On our main employee survey platform, PRISM plays a crucial role in summarising comments, categorising feedback, and suggesting actionable steps. However, we have implemented a soft-threshold of 150+ employees for using PRISM for several important reasons:
Efficacy: Research indicates that AI-driven analysis is significantly more reliable with larger datasets. For example, studies published in IEEE Transactions on Affective Computing highlight that AI models improve their accuracy and reliability when applied to larger datasets, particularly in sentiment and text categorisation tasks, as larger datasets provide more diverse and representative data for training the models.
Cost-effectiveness: We have set the threshold at 150+ employees to ensure that the cost-benefit ratio is optimised. AI analytics becomes significantly more efficient and cost-effective when applied to larger datasets, automating repetitive processes and reducing the need for manual intervention. For smaller datasets, the operational cost of AI-driven analysis may outweigh the benefits, making it more practical and insightful to use human analysis. Thus, the 150-employee threshold ensures that we balance the efficiency of AI with cost considerations, avoiding unnecessary expenses for smaller datasets.
Human Expertise: For smaller datasets, human analysis often provides more insightful and efficient results. This approach supports our commitment to combining technological innovation with valuable human insight, ensuring that our clients get the best of both worlds.
Future Developments
We are working on integrating quantitative data analysis into PRISM on the main platform. This will enhance PRISM's capabilities and bring it in line with the functionality currently available for 360 assessments, further expanding the value it offers to our clients.