· A further step in making the recommendations trustworthy would be to have experienced peers involved in the initial stage of launching the system, and having an industry partner who could support development and lend legitimacy, and finally;
· Such a system could possibly be charged for, although it was thought to be most useful to those at the start of their careers, who are least able to pay. The possibility of funding the system through advertisements was not attractive to the stakeholders.
We then used these parameters to build a prototype of the system, which we then tested with a group of 7 creative freelancers (6 from the original group, and 1 replacement for a participant who was unable to continue). These evaluation interviews involved a walkthrough of the system, the application of a System Usability Scale, and a series of more qualitative questions on system design, types of advice provided and views on key features. This allowed us to undertake a review of the prototype and to develop a final overview of the potential system.
One of the key challenges and learning opportunities during this project arose from the multi-disciplinary nature of the research team. Much of the overall methodology adopted came from the field of computer science, but the initial problem from a qualitative study in the field of business. As such, even in a comparatively small project, there was much to translate across disciplinary practices and vocabularies. This was an important developmental opportunity for the research team.
Project Outcomes
They key outcome of this project is a HI-Fi prototype for a system that addresses a key issue of knowledge transmission in the freelancer ecosystem, and which has demonstrated usability and attractiveness for potential users. We intend to advance this project by seeking partners to develop and trial the system, and we would be pleased to hear from anyone interested in collaborating in this development.
We would like to note that in addition to the excellent support we received from Creative Informatics (without which we would not have even imagined this project), we were happy to receive advice and support for this project from Creative Edinburgh, Creative Bridge and colleagues from the British Academy of Management.
Stage One Research Process