Days in the Life of a Senior Postdoc
23.10.2024
Elisa Granato, Sofia van Moorsel, Christoph Ziegenhain
[This article is a follow-up to “Days in the Life of a Postdoc“, where we outlined our lives as newly-minted postdocs.]
Work hours and time management are a recurring topic of discussion in academia. Reasons for this include the expectation for academics to fulfill multiple different roles with little external oversight, combined with an often high degree of freedom with regards to personal schedules. One factor that complicates such discussions is that the vast majority of academics estimate their work hours purely based on their own memory, which is notoriously unreliable.
To combat this, we conducted a time tracking project to more objectively document our work time over several weeks. Back then, we were freshly minted postdocs (1st year post PhD), and our lives were largely dominated by lab work, data analysis and getting our research projects off the ground. While this pilot project ended up being very fun and informative, the insights were somewhat limited due to the short tracking time.
We therefore decided to one-up ourselves: we repeated the time tracking project, but this time we tracked over several months (total weeks tracked: 84 weeks, or 1 year and 7 months) as senior postdocs (5th year post PhD). Since we were much more seasoned researchers now, we were curious to see if and how our schedules and work styles had changed as our jobs evolved over the years. Is it still the same job we originally embarked on? We were also excited to receive time-tracking data from a third researcher (SVM) who is not just a postdoc but also a parent to two small children, which greatly expanded our dataset for this follow-up project. And finally, we have some exciting career updates on where all three of our participants ended up after their postdoctoral time … Keep reading to find out!
METHODS
Using a time tracking app (https://toggl.com/), three postdoctoral researchers in biology at three different universities in Europe tracked their work days in small increments (minimum “session length”: 30 s) over several weeks (42, 30 and 12 weeks respectively). Each session was categorized into one of 12 task categories (see Figures), including at-work break time. At the time of tracking, we were all in the same stage of our career (~5 years post PhD), but worked in different countries and scientific fields: Elisa Granato (EG) – microbial ecology; Oxford, UK. Christoph Ziegenhain (CZ) – single-cell genomics; Stockholm, Sweden. Sofia van Moorsel (SVM) – plant ecology; Zurich, Switzerland. At the time of tracking, SVM had caring responsibilities for two small children, whereas neither EG nor CZ had caring responsibilities.
RESULTS
Total work hours differ strongly between individuals and change over time
In full work weeks (i.e. weeks without holidays), EG and SVM worked an average of ~40 and 43 h, respectively, while CZ worked an average of 53 h (including at-work break time; Fig. 1A). Of note here is the large difference in variance: EG worked anywhere from 23 to 54 h and CZ from 37 to 65 h, while SVM rarely deviated from their average. When looking at how hours were distributed between weekdays and the weekend (Fig. 1B), even more differences become apparent. EG and CZ regularly worked on weekends (purple) and sometimes compensated by taking weekdays (yellow) off. In contrast, SVM largely followed a strict Mo-Fr routine with only a single recorded instance of weekend work. This stark difference was mainly driven by SVM’s childcare responsibilities which severely restricted their schedule.
Compared to our junior postdoc days, EG worked 5 h less in an average week, but still worked most weekends. Their research was largely independent of their colleagues, resulting in a lot of flexibility and incentives to work on weekends when shared resources were more freely available. EG’s personal preference for working shorter days and during weekends also remained strongly reflected in their work patterns. In contrast, CZ worked 2 h more per week compared to a few years ago, and more of their work time has shifted to the weekend. This was likely driven by CZ’s responsibilities becoming more managerial in nature, leading to increased demands on their time by students and colleagues during the week.

Work pattern differences extend to daytime distributions
When looking at the entire dataset for each researcher (Fig. 2), we see several differences in general work patterns. EG (Fig. 2A) and SVM (Fig. 2C) largely followed a traditional 9-to-5 schedule, occasionally punctuated by late evening work. In contrast, CZ (Fig. 2B) started work slightly earlier and finished much later on most days. There is also a pronounced pattern where CZ frequently conducted 1-2 additional hours of work after an evening break.
Compared to when we were junior postdocs, EG’s schedule now exhibited a more pronounced stereotypical postdoc pattern, which often develops organically as one gains more experience: highly irregular hours, late starts in the morning, and frequent last-minute preparation of slides the evening before a presentation (Fig. 2A). CZ’s work patterns also demonstrate a drastic shift over several months of time tracking (Fig. 2B). Where their typical workday was initially mostly comprised of lab work and data analysis, as CZ became more senior, their days were more and more dominated by meetings, administrative work and (grant) writing. This prompted us to delve further into the relative investment into different tasks, which we will explore next.

Different postdocs prioritize different tasks
Fig. 3 depicts how we overall split our time between different work categories. With EG and CZ being wet-lab researchers, “lab work” unsurprisingly took up the biggest chunk of our time (~18% for EG and 27% for CZ; Fig. 3A). Interestingly, for both of us the lab work fraction used to be significantly higher (25% and 32%, respectively) when we were junior postdocs.
Aside from lab work, EG and CZ invested their time very differently. For example, CZ spent around twice as much time (~14%) in meetings compared to EG (~7%). EG spent 8% of their time in seminars, whereas for CZ this number was negligibly small (<1%). Overall, EG’s and SVM’s time investment was generally more evenly distributed across different tasks, whereas CZ more strongly focussed their time on a few different aspects of their work (lab work, meetings, and admin; Fig. 3B).
Notably, and despite the stark differences in work style, CZ and EG both spent an equal amount of time taking breaks at work (~14%; equivalent to ca. 67 min of break time in an 8 h day). Interestingly, average break time has not changed at all from when we were freshly-minted postdocs, suggesting that this represents a consistent need for downtime that is largely independent of the rest of our schedules.

Task prioritization is largely consistent over time
Next, we looked at how our work patterns changed over the course of several months (Fig. 4). For all three researchers, there was a clear oscillation between periods of lots of lab work vs. little to none (Fig. 4A, dark purple). These waves of more focussed lab work lasted from a few weeks to several months and were typically followed by a period of “lab downtime” with little to no lab work conducted. This likely represents a common work pattern among empirical scientists, where many experiments follow each other for e.g. troubleshooting or exploration purposes, after a breakthrough that accelerates the work, or simply because of a looming deadline. This pattern was not visible when we time tracked as junior postdocs, likely due to the much shorter tracking period.
When looking at hours worked per week over time (Fig. 4B), it also becomes obvious that both EG and CZ frequently oscillate between working much less and much more than their respective averages. This was driven by our schedule being largely self-determined, and shaped by balancing the needs of a given project with one’s energy levels and personal life. In contrast, SVM showed little to no variance in weekly work hours (Fig. 4B, right) despite working in a pretty much identical position to EG. This was largely driven by childcare responsibilities, which added a rigid structure to each work day due to school starting and ending the same time every day. This leaves little flexibility for parent-researchers, frequently having to stop working at a given time even if important tasks are still pending.

Academic work days are highly fragmented
Zooming in on individual work days (Fig. 5) allows us to get a sense of how the researchers’ work schedules were structured at very short timescales. Much like when we tracked as junior postdocs, our work days were still extremely fragmented, showing evidence of rapid task switching dozens of times per day. EG’s days appear to be the most fragmented, with CZ and SVM displaying occasional periods of longer focus on specific tasks.

To further assess how often researchers switch tasks during their work days, we looked at uninterrupted session lengths per task category (Fig. 6). For EG and CZ, the median uninterrupted time spent on a task was < 1 h for all categories except seminars. This pattern was already present when we time-tracked earlier in our careers, highlighting a persistent struggle to balance the number of different tasks we are required to perform with the need for deeper, more focussed work. The most obvious outlier in this regard is SVM’s extended session lengths in the “lab work” category. This is due to the nature of their research (experimental plant ecology), which sometimes involves extended blocks of uninterrupted field work.

Senior postdocs invest their time differently than junior postdocs
To explore whether we (had to) prioritize different things now compared to when we just started postdoc-ing, we analyzed changes in relative weekly time investment in different task categories (Fig. 7). For both EG and CZ, we observed big increases in time spent on administrative tasks, reading/searching, and teaching/supervision. This development tracks with the perceived changes to our work lives: as we became more senior, we conducted more managerial tasks and took on a larger load of teaching and supervisory duties. But there were also differences in our respective trajectories. Only CZ displayed large increases in time spent on meetings, e-mails, and writing, while significantly reducing “core” research work such as data analysis. This reflects our diverging career paths while time tracking: EG transitioned into being an “independent research fellow”, which represents more seniority, but little actual change in job description and daily tasks. In contrast, CZ formally transitioned into a junior group leader role which led to bigger changes in their responsibilities and work schedule.

CONCLUSIONS
This extended time-tracking experiment greatly expanded our insights into how we, as now senior postdocs, invest our time. Much like when we time-tracked as junior postdocs, our work days are highly fragmented, our weekly hours highly variable, and a large chunk of our time is spent on non-research tasks like administration, emails, and meetings. (For a deeper discussion on the trials and tribulations of postdoctoral work, see our original article).
For the first time since starting this project, we also had the privilege of receiving data from a third participant (SVM). While we had no “junior postdoc” reference data for them, we are stoked that we were able to include their dataset because it highlighted the hurdles put in front of researchers with caring responsibilities. While EG and CZ often refer to our significant freedom in making our own schedule, SVM was (and is) heavily limited by the school times of their children. This meant that they had to bear the additional load of work time restrictions, while experiencing all the same pressures and productivity incentives as those of us without dependents.
This time tracking project also highlighted the differences in people’s career trajectories. As we progressed from junior to senior postdocs, some of us worked slightly more, some of us slightly less, but we all experienced a significant shift in time investment in different aspects of our work. Most noticeably, as they became more senior, CZ experienced a ~3-, 4- and 6-fold increase in relative workload associated with emails, meetings, and teaching/supervision, respectively. This extends one of the conclusions we drew from our original dataset: not only are no two postdoc jobs the same, – even if they exist in the same scientific field –, but postdoc job descriptions can also change over time. The daily duties of a junior postdoc can be significantly different from those of a senior postdoc, even if the person stays in the same research group (as was the case for EG and CZ).
This pattern of shifting duties to more managerial tasks as one becomes a more senior scientist also directly reflects an ongoing phenomenon in many academic fields, especially the empirical sciences. Young scientists are trained to do highly technical tasks extremely well, yet the only way to progress in their careers at a university is to shift their work to become a group leader / manager, with significantly reduced opportunities to apply their technical skills. We ask ourselves if this kind of “forced managerial” career trajectory is beneficial for science overall, or a design flaw that could be circumvented by a restructuring of research at universities.
To conclude: what we learned from this time tracking experiment is that i) postdocs structure their days very differently depending on their own life and project circumstances; that ii) they tend to work highly variable hours involving rapid task-switching, and that iii) research activities are increasingly diluted by administrative tasks as scientists become more senior in their roles. At this point we would also like to emphasize that our documented schedules are neither “ideal” nor representative of postdocs as a whole. It is simply how we spent our work time over several months, for better or for worse. Everyone’s work life is different, even if we happen to have the same job title, and only you can decide what works best for you and your project. Personally, time tracking turned out to be a highly useful exercise for us. It taught us to reflect more on how we spend our time at work, and we think it will help us manage our work lives more efficiently and sustainably going forward.
WHERE ARE THEY NOW?
At the time of writing this article, the period when we tracked our time is several months behind us. Since all participants were senior postdocs at the time of the experiment, perhaps you are wondering where our careers have taken us. We are happy to report that both the person with the lowest average work hours (SVM) and the one with the longest work hours (CZ) are now group leaders at the universities they did their postdoctoral work at (University of Zurich & Karolinska Institute, respectively). EG has happily left academia to pursue other things (science communication). And now there is nothing left but to leave you with our main take-away from this experiment – there is no one way to success, do what works for you!