Thesis
Revisiting Ansoff’s weak signal theory : exploring the effects of machine learning on filters of weak signals
- Creator
- Rights statement
- Awarding institution
- University of Strathclyde
- Date of award
- 2025
- Thesis identifier
- T17250
- Person Identifier (Local)
- 201866624
- Qualification Level
- Qualification Name
- Department, School or Faculty
- Abstract
- Strategic foresight is an important managerial activity to develop coping strategies for future uncertainties. Ansoff in the 70s proposed weak signals theory to manage uncertainties. The literature on Ansoff’s weak signals has seen an increase in recent years. It is argued that weak signal analysis can aid an organisation to weather disruptions and remain competitive. As disruption are preceded by subtle signals called weak signals. By monitoring, detecting, amplifying, and reacting to these signals, organisations can adapt to the dynamic environment and be relevant in the long term. This is where forecasting tools fail as it is impossible to forecast into long term future. But the research in this field is in its infancy compared to other fields such as scenario planning or more broadly strategic management. This research reviewed the extant literature on weak signals through a systematic literature review and identified the current research gaps. Based on the research gaps found, a research agenda was set to explore how machines enhance the detection and alter the perceiving of weak signals (filters of weak signals). Through the lens of pragmatism, this research adopted a multi methods approach. The methods included semi-structured interviews, participant observation, and Miro board observations. The findings from the research showed that it does not seem that machines can enhance weak signal detection. But there is a nuance to this, as the value of AI output depends on who the user is. If an expert is using it, then perhaps not as they seem to contain information already known. But if it is a novice who is using AI then there seems to be value in the output. Filters of weak signal do remain even with the use of machines. It also seemed that machines amplified filters. Findings from the research allowed for contribution to practice such as, foresight experts need to up-skill with respect to AI. And prompt engineering is an important skill required whilst using AI models such as LLMs.
- Advisor / supervisor
- Ates, Aylin
- Resource Type
- DOI
Relations
Items
| Thumbnail | Title | Date Uploaded | Visibility | Actions |
|---|---|---|---|---|
|
|
PDF of thesis T17250 | 2025-04-25 | Public | Download |