The Digital Transformation Paradox
Organizations invest billions in digital technology, yet 70% of transformations fail (McKinsey 2023) and 88% miss their targets (Bain 2024). This isn't a technology problem — it's a constraint coupling problem that existing theories address only partially.
TOC & NAT: Incomplete
Goldratt's Theory of Constraints assumes singular, sequential bottlenecks. Perrow's Normal Accident Theory diagnoses coupling but doesn't prescribe resolution.
TAM/UTAUT Miss It
Technology adoption models explain individual acceptance, not systemic organizational failures caused by pre-existing bottlenecks.
The Gap
No theory explains what happens when you deploy technology on a system where human and IS constraints are structurally interdependent.
Constraint Coupling Theory
CCT integrates three theoretical traditions — the Theory of Constraints (Goldratt 1984), Sociotechnical Systems theory (Trist 1951), and Normal Accident Theory (Perrow 1984) — to explain why technology deployment on coupled constraints amplifies dysfunction rather than resolving it. The model is generalized: SMEs are the simplified case (n=2 constraints), while ETIs and large enterprises exhibit distributed coupling across multiple organizational layers.
Constraint Coupling
When organizational constraints of different natures — human (owner-manager SPOF, middle management bottleneck) and technical (fragmented IS, legacy systems) — are structurally interdependent. Resolving one requires changes in the other, creating a circular dependency that Perrow (1984) identified as the root of cascading failures.
The Constraint Amplification Effect
Deploying technology on coupled constraints degrades rather than improves performance. In SMEs, this manifests as owner-manager overload. In large enterprises, it manifests as cascading silos and project paralysis. The mechanism is identical — constraint coupling amplifies dysfunction at every organizational scale.
Three Failure Modes
Amplification — increased data volume and coordination overhead. Rejection — users resist technology that disrupts their workflow. Workaround — shadow systems that fragment information further.
"You cannot automate chaos. You must first align the flows, then assist the people, then — and only then — automate."
The A→B→C Invariable Sequence
CCT proposes that successful technology deployment in coupled constraint systems requires a strict three-phase sequence:
Align
Stabilize data flows. Eliminate double entry. Connect existing tools before adding new ones.
Assist
Deploy technology that reduces the human bottleneck's cognitive load. Build team readiness.
Automate
Only now can AI and automation succeed — clean data and prepared teams are in place.
Skipping phases produces systematic failures: Skip A → technical failures. Skip B → change resistance. Skip A and B → both.
Empirical Evidence
CCT is grounded in multi-case study research following Eisenhardt (1989) and Yin (2018) methodologies. The cases span from a 10-person SME to a 200-employee cooperative, testing CCT across different coupling levels and organizational scales.
PLC Conseil — Accounting Firm
Managing director = strategic + operational bottleneck. 5 disconnected tools (accounting, CRM, docs, email, spreadsheets). AI chatbot deployed at C level for client communications — without stabilizing data flows (A) or building team readiness (B).
Marginal gains (H2 confirmed)JLM Menuiserie — Construction (€5M)
Project coordinator = single point of failure managing 15-25 concurrent construction sites alone. 11 disconnected tools across quoting, planning, procurement, and field coordination. A→B→C sequence followed deliberately over 6 months with the Saxium platform.
Positive indicators (H3 testing)Cerfrance — Agricultural Cooperative
Field advisors = distributed individual bottlenecks across agricultural cooperative network. Custom LDM tool deployed directly at C level for advisor workflow — without prior data alignment or change management.
Bugs + resistance (H4, H5 confirmed)Falsifiable Hypotheses
CCT proposes five testable predictions, distinguishing it from frameworks that describe but cannot predict:
Research Agenda
About the Researcher
Working Paper
The foundational paper for Constraint Coupling Theory has been submitted to ICIS 2026 (International Conference on Information Systems, Lisbon) in the "Organizational Strategy, Governance, and Transformation" track.
A proposal has also been submitted to MIT Sloan Management Review. The theory builds on 13 years of fieldwork across 80+ SME audits.
How to Cite
Fritsch, T. (2026). Constraint Coupling Theory: Why Technology Deployment on Coupled Human-IS Constraints Produces the Opposite of the Expected Effect. Working Paper. Available at: constraintcoupling.com
Key References
- Goldratt, E.M. & Cox, J. (1984). The Goal: A Process of Ongoing Improvement. North River Press.
- Brynjolfsson, E. (1993). The Productivity Paradox of Information Technology. Communications of the ACM, 36(12), 66–77.
- Eisenhardt, K.M. (1989). Building theories from case study research. Academy of Management Review, 14(4), 532–550.
- Yin, R.K. (2018). Case Study Research and Applications: Design and Methods (6th ed.). Sage.
- Venkatesh, V. et al. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478.
- DeLone, W.H. & McLean, E.R. (2003). The DeLone and McLean Model of IS Success. JMIS, 19(4), 9–30.