From Stakeholder Feedback to Strategic Action

Research in Organizational Behavior and Stakeholder Governance

Paper III of the Series · Feedback Events and Their Handling

From Stakeholder Feedback

to Strategic Action

How the Same Quality Issue Acquires Budget,

an Accountable Owner, and an Implementation Record

Andrew David W.Y.

WISERUNION / weitengfei.com

Grok

xAI

Corresponding author: Andrew David W.Y. September 2026

Note on the Manuscript

This is the third paper in the series. The unit of analysis is the feedback event and its cross-functional handling. The preferred path is a longitudinal comparative case study. The explanatory task is to show how feedback enters action. Strategic action is operationalized as an organizational response that simultaneously has a budget, a named owner, and an implementation record. An ordinary after-sales reply, a survey response rate, or the number of channels is not, by itself, strategic action. The manuscript specifies a process model, a comparative logic, and a data protocol. Because cases have not yet been collected, no fabricated process findings are reported.

Abstract

Growing manufacturers rarely lack feedback channels. Employees can report a product-quality defect through teams, inspection, and internal systems; customers can report the same defect through complaints, returns, reviews, and after-sales files. Firms therefore accumulate remarks without being able to show which remark changed a standard, a budget, or a cross-functional process. Voice and issue-selling research explains why people speak and how middle managers place problems on agendas.[1][2] Quality-management work describes how complaints are classified. The remaining gap is how, for the same quality issue arriving from internal and external sources, evidential standing, organizational translation, agenda authority, and resource handover connect or break—and why channel volume does not guarantee conversion.

Conversion is defined as the process by which a feedback event is translated into a verifiable problem, enters an agenda with authority to allocate resources, receives a continuous coupling of ownership and means, and leaves an inspectable implementation record. A five-link model is proposed: reception, translation, agenda entry, coupling of responsibility and resources, and implementation with notice back to the source. The planned design follows paired employee-source and customer-source events on one class of quality problem in two to four growing manufacturers that already operate feedback channels and are expanding or changing product or process. Process tracing compares four endings: successful conversion, stoppage in after-sales, stoppage on the agenda, and blockage by resource constraint.

The intended contribution is to move stakeholder participation from collecting opinions to a decision response that can be evidenced, and to test whether the intelligence and strategic-control tools in the companion treatise actually connect at the event level. Practically, the study helps quality and strategy leads judge whether the missing piece is channels, translation capacity, agenda authority, or budgetary handover. Channel participation rates are not written as governance improvement.

Keywords: stakeholder feedback; strategic action; issue selling; process tracing; product quality; growing manufacturers

1. Introduction

Firms have become skilled at collecting opinions. Surveys, hotlines, dealer weeklies, andon cords, suggestion boxes, and customer reviews can run at once. Collection does not answer a harder question: when does a remark about product quality count as having entered strategic action? If action is defined as “someone replied,” participation governance stops at courtesy. If action is defined as a recordable change in standards, resource allocation, or cross-functional process, the study must follow the event through translation, agenda, budget, and execution.

The Luck Companies material in the base reader suggests that 360-degree assessment and anonymous feedback can be linked to managerial action, but the public account is not a complete event file and cannot show how each remark entered a budget and an implementation record.[3] Voice research is now large. Morrison’s decade review is explicit: speaking, remaining silent, being heard, and being acted on are not the same thing, and scales must not be mixed.[2] Dutton and Ashford treat middle managers’ selling of issues to the top as an early mechanism of strategic agenda formation, yet resources and execution can still break after an issue is seen.[1]

The same quality problem is especially good at exposing those breaks. Employee reports often concern process windows, incoming variation, and inspection gaps. Customer reports often concern field failure, delivery, and on-site loss. Both may point to one defect and still be sent separately into inspection, after-sales, sales, and strategy meetings, under different evidence standards. Four endings then become common: after-sales closure with no change of standard; minutes without a budget; a budget without an owner; an owner without an implementation record. More channels fragment remarks. Conversion does not rise by itself.

The one-sentence question is therefore: in growing manufacturing firms, through which organizational processing steps can employee and customer feedback about the same product-quality issue be converted into strategic action that has a budget, an accountable owner, and an implementation record? Four boundaries follow. First, only actions that affect product standards, resource allocation, or cross-functional process are studied; an ordinary after-sales reply is not automatically strategic action. Second, public opinion campaigns and crisis communication as a whole are out of scope. Third, channel counts, participation rates, and satisfaction scores are not written as proof that governance works. Fourth, each link in the process model is a connection to be tested, not a necessity announced in advance; if action repeatedly occurs without a given link, the chain should be rebuilt.

2. Theoretical Background

2.1 Voice, Silence, and Organizational Response

Employee-voice research has long asked when people speak, when they stay silent, and how leadership, psychological safety, and identification shape expression.[2] That work is necessary for understanding why feedback appears. It is not sufficient for understanding why feedback becomes action. After the mouth opens, a remark can still be downgraded to venting, rewritten as an isolated operating error, or lost in a handoff. Customer-complaint research is equally rich and often stops at satisfaction repair, without tracing whether a standard was rewritten. This paper takes “expression has already occurred” as the starting point of analysis, not as evidence that conversion is complete.

2.2 Issue Selling and the Allocation of Attention

Dutton and Ashford note that top-management time and attention are scarce, and that how middle managers package, bundle, and send problems upward affects what enters the strategic agenda.[1] Later work added contextual reading, impression management, and the moves of successful sellers. That line is useful for explaining why a given quality issue is seen. It still leaves the second half: once seen, does the issue obtain a budget, a named owner, and an inspectable implementation? Equating successful selling with strategic action writes minutes as execution. This paper treats agenda entry as a necessary but not sufficient intermediate state.

2.3 Quality Loops and Strategic Control

Manufacturers usually already run nonconformance, corrective-and-preventive-action, and complaint processes. Those processes are good at turning problems into tickets. They are not always good at turning repeated tickets into a standard or a capacity decision. The interest-understanding, managerial-intelligence, and strategic-control tools in Freeman, Harrison, and Zyglidopoulos suggest that external signals should enter the control cycle.[3] The companion treatise treats feedback as one entrance through which co-creation reaches governance, and stresses recognition, responsibility, and traceability.[4] What must be tested at event level is where those tools actually break in growing firms that are expanding, changing models, and short of staff.

2.4 Dual Sources: Employees and Customers

Internal and external reports of the same defect do not carry the same evidential standing. Employee material is often closer to process parameters and may still be heard as internal complaint. Customer material often arrives with claims and relationship pressure and may still lack reproducible technical detail. When the two meet, the organization may trust the complaint that carries money, or the internal report that carries test data. Process research must code source into the event rather than write “stakeholder feedback” as a single input. That is also the difference from Papers I and VIII: Paper I asks why individuals keep contributing; Paper VIII asks who gets scarce attention first; this paper asks how one problem already inside a channel travels the rest of the conversion chain.

3. Process Model and Propositions

3.1 A Five-Link Model

Conversion is the process by which a feedback event moves from entry into the organization to strategic action. Strategic action has three operational tests, all required: an identifiable budget or explicit claim on capacity; a named owner with authority; and an implementation record that shows a standard, a process, or a configuration has changed. If any test fails, the event is coded as incomplete conversion, with a mark for the link at which it stopped.

Table 1. Five Links from Feedback to Strategic Action

Note: The five links are an analytic frame, not a ritual that must occur in this order. The comparative cases observe which connections appear, which are skipped, and which are replaced by other pressures.

3.2 Propositions

Proposition 1. Raising the number of channels and the participation rate does not, by itself, raise the conversion rate into strategic action. Conversion depends on whether the event is translated and reaches an agenda that can allocate resources.

Proposition 2. The same quality issue is less likely to convert if it arrives only as a customer complaint or only as an employee report than if the two sources are joined at the translation link.

Proposition 3. Translation quality—reproducibility, checkable evidence, and countable recurrence—positively affects agenda entry. Source identity is weaker than translation quality unless source directly changes access to evidence.

Proposition 4. After agenda entry, events that do not couple an owner with a budget or capacity window in the same decision are more likely to stop at “discussed, not acted.”

Proposition 5. External regulation, a customer cutoff, or a safety incident can bypass earlier links and start action directly. Such action should be coded separately and must not be used as proof that the everyday feedback chain works.

Proposition 6. If high-quality translations are blocked for long periods by resource constraint, the model should add that constraint rather than keep writing failure as “poor feedback quality.”

Proposition 7. If strategic action that meets all three operational tests repeatedly occurs without translation, agenda entry, or coupling, the model should be rebuilt; a skipped link should not be kept as a necessity.

4. Method

4.1 Overall Strategy

The preferred path is a longitudinal comparative case study. The unit is the feedback event, not an abstract firm culture. Two to four growing manufacturers in one segment are sought. They should already run employee and customer channels, be expanding, changing models, or adjusting process, and allow tracking of one class of quality problem over about six to ten months. Each firm should contribute at least eight to twelve events, including employee-only, customer-only, and joined-source cases where possible. Endings are compared across successful conversion, after-sales stoppage, agenda stoppage, and resource blockage.

4.2 Evidence

Each event receives a file built, as far as possible, by triangulation: original channel records (tickets, complaints, andon, proposals); minutes and approval flows; budget or work-order changes; versions of standards or work instructions; and interviews with quality, production, and sales leads plus at least one front-line employee and one customer-facing counterpart. Interviews fill timelines and reasons. They do not replace original records. A firm that will give only summary dashboards, and will not open event files, is kept as background and does not enter the main comparison.

4.3 Coding and Process Tracing

Structured coding first marks each of the five links as occurred or not, with a timestamp, a responsible unit, and an evidence type. Process tracing then asks, for each successful conversion, whether action would still have occurred if a given link were removed; and, for each failure, where the earliest stable break sits. Source, defect class, money or line-stop loss, and whether safety or regulation is touched are treated as contextual conditions, not as dependent variables. Two coders first work a subset independently; agreement is checked before the full set is coded.

4.4 Comparative Logic

Across firms the comparison is institutional: whether channels connect, whether translation has a dedicated role, whether the agenda can commit resources, and whether notice back to the source is mandatory. Inside a firm the comparison is among events under the same rules but different sources and loss magnitudes, so that institutional effects can be separated from case-level pressure. Action started by an external shock is grouped apart to test Proposition 5. Firms whose resource constraint rises inside the window are used to test Proposition 6.

4.5 Ethics

Firm agreements and consent from each class of interviewee are obtained separately. Identifiable individual criticism is not returned to management. Customer names, claim amounts, and unpublished defects are de-identified. The study does not enter live disciplinary proceedings. Compensation is not tied to supplying success stories that match expectations.

5. Analysis and Falsification

Analysis produces three tables. An event-path table shows which links each event passed, where it stopped, and whether the three tests of strategic action were met. A configuration table shows how translation quality, joined sources, agenda authority, resource coupling, and external shock covary with endings. A counterexample table lists successes without a given link, failures with all links present, and successes driven only by regulation.

Falsification rules are written first. If successful conversions commonly lack translation or coupling, Propositions 3 and 4 should be withdrawn or reduced to probabilistic conditions. If firms with more channels convert at higher rates and the process does not differ, Proposition 1 should be revised. If employee-source and customer-source events do not differ before and after joining, Proposition 2 should be dropped. If every “success” is driven by cutoff or accident, the everyday feedback chain cannot be declared effective. Process cases cannot estimate a general rate, and a single firm’s practice must not be written as best practice.

Table 2. Propositions and Falsification Rules

6. Intended Theoretical Contributions

First, the measurement of stakeholder participation is moved from the input side to the output side. The input side asks how many opinions were collected. The output side asks which opinion left a budget, an owner, and a changed standard. Without that move, participation governance is easily replaced by channel construction.

Second, issue selling is extended forward into resource coupling and implementation residue. Dutton and Ashford explain how a problem is seen;[1] this paper explains how seeing becomes inspectable action, and allows some action to bypass earlier links under regulation or cutoff.

Third, employees and customers are written as dual sources on one defect, not as two parallel “stakeholder programs.” If co-creation in the companion treatise is to enter governance, the text must show how heterogeneous evidence is joined, rather than assume that every voice carries the same probative force.[4]

Fourth, a contrast is prepared for Paper VIII. This paper follows events already inside a channel until conversion is complete or broken. Paper VIII explains which issues never enter a channel when attention is scarce. Both concern ordering. One explains how a process finishes; the other explains how a process starts.

7. Managerial Implications

If the comparisons support the model, quality and strategy leads should not treat “another suggestion box” as a remedy. What is more often missing are four concrete capacities: a translation role that can write complaints and line reports as reproducible problems; a problem ledger that can join the two sources; an agenda that can actually approve a budget or change a standard, rather than a briefing; and an implementation record that forces notice back to the source. A local patch can cut today’s loss. It should not be registered as strategic action.

Growing firms in particular must admit resource constraint. When high-quality feedback is blocked, the honest managerial move is to record the reason for delay and a date for review, not to blame employees for poor expression or customers for being too demanding. Regulation-driven correction should be listed separately, so that failure of the everyday chain is not hidden by success in a crisis.

8. Limitations, Boundaries, and Further Research

This paper is a case design, not a completed verification of process theory. Accessible firms may tilt toward those willing to cooperate and those with relatively complete records, so failed conversion will be undercounted. Retrospective interviews will polish “we have always taken quality seriously.” Event files must therefore remain the primary evidence. External validity is limited to growing manufacturers and quality-class issues; it does not automatically travel to pay, community, or public crisis.

Boundaries with the rest of the series: Paper I asks why employees continue to contribute; Paper II asks how one increment is divided; Paper VII asks how incentive systems change the consistency of words and deeds; Paper VIII asks how owner attention ranks issues. If this paper can obtain only agenda rankings and no implementation or budget trace, the prospectus already says it should be considered for merger with Paper VIII rather than have the operational definition of strategic action lowered in order to keep a separate article.

9. Conclusion

Feedback that enters the organization shows only that someone spoke. Strategic action requires the organization to write that speech as a verifiable problem, send it to an agenda that can allocate resources, name an owner and a budget, and leave a record that a standard or a process has changed. There can be many channels, and each of those four steps can still break. Employee and customer reports of the same defect give a natural contrast for watching the breaks: the evidence differs, the pressure differs, and so does the way each report is believed.

The conversion mechanism is kept as a process proposition to be traced. If the cases show that action is always started by cutoff and accident, the everyday participation apparatus has not yet been built. If standards are frequently rewritten without the coupling of budget and owner, the model should change rather than insist that all five links remain necessary. The task of the research is to make the sentence “we listened to stakeholders” require an event-level implementation record before it can stand.

References

[1] Dutton, J. E., & Ashford, S. J. (1993). Selling issues to top management. Academy of Management Review, 18(3), 397–428.

[2] Morrison, E. W. (2023). Employee voice and silence: Taking stock a decade later. Annual Review of Organizational Psychology and Organizational Behavior, 10, 79–107.

[3] Freeman, R. E., Harrison, J. S., & Zyglidopoulos, S. (2018). Stakeholder theory: Concepts and strategies. Cambridge University Press.

[4] Andrew David W.Y. (2026). Stakeholder capital theory (1st ed. working manuscript). WISERUNION.

[5] Andrew David W.Y. et al. (2026). Prospectus for eight papers derived from stakeholder theory (discussion draft, 10 September 2026).

[6] Dutton, J. E., & Duncan, R. B. (1987). The creation of momentum for change through the process of strategic issue diagnosis. Strategic Management Journal, 8(3), 279–295.

[7] Langley, A. (1999). Strategies for theorizing from process data. Academy of Management Review, 24(4), 691–710.

[8] Yin, R. K. (2018). Case study research and applications (6th ed.). Sage.

[9] Beach, D., & Pedersen, R. B. (2019). Process-tracing methods (2nd ed.). University of Michigan Press.

[10] Harrison, J. S., Bosse, D. A., & Phillips, R. A. (2010). Managing for stakeholders, stakeholder utility functions, and competitive advantage. Strategic Management Journal, 31(1), 58–74.

[11] Phillips, R. (2003). Stakeholder theory and organizational ethics. Berrett-Koehler.

[12] Mitchell, R. K., Agle, B. R., & Wood, D. J. (1997). Toward a theory of stakeholder identification and salience. Academy of Management Review, 22(4), 853–886.

[13] Wood, D. J., Mitchell, R. K., Agle, B. R., & Bryan, L. M. (2021). Stakeholder identification and salience after 20 years. Business & Society.

Appendix. Case-Implementation Notes

Appendix A. Event sampling: first fix one recurring quality defect (seal, tolerance, coating, or electrical performance), then draw employee-only, customer-only, and joined-source events. Do not draw only the success stories that leaders remember. Appendix B. Timeline template: date of reception; date translation is complete; date of first entry onto an authoritative agenda; date owner and budget are named; date a standard or process version changes; date of notice back. Appendix C. Merger condition with Paper VIII: if implementation records and budget traces cannot be obtained and only issue rankings remain, the operational definition of strategic action should not be lowered merely to keep a separate paper.

— End of manuscript —