Strategic discourse as competitive capability: measuring resilience and antifragility in B2B firms using NLP
Piotr Kwiatek & Sławomir Paweł Zapłata
What the paper says
Purpose This study aims to operationalize and empirically examine the concepts of resilience and antifragility in the context of business-to-business firms by analyzing how organizations discursively frame risk and adaptation during earnings calls. We address the gap in scalable, firm-level measurement of these dynamic capabilities and investigate their temporal evolution across crisis periods. Design/methodology/approach We develop a novel NLP-based pipeline that processes over 17,000 earnings call transcripts (2015–2021) from production-sector firms. The pipeline combines rule-based text processing, transformer-based sentiment models (FinBERT), and domain-specific lexicons to extract sentence-level indicators. Composite scores for risk exposure, resilience, and antifragility are computed and analyzed longitudinally. A clustering approach segments firms into four adaptive archetypes. Findings The study reveals four distinct discursive archetypes that vary in their orientation toward resilience and risk. Firms shift between these archetypes over time, suggesting a dynamic discursive adaptation process. High-performing firms exhibit narrative strategies that frame adversity as an opportunity, reinforcing stakeholder confidence and signaling adaptive capability. Originality/value This research extends dynamic capabilities theory by conceptualizing discourse as a strategic intangible asset. It introduces a novel methodology for quantifying organizational discourse and links discursive agility to firm competitiveness. The findings have practical implications for B2B firms seeking to manage stakeholder perceptions and enhance adaptive marketing capabilities through strategic communication.
Evidence weight
Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40
| F · citation impact | 0.50 × 0.4 = 0.20 |
| M · momentum | 0.50 × 0.15 = 0.07 |
| V · venue signal | 0.50 × 0.05 = 0.03 |
| R · text relevance † | 0.50 × 0.4 = 0.20 |
† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.