Motivation
Motivation is what makes a person willing to spend effort on the work — and in software teams the strongest reported source of it is not pay or praise but identifying with the task: knowing what the work is for and finding it interesting [1].
1. Herzberg’s two factors
Herzberg separates what makes people satisfied from what makes them dissatisfied, and argues they are different lists [2]. Motivators are intrinsic to the work: achievement, recognition for achievement, the work itself, responsibility, advancement and growth. Hygiene factors are the surroundings: company policy and administration, supervision, working conditions, salary, relationships with colleagues, status and security. Removing dissatisfaction is not the same as motivating. Example: a developer trusted to design a service stays engaged (motivator) while a flaky test suite grinds them down (hygiene) — fixing the suite buys back attention but does not create commitment.
His evidence pools 12 investigations across 1,685 employees: of the factors contributing to satisfaction 81% were motivators, and of those contributing to dissatisfaction 69% involved hygiene. The theory “was first drawn from an examination of events in the lives of engineers and accountants” (p. 6) — a closer population to a software team than most classics here.
2. What software engineers actually report
Beecham and colleagues reviewed 92 papers spanning 1980–2006 across 16 countries and extracted 21 motivators [1]. The most-cited one is not on Herzberg’s list:
| Motivator | Studies reporting it |
|---|---|
| Identify with the task — clear goals, personal interest, knowing the purpose | 20 |
| Good management · employee participation | 16 each |
| Career path | 15 |
| Rewards and incentives · variety of work · sense of belonging | 14 each |
| Recognition | 12 |
| Autonomy | 9 |
The leading de-motivators are a poor working environment and lack of resources (9), poor management (7), uncompetitive pay (6), stress (5) and poor communication (5). Herzberg’s frame survives — working conditions sit on the hygiene side — but the content shifts: pay and recognition rank well behind identifying with the task.
3. Two frameworks worth the vocabulary
Self-determination theory (Deci and Ryan) holds that people are motivated when three needs are met — autonomy over how the work is done, competence at it, and relatedness to the people doing it. Choice of framework, time to learn it and a stand-up worth attending address all three.
Expectancy theory (Vroom) multiplies three beliefs: expectancy that effort produces performance, instrumentality that performance produces a reward, and valence that the reward is worth having. One term at zero zeroes the product, which is why a bonus scheme nobody believes in motivates nobody.
4. Practical techniques
Most workable techniques make effort visible and tie it to the goal: progress on a board or burndown, end-of-sprint demos, peer-recognition channels, rotation between front-end, back-end and test work, and rituals such as retrospectives that mark what changed. These are largely identifiability measures, so Thompson’s qualification applies to all of them — visibility only changes behaviour if somebody acts on what it shows (p. 51) [3]. See Social Loafing.
5. When the collaborator is a model
Two recent studies asked whether generative AI helps or harms intrinsic motivation and reached opposite answers. Both are sound; they measured different things, and separating them is the skill worth taking away.
| Wu et al. [4] | Corgnet et al. [5] | |
|---|---|---|
| Compares | AI-assisted work → then unassisted | Working with AI vs with a human |
| Design | 4 pre-registered experiments, N = 3,562 | n = 1,091 plus 180 raters |
| Finding | Performance rose, but sense of control and intrinsic motivation fell and boredom rose on the handoff back | Relatedness, competence, autonomy, satisfaction and interest all higher with AI (d = 0.2–0.5) |
With their bounds attached the two are consistent: collaborating with AI is motivating while it lasts, and losing it costs something. Herzberg predicts the shape of Wu’s result by his own mechanism — three of his six motivators are the work itself, responsibility and achievement, so stripping the challenge out of a job takes intrinsic motivation with it. That reading is a synthesis, not a claim either paper makes.
How solid is this?
- Herzberg is a method problem, not just an old one. It rests entirely on the critical-incident interview — what made you extremely happy or unhappy at work — which invites attribution bias (credit the self, blame the environment) and would produce the 81/69 pattern whether or not the two-factor structure is real. No effect sizes, significance tests or replication failures are reported, and software developers are not among the pooled populations.
- Beecham counts studies, not effects. Its window closes at 2006, before agile and distributed work were normal, and the authors warn the aggregated frequencies “need to be treated with caution”. It found no dominant motivation model in software engineering, and 56% of studies distinguish software engineers from other occupations — so roughly half do not.
- §3 summarises two theories rather than testing them. Self-determination theory (Deci and Ryan) and expectancy theory (Vroom) are given here as vocabulary; how well either predicts behaviour on a software team is still to be verified. Go to their original statements before quoting a number from them.
- Do not stretch the AI studies. Wu’s effect concerns the handoff back to unassisted work, on writing and idea generation, not software. Corgnet’s authors state their result does not extend to replacing a familiar teammate, nor to analytical work; the arms were unbalanced (698 human–human against 393 human–AI) and the paper is not peer-reviewed.
Acknowledgments
This page adapts material from lectures by Eduardo Miranda and David Root [6] on software project management.
References
- S. Beecham, N. Baddoo, T. Hall, H. Robinson, and H. Sharp,“Motivation in Software Engineering: A Systematic Literature Review,” Information and Software Technology, vol. 50, no. 9–10, pp. 860–878, 2008, doi: 10.1016/j.infsof.2007.09.004.
- F. Herzberg,“One More Time: How Do You Motivate Employees?,” Harvard Business Review, vol. 46, no. 1, pp. 53–62, 1968.
- L. L. Thompson,Making the Team: A Guide for Managers, 5, Global. Pearson Education Limited, 2015.
- S. Wu, Y. Liu, M. Ruan, S. Chen, and X.-Y. Xie,“Human-Generative AI Collaboration Enhances Task Performance but Undermines Human’s Intrinsic Motivation,” Scientific Reports, 2025, doi: 10.1038/s41598-025-98385-2.
- B. Corgnet, E. Fumagalli, and S. Rezaei,“Beyond Performance: Replacing a Human Coworker with AI Enhances Motivation,” SSRN Electronic Journal, 2026.
- D. Root,Managing Software Development. Carnegie Mellon University / Innopolis University lecture series, 2014.
Disclaimer: AI is used for text summarization, polishing and explaining. Authors have verified all facts and claims. In case of an error, feel free to file an issue.