The Psychology of Technology Adoption

Organizations routinely deploy technologies that are demonstrably better than what they replace and watch them go unused. The reflex explanation is resistance to change, which explains nothing. Four decades of adoption research offer something more specific: people are evaluating the technology against criteria the deploying organization usually has not measured.

Two perceptions carry most of the variance

Davis (1989) proposed that adoption intention is driven principally by perceived usefulness — whether using the system will improve performance — and perceived ease of use. The Technology Acceptance Model that followed became one of the most heavily replicated frameworks in information systems research, and its central finding has held up: these are perceptions, not properties.

The distinction matters. A system can be objectively faster and be perceived as less useful because its benefit accrues to a different part of the organization than the one operating it. Perceived usefulness is usefulness to the user, evaluated against their own goals and their own accountability.

Venkatesh and Davis (2000) extended the model to include social influence and cognitive instrumental processes, and Venkatesh, Morris, Davis and Davis (2003) consolidated eight competing models into UTAUT, with performance expectancy, effort expectancy, social influence and facilitating conditions as the core predictors. The consolidation is telling: independent research traditions kept rediscovering approximately the same small set of factors.

Diffusion is a social process, not a sum of individual decisions

Rogers (2003) characterised adoption as spreading through a social system over time, with the perceived attributes of an innovation — relative advantage, compatibility, complexity, trialability and observability — governing the rate.

Two of those are habitually neglected. Trialability is whether a person can experiment at low cost before committing; observability is whether they can see others using it and see the results. Enterprise rollouts often eliminate both, by launching organization-wide on a fixed date with no reversible trial period. The deployment method itself suppresses two of the five attributes that drive uptake.

Adoption and continuance are different problems

Bhattacherjee (2001) argued that continued use is governed by a different process than initial acceptance: confirmation of pre-adoption expectations, leading to satisfaction, leading to continuance intention. Initial adoption can be driven by mandate or novelty. Continuance cannot.

This has an uncomfortable implication for launch marketing. Overselling a technology raises initial adoption and raises expectations, and if expectations are not confirmed, satisfaction falls further than it would have with modest claims. High-expectation launches purchase early numbers at the cost of the retention curve.

Risk and switching costs

Adoption also carries perceived risk: performance risk, financial risk, and — in a professional setting — the social risk of being visibly associated with a failure. Where the personal downside of a failed adoption exceeds the personal upside of a successful one, non-adoption is the individually rational choice regardless of the organizational case.

This asymmetry is often the whole story behind a stalled rollout. The organization is calculating expected value at the aggregate level; the individual is calculating exposure at the personal level. No amount of training addresses that mismatch, because it is not a knowledge problem.

What this means in practice

Diagnose before intervening. Low uptake caused by unclear usefulness, by effort cost, by absent social proof, and by personal risk exposure look identical in a dashboard and require entirely different responses. Ask users what they think the technology is for and what they think happens to them if it fails — the answers usually locate the constraint within a few conversations. The companion piece on algorithm aversion covers a specific version of this problem for decision-support systems.

References

1. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340.
2. Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model. Management Science, 46(2), 186–204.
3. Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478.
4. Rogers, E. M. (2003). Diffusion of Innovations (5th ed.). Free Press.
5. Bhattacherjee, A. (2001). Understanding information systems continuance: An expectation-confirmation model. MIS Quarterly, 25(3), 351–370.