Lilla Hortoványi

Strategy Without Templates

Adaptation in Digital Environments

9.5 Learning Through Inference and Practical Testing

If firms cannot fully see the environment, cannot easily identify causes, and cannot treat demand as an external given, then how do they learn at all? The fifth mechanism answers that question. Organizations often learn through inference, pattern recognition, and practical testing rather than through full causal understanding. This learning mode has several features.

First, organizations engage in repeated adjustment. They try a change, observe the response, and then decide whether to keep, revise, or reverse the move. This is not new in itself, but in algorithmically mediated settings the role of repeated testing becomes much more central because full explanation is often unavailable. Möhlmann et al. (2023) show that platform workers build and test informal theories about how algorithmic systems behave. Their learning resembles ongoing field inquiry rather than straightforward rule-following.

Second, organizations build heuristics around recurring patterns. A seller may notice that certain titles, images, keywords, or pricing moves tend to improve visibility. A content creator may observe that posting rhythm affects reach. In each case, the actor learns something useful without fully knowing why it works. The result is functional adaptation, but not stable causal knowledge.

Third, selective compliance becomes part of learning. Liu et al. (2022) show that ridesharing drivers do not simply accept algorithmic guidance when it conflicts with their experience. They compare platform suggestions with local knowledge and decide when to follow and when to ignore them. Learning, therefore, becomes comparative and judgment-based rather than purely receptive.

Fourth, this mode of learning can be emotionally demanding. Register et al. (2023) show that users on Instagram experience algorithmic uncertainty as a source of precarity and stress. Because rewards and penalties are hard to predict, actors become highly attentive to weak signals and engage in collective theorizing about what the algorithm “wants.” This matters because learning is not only cognitive. Persistent opacity can also shape attention, vigilance, and overinterpretation.

The key distinction here is between causal understanding and practical inference. Classical learning theories often assume that organizations improve by developing better explanations of cause and effect. In algorithmically mediated settings, that ideal is often unrealistic. Organizations improve by forming provisional rules that help them cope with recurring signal patterns, even when the deeper mechanism remains partly hidden.

This also creates fragility. Causal knowledge often travels across contexts because it explains why something works. Pattern-based heuristics are less transferable. A tactic that works on one platform, or under one version of an algorithm, may not work elsewhere. Therefore, firms can become highly competent in using a given system without gaining generalizable knowledge about the broader environment.

This point helps distinguish Chapter 9 from earlier chapters. Chapter 6 focused on experimentation as a way of acting without stable templates. The present chapter focuses on learning under mediated feedback. The issue here is not simply that firms experiment. It is that the signals guiding adjustment are themselves filtered and difficult to explain.

Tartalomjegyzék

Kiadó: Akadémiai Kiadó

Online megjelenés éve: 2026

ISBN: 978 963 664 204 4

Gazdaság

What happens when understanding comes only after action has already begun?

Traditional strategy rests on the assumption that organizations can understand their environment before deciding how to act. Yet the conditions that once allowed organizations to rely on benchmarking, best practices, and proven strategic templates can no longer be taken for granted. Today, organizations increasingly face situations for which no clear roadmap exists. Established assumptions become less reliable, familiar reference points lose their clarity, and strategic decisions must be made before their consequences can be fully understood.

Strategy Without Templates explores how organizations learn, adapt, and navigate environments in which uncertainty is pervasive and established templates are absent or no longer sufficient. Instead of treating strategy as a process of prediction and planning, the book explores how strategic paths take shape through action, experimentation, adjustment, and learning.

A central insight in the book is that temporary solutions are often necessary. What begins as a practical response to an immediate challenge may gradually shape future possibilities in unexpected ways. Some solutions create new opportunities and sources of advantage. Others become constraints that are difficult to overcome.

Hivatkozás: https://mersz.hu/hortovanyi-strategy-without-templates//

BibTeX EndNote Mendeley Zotero