Motivation, habit and adherence
Evidence: moderate
The direction of travel is well supported: self-driven motivation and well-built habits predict who keeps training, and a few practical techniques reliably help. The effects are modest, some of the evidence is correlational, and much of it comes from general exercise rather than running.
For most runners the binding constraint is not the quality of the training plan but whether they follow it, week after week, for years. The basics are worth little unless you do them consistently, and the single biggest lever on long-term progress is continuing to show up. That makes motivation and adherence a training topic in its own right, arguably the central one.
Why self-driven motivation lasts and pressure does not
The most useful framework is self-determination theory. It separates motivation that comes from inside, such as running because you enjoy it or because being a runner matters to you, from motivation driven by outside pressure, guilt or reward. Across 66 studies of exercise, adherence was predicted by autonomous motivation and by the satisfaction of three basic needs: autonomy, a sense of choice; competence, a sense of getting better; and relatedness, a sense of connection to others (Teixeira et al. 2012). The practical implication is to build training you actually want to do, that gives regular evidence of progress, and that connects you to other people, rather than training you force yourself through on willpower.
The theory also explains why heavy-handed pressure backfires. In a meta-analysis of 131 samples covering nearly 39,000 athletes, an autonomy-supportive coaching style, one that offers choice and a reason rather than control, was linked to greater wellbeing and more self-driven motivation, while a controlling style was linked to distress and less self-determined motivation (Mossman et al. 2022). Relatedness is also why training with others helps so many people keep going; the social structures around running do much of the adherence work, even where their direct effect on performance is hard to isolate.
Habits do the work willpower cannot
Motivation rises and falls; habit carries you on the days it is low. A habit is a behaviour that a consistent context has made near-automatic, so it no longer needs a fresh decision each time. In a field study of people taking up a new daily behaviour, automaticity built up along a curve that rose steeply and then flattened, reaching its plateau after a median of about 66 days, with a wide spread from roughly 18 days to 254 (Lally et al. 2010). The popular ‘21 days to a habit’ figure is a myth: it takes longer, and how long varies hugely from one person to the next. What builds the habit is repetition in a consistent cue or context, the same time of day or the same trigger, and missing the odd day does not undo the progress, so one lapse is not a failure.
Interventions that deliberately build exercise habits work, modestly. A meta-analysis of 10 randomised trials found a small-to-moderate increase in physical-activity habit strength, stronger in the first few months and weaker later, which suggests habits can fade without continued support (Ma et al. 2023). Tellingly, interventions that taught problem-solving did better, while those that leaned on external rewards did worse. That echoes the warning from self-determination theory that rewards and pressure can crowd out the motivation you actually want.
The idea of an identity-based habit, coming to see yourself as a runner so that running expresses who you are rather than sitting on a to-do list, fits plausibly where self-determination theory and habit formation meet. No trial has isolated and tested it, though, so it is worth holding as a useful frame rather than an established finding.
The small techniques that reliably help
One practical tool with good evidence is the implementation intention: a concrete ‘if this situation, then this action’ plan that fixes in advance when, where and how you will run, such as ‘if it is 7 am on a weekday, I run from the front door’. Across 94 tests, forming such a plan had a medium-to-large effect on whether people carried through, by binding the action to a specific cue so it does not depend on a decision in the moment (Gollwitzer & Sheeran 2006). It uses the same cue-and-context mechanism as habit formation, which is why the two reinforce each other.
Goal-setting is the other, with a catch about which goals. A meta-analysis in sport found small but reliable benefits, and found that the type of goal matters most. Process goals, aimed at the actions within your control, produced by far the largest effects; performance goals produced moderate effects; and outcome goals, such as beating a rival or placing in a race, produced almost nothing (Williamson et al. 2022). So set goals about what you will do, the sessions and the weekly consistency, rather than only about results you cannot fully control. Mental training covers the wider psychology of performance.
Where the honest limits are
The apps that gamify running with points, badges, streaks and leaderboards lean on these mechanisms, and their effect is real but small and short-lived. A meta-analysis of gamified physical-activity interventions found a small-to-moderate boost that weakened substantially once the programme ended (Mazéas et al. 2022), and a larger, high-certainty analysis of gamified health apps found they added only about 489 steps a day, with no effect on moderate-to-vigorous activity (Nishi et al. 2024). External nudges can help you start; the durable motivation is the internal kind, and the durable structure is the habit. Build both, and the plan takes care of itself. Training apps covers what these platforms do and do not deliver.