← Back to main theory

The 8-Article Series on Constraint Coupling

One article per week exploring why AI fails in SMEs and what to do about it.
1
The AI Productivity Paradox in SMEs: Why 88% of Deployments Fail

Despite massive investments, 88% of AI deployments in SMEs fail to deliver measurable productivity gains. This article examines the structural reasons behind this paradox, drawing from 80+ field audits. We argue that the failure is not technological but organizational, rooted in unresolved constraint couplings.

April 2026
2
Beyond TAM and UTAUT: Why Adoption Models Miss the Point

Traditional adoption models (TAM, UTAUT) focus on individual acceptance factors. We demonstrate their inadequacy for SME contexts where a single decision-maker concentrates operational, strategic, and technological constraints. A new unit of analysis is needed.

April 2026
3
Constraint Coupling: When Human and IS Constraints Amplify Each Other

We introduce the concept of constraint coupling, extending Goldratt's Theory of Constraints. When a human bottleneck (dirigeant-goulot) is coupled with fragmented IS, technology deployment produces amplification, rejection, or workarounds rather than improvement.

May 2026
4
The Dirigeant-Goulot: Understanding the SME Owner as System Bottleneck

In 95% of SMEs audited, the owner-manager operates as the system's primary constraint. This article characterizes the dirigeant-goulot phenomenon and its implications for any technology deployment strategy.

May 2026
5
The Invariable Sequence A→B→C: A Deployment Protocol

We propose a three-phase deployment sequence: Align data flows (A), Assist teams (B), then Automate (C). Empirical evidence from 3 longitudinal case studies shows that violating this sequence produces predictable failure patterns.

June 2026
6
Three Case Studies: PLC Conseil, JLM Menuiserie, Cerfrance

Detailed multi-case analysis of three SMEs (accounting, construction, agriculture) demonstrating how constraint coupling manifests differently across industries while producing isomorphic failure patterns.

June 2026
7
Five Falsifiable Hypotheses for CCT

We formalize CCT into five testable predictions (H1–H5) with operationalized metrics. Each hypothesis specifies measurement instruments, expected effect sizes, and falsification criteria, establishing CCT as a refutable theoretical contribution.

July 2026
8
Toward a General Theory: Research Agenda and ICIS 2026

We outline a research program for CCT formalization: 10 instrumented audits over 12 months, mixed-methods design, and submission to ICIS 2026. This article serves as both a research agenda and a call for collaboration.

July 2026

Stay updated with our newsletter