Does tracking progress make goals more achievable?

A goal changes when it starts giving feedback

A target without feedback can remain abstract. Progress monitoring compares what is happening now with a desired standard: steps against a weekly target, spending against a budget or completed practice against a plan. That comparison can expose a gap while there is still time to respond.

The question is not whether successful people happen to track more. It is whether prompting people to monitor more changes what they subsequently do or achieve.

The review isolated randomized interventions

A 2016 meta-analysis began with more than 22,000 records and retained 138 studies that randomly assigned adults to an intervention designed to increase progress monitoring or to a control condition. The monitoring-frequency analysis included 19,951 participants; the goal-attainment analysis included 18,398.

The goals, tools and outcomes varied widely. Many studies addressed health: physical activity, weight, diet, blood pressure, blood glucose or management of a condition. Interventions could involve diaries, devices, feedback or public reporting, often alongside other behavior-change techniques.

That breadth makes the average informative, but it also means tracking was not one consistent product feature.

Tracking helped on average—not by one fixed amount

The interventions had a very large average effect on monitoring frequency, d+ = 1.98 with a 95% confidence interval from 1.71 to 2.24. Their average effect on goal attainment was smaller: d+ = 0.40, with an interval from 0.32 to 0.48.

Both estimates varied substantially across studies. A trim-and-fill analysis for possible small-study or publication bias reduced the goal-attainment estimate to d+ = 0.19. Among 70 studies meeting the authors’ statistical-power criterion, the estimate was d+ = 0.33.

The responsible summary is therefore directional rather than absolute: interventions that increased monitoring improved attainment on average, while the likely magnitude depends heavily on the evidence and design included.

How feedback was designed mattered

Effects differed by domain, measurement and study quality. Goal-attainment effects were larger when outcomes were measured objectively rather than by self-report. Studies with stronger blinding tended to produce smaller effects. Some interventions combined monitoring with goal setting, action planning or immediate feedback, making it difficult to assign every gain to measurement alone.

Cross-study moderator estimates were larger when progress information was made public (d+ = 0.55) or reported to someone else (d+ = 0.47) than when kept private (d+ = 0.19). Physically recording it was also associated with a larger estimate than not recording it (d+ = 0.43 versus 0.29).

These rows summarize moderator comparisons across different studies; they do not isolate the feature through randomized head-to-head tests.

Several other distinctions did not significantly moderate attainment: tracking behavior versus outcomes, monitoring rate versus distance from a goal, and passive versus active monitoring.

The study-level mediation analysis was consistent with more frequent monitoring carrying part of the intervention effect to attainment. But only a subset of studies provided the necessary correlations, and aggregated mediation cannot reconstruct the process for every individual participant.

No pooled effect can promise your result

Most included evidence concerned health-related goals, and some domains had very few studies. Follow-up periods, comparison conditions and target outcomes were mixed. In control groups described as no monitoring, spontaneous monitoring was not always measured, so the analysis sometimes assumed it was zero.

The review cannot tell whether an intervention also added burden, anxiety or pressure, or changed the experience of an activity. Attainment and wellbeing are related, not identical.

Track only what can change the next move

Our editorial inference: a tracker is most defensible when its purpose is explicit. Measure the behavior or outcome that can inform a decision, at a frequency that leaves room to act. If the number changes nothing, it may be observation without feedback.

The open question is not whether measurement can work. Across randomized studies, it often did. The sharper question is what else measurement changes while a target outcome improves—including whether an enjoyable activity begins to feel like work.

Standardized mean difference · d+

All included studies95% CI 0.32 to 0.48

0.40

Adequately powered studies70 studies

0.33

Trim-and-fill sensitivity95% CI 0.10 to 0.28

0.19

Harkin et al. (2016), 138 randomized studies. The authors cautioned that substantial heterogeneity limits any single expected effect.