TL;DR. The original version of this article listed ten morning routines and put the phrase “according to research” under nearly every one. In this update we went looking for that research. What came back was more instructive than the list itself. For most of the ten items we could not verify a primary source supporting the claim as stated. The strongest evidence we did find belongs to the morning exercise item: a 2024 Bayesian meta-analysis synthesising 651 effect sizes from 113 studies and 4,390 participants finds a small benefit of acute exercise on cognition, g = 0.13. That sits below Cohen’s threshold for a small effect, which is 0.20. More striking is item eight. The famous 2009 study behind the advice to stop multitasking did not survive two replication attempts and a meta-analysis published in 2017. This article no longer gives you ten pieces of advice. It gives you a table and a method for auditing your own morning routines by strength of evidence.
The first rule of running yourself like a company is knowing who wrote your operating system. Morning routines are the most crowded corner of the advice market, and most of the claims in that corner circulate in sentences whose sources have never been opened.
Why we rewrote this article
The original text was published in 2015 and carried the list format of its era: ten items, a confident justification under each, a source under none.
A claim without a source is not necessarily wrong. But it is unverifiable. And building your daily architecture on an unverifiable claim is the opposite of learning like a student.
So instead of deleting the article, we audited it. The table below takes each of the ten items from the original list and answers a single question for each: could we find a named, openable primary source supporting this claim?
The evidence audit of the ten routines
| # | Routine in the original list | Verified primary source found? | What the evidence says |
|---|---|---|---|
| 1 | Start the day with exercise | Yes | 2024 Bayesian meta-analysis: acute exercise has a small beneficial effect on cognition, g = 0.13 ± 0.04. An analysis restricted to executive function tasks shows improvements in working memory and inhibition |
| 2 | Drink lemon water in the morning | No | We could not verify a primary source for the claims that it “sharpens energy” or “increases nutrient absorption”. Treat as folklore |
| 3 | Stay off screens until breakfast | No | We could not verify a primary source for this specific claim. The attention and notification literature is large, but the “until breakfast” threshold does not come out of it |
| 4 | Eat a real breakfast | No, not as stated | The relationship between breakfast and cognition is a contested area; we could not verify a primary source supporting the certainty of the original text |
| 5 | Set your daily goals | Partly | Goal setting has a deep literature in organisational psychology, but we did not verify a primary source here supporting the specific morning to-do-list claim |
| 6 | Keep your workspace clean | No | No verifiable primary source found |
| 7 | Keep morning meetings under control | No | No verifiable primary source found. A reasonable operational suggestion, but not an evidence claim |
| 8 | Do not multitask | Yes, and it points the other way | The 2009 study behind this advice was tested in 2017 by two replication studies and a meta-analysis. Only 5 of 14 tests were significant, and only 2 of those survived a more conservative Bayesian analysis; the meta-analysis across 39 effect sizes became non-significant after correction for small-study effects |
| 9 | Learn to say no | No | No verifiable primary source found |
| 10 | Put it all together | Not applicable | Not an evidence claim |
CEOtudent editorial framework. The columns record the result of the source search carried out during this update on the items of the original 2015 list. “No verified primary source found” does not mean the claim is false; it means that in this update we could not reach an openable primary source supporting it as stated.
Out of ten items, two have a verifiable primary source, and one of those undercuts the advice rather than supporting it. That is not a flaw peculiar to this article’s original version. It is a description of how the advice format works.
A closer look at item one: how much does exercise actually help?
Start with the item that has evidence, because the evidence is more interesting than the advice.
The systematic review and Bayesian meta-analysis published in Communications Psychology in 2024 synthesised 651 effect sizes from 113 studies, representing 4,390 participants, to determine the impact of acute exercise on cognitive task performance in healthy young adults. Publication bias was mitigated using the trim-and-fill method.
The finding: acute exercise has a small beneficial effect on cognition, g = 0.13 ± 0.04, with a Bayes factor of 3.67, and it decreases reaction time. A meta-analysis restricted to executive function tasks revealed improvements in working memory and inhibition.
Now let us translate what that number means, because “small beneficial effect” on its own changes no decisions.
| Measure | Value | What it means |
|---|---|---|
| Pooled effect size (Hedges g) | 0.13 | Cohen’s threshold for “small” is 0.20; this is roughly 65% of it |
| Approximate range | 0.09 to 0.17 | Based on the reported standard error; excludes zero, but sits entirely in small-effect territory |
| Cohen’s U3 | 55.2% | The average person who exercised outperforms about 55% of those who did not |
| Common language effect size | 53.7% | The probability that a randomly chosen exercised person beats a randomly chosen non-exercised one |
CEOtudent editorial framework. The value g = 0.13 ± 0.04 is taken verbatim from the source meta-analysis. U3 and the common language effect size are our conversions via the standard normal distribution and are not reported in the source study.
The last two rows make visible what morning routine advice usually hides. A coin flip is 50%. The best estimate from this literature moves you to roughly 54 to 55%.
This is not an argument against exercising. The health case sits entirely outside this discussion and is far stronger. It is an argument that any text selling morning exercise as a lever that will transform your day is claiming more than the best available evidence supports.
For someone managing themselves, the right reading is this: you are buying a small but real effect at a small but real cost. That can be a good trade. But the reason people quit in week three is that the advice was oversold going in.
A closer look at item eight: what was the multitasking advice based on?
“Do not multitask” may be the most repeated sentence in productivity writing. Its source is usually a single study: Ophir, Nass and Wagner, published in the Proceedings of the National Academy of Sciences in 2009. It reported that heavy media multitaskers were more susceptible to interference from irrelevant stimuli and performed worse on a test of task-switching ability.
In 2017, Wiradhany and Nieuwenstein tested that finding in Attention, Perception and Psychophysics. They ran two replication studies comprising 14 tests with an average replication power of 0.81. Of those 14 tests, only five yielded a statistically significant effect in the direction of increased distractibility for people scoring higher on the media-use questionnaire. Only two of those effects held under a more conservative Bayesian analysis.
They then ran a meta-analysis including all published studies on the relationship, covering 39 effect sizes. The result was a weak but significant association that turned non-significant after correction for small-study effects. In the authors’ own words, these findings lead them to question the existence of an association between media multitasking and distractibility in laboratory tasks of information processing.
The conclusion is not “multitask all you like”. It is narrower and more useful: the evidence behind this advice is not as firm as it was presented to you. You need to observe your own distractibility and make your own call, because the literature has not made it for you.
What to do instead
The lesson from the audit is not to replace ten routines with ten different ones. It is to change how you choose a routine.
One. Set the routine up as an experiment, not a rule. Run one item for two weeks, measure a single outcome, and write that outcome down in advance. If you pick what to measure afterwards, every routine looks like it worked.
Two. Ask about effect size, not direction. The answer to “does it help” is almost always yes and almost always meaningless. The question “by how much”, as the exercise example shows, turns the advice into something entirely different.
Three. Do not put unsourced claims at the core of your routine. An unsourced suggestion can be tried, but it cannot be the beam holding up your day’s architecture. The eight rows marked “no” above are a list of things worth trying, not a list of evidence.
Four. Write down the cost too. Morning exercise costs you twenty minutes and some willpower. A gain of g = 0.13 may or may not be worth that cost. Nobody can run that calculation for you, because the cost side is entirely specific to your life.
Five. Watch what happens when you stop. Removing a routine is more informative than adding one. The effect of adding gets tangled up with expectation; the effect of removing does not.
What changed in this update
For transparency: the 2015 original supported each of the ten items with the phrase “according to research” and gave a source for none of them. In this update those sources were searched for, the ones found are named, and the ones not found are explicitly marked as not found. No statistic whose source could not be verified was republished.
The main idea of the article changed too. It no longer says “do these ten things”. It says “know how to audit it when you are told ten things”. That is exactly where running yourself like a CEO and learning like a student intersect: demanding the justification for every rule that enters your own operating system.
Frequently asked questions
Should I stop exercising in the morning?
No. The finding is that the cognitive benefit is smaller than advertised, not that it is harmful. The health case for exercise is independent of this discussion and far stronger. The only thing that changes is how much weight you put on morning exercise as a productivity lever.
Are the items with no source found false?
We do not know, and the table says exactly that. “No verified primary source found” and “false” are different things. Not conflating them is the discipline this article is recommending.
Is g = 0.13 really negligible?
Not negligible, small. At population scale, small effects matter. At individual scale, it may not be large enough to justify redesigning your morning around it. The decision turns on the ratio of gain to cost, not on the size of the gain alone.
Is the multitasking advice worthless then?
No. The replication work questions a specific laboratory association, not every experience of distraction in daily life. The right conclusion is not to abandon the advice but to stop presenting it as a settled scientific finding.
How do I test my own routine?
Change one routine, write down one metric in advance, run it for two weeks, then stop and see what happens. If you change three things at once, you will never learn which one worked.
Sources
- Communications Psychology, 2024. Systematic review and Bayesian meta-analysis of the effect of acute physical activity on cognition in healthy young adults: 651 effect sizes from 113 studies, 4,390 participants; pooled effect g = 0.13 ± 0.04, Bayes factor 3.67; publication bias addressed with trim-and-fill.
- Wiradhany and Nieuwenstein, Attention, Perception and Psychophysics, 2017. Cognitive control in media multitaskers: two replication studies and a meta-analysis. 5 of 14 tests significant, 2 surviving Bayesian analysis; meta-analysis across 39 effect sizes non-significant after correction for small-study effects.
- Ophir, Nass and Wagner, Proceedings of the National Academy of Sciences, 2009. Cognitive control in media multitaskers; the original finding tested by the replication work above.
- Jacob Cohen, work on statistical power analysis for the behavioural sciences; the source of the effect size thresholds and the U3 measure.
- World Health Organization, guidelines on physical activity and sedentary behaviour; for the health case for exercise.
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
