Our current workplace - with its increasing digitalization, remoteness, connectivity, flexible work schedules, and large proportion of gig workers - places conflicting demands on us. Because of this we are forced to deal with many trade-offs and make sometimes quite tough choices and compromises that have a direct impact on our productivity and well-being. One of the most prominent trade-offs that characterize the current workplace is a tension between responsiveness and focus time. The former builds trust, the latter is necessary for high quality work. We will examine in this blog post whether computer science and data could help us a little bit with alleviating the tension caused by this specific trade-off.
Our everyday trade-offs
In your personal and professional life, wherever you look, sooner or later you will hit upon some kind of trade-off you have to deal with on a regular basis. We can borrow some examples of such problems from Brian Christian and Tom Griffiths’ book Algorithms to Live By (2016), that tries to look at our common everyday challenges through the lens of computer science:
- How long should I hunt for a new apartment to minimise risk of missing a good opportunity? Are we running the risk of stopping the hunt too early or too late? [example of optimal stopping problem]
- Should we experiment and try new restaurants, or revisit our tried and tested? [example of exploring/exploiting trade-off]
- Is it reasonable to invest part of my valuable and scarce time resources into sorting my email inbox? [example of sorting/searching trade-off]
- How do I schedule my working day to be able to do as many tasks as possible, as quickly as possible? [example of scheduling problem]
- How to plan my tasks when there are so many that I am completely stuck, and unable to complete them? [example of trashing as a result of trade-off between work and metawork]
What is difficult about these trade-offs (and about trade-offs in general) is that one has to choose or make a compromise between outcomes that are all desirable and worth achieving. An additional source of difficulty for this type of problems lies in the fact that optimal choice or compromise may differ, depending on the circumstances under which we make our decision.
In the case of the specific trade-offs listed above we are lucky that there are some algorithms or heuristics that make these trade-offs “solvable”. This means that we can consult some rigorous and data-based recipes for dealing with these challenges, with guarantee of near optimal or, at least, good enough outcomes that are achievable given our limited resources, be it time, information or cognitive power. You can check some of these recipes below and even try to implement them in your everyday life:
Hunting for an apartment:
- Use the Look-Then-Leap rule with 37% threshold value. Gather data for the first 37% of total search time (e.g. 22/60 days) to establish a benchmark, and then make a decision as soon as you find an option better than anything you found in that first 37% of your search.
Choosing a restaurant:
- The answer depends on the time interval, i.e. how long after making the decision will you be going out for a meal. The longer this interval, the more worthwhile it is to explore new options, because it gives you the opportunity to discover something that will be significantly more valuable than what you have been able to discover so far, and you will be able to re-use it repeatedly later. On the contrary, the shorter the interval, the more worthwhile it is to exploit known options - when exploring you are more likely to discover something that is less valuable than what you have discovered in the past.
Sorting your inbox:
- You need to think about it in terms of relative time costs of sorting and future searching because sorting is nothing more than prophylaxis for searching in the future. Thus, sorting makes sense only when its time costs are lower than the time costs of searching. From this accounting perspective, sorting emails is not immensely productive, the same goes for sorting your domestic bookshelf.
Prioritising your tasks:
- If your goal is to do as much as possible in the shortest possible time, then rely on the Weighted Shortest Processing Time algorithm, which minimizes the sum of completion times. You should “divide the weight of each task by how long it will take to finish, and then work in order from the highest resulting importance-per-unit-time to the lowest. [...] Each time a new piece of work comes in, divide its importance by the amount of time it will take to complete. If that figure is higher than for the task you’re currently doing, switch to the new one; otherwise stick with the current task.” (Christian & Griffiths, 2016, p. 109, 118) You may recognize in this recommendation an enhanced version of the “two-minute rule” popularized by a time management method called Getting Things Done (“Whatever you can do till two minutes, do it now.”).
- When trying to achieve the goal described in the previous paragraph, be aware of the costs caused by switches between tasks and contexts. If you are constantly switching between tasks, actual work is usually not carried out. Recall the minutes spent reprioritising, the time spent preparing a new assignment, orienting where you finished off before picking it up again. And even if the work is done under these circumstances, then, as demonstrated by task-switching experiments, it is often at the expense of slower work rate and higher error rate. In terms of computer science, you are at risk of thrashing. When the number of tasks you are simultaneously working on exceeds a certain amount, the metawork caused by the task prioritization and coordination, combined with the switching itself, severely cripples the time you have for said tasks. To prevent this, regulate and reduce the number of incoming tasks (i.e. learn the art of saying no) and/or to make the prioritization metatask simpler (e.g. answering emails in whatever order they happen to appear in the email inbox).
As mentioned in the introduction of this blog post, the current workplace is, besides other things, characterized by trade-off between responsiveness to others that builds trust on one side, and the focus time one needs for doing high quality work on the other side. Are there some insights from computer science that could help us to deal with this trade-off more successfully? Fortunately, the answer to this question seems to be positive, as computer science deals with analogous problem with responsiveness and throughput: “How quickly a computer can respond to things?” vs. “How much computer can get done overall?” For example, when a computer is monitoring and fulfilling user’s requests as quickly as possible while trying to calculate all the specified tasks in the background. Let’s check the details below.
First of all, it is good to be aware of the fact that the need to have enough focus time during working hours is not spread evenly across jobs. It depends on the nature of the role’s responsibilities. There are jobs that need less focus time (e.g. salesmen or call-center operators) and there are jobs where it is a crucial prerequisite for doing high-quality work (e.g. programmers or higher management). Actually, this fact represents one possible solution to the responsiveness/throughput trade-off: having specialized job positions where it is specified in advance that responsiveness will be highly favored at the expense of throughput. That’s why many companies have call-centers where one of the most important KPIs is the First Response Time (some data suggests that the companies responding to emails within an hour are 7x more successful than those that take two hours, and they are 60x more successful than companies that take a day or longer to respond).
Outsourcing responsiveness is thus one viable option for companies that wish to reduce the compromises employees have to make on their focus time.
Divide and conquer
Unfortunately, the above solution is relevant only when you don’t have to make the decision about a specific form of compromise yourself. Then you need to take a different, personal approach. Here we can be inspired by a solution for computers that defines a minimal period of time during which every program is guaranteed to be executed at least a little bit. When there are too many programs shrinking this period under the prespecified minimal length, the waiting time for execution of some programs simply gets longer, but with guarantee that they will be processed eventually, at least to a small degree.
“Establishing a minimum amount of time to spend on any one task helps to prevent a commitment to responsiveness from obliterating throughput entirely: if the minimum slice is longer than the time it takes to context-switch, then the system can never get into a state where context switching is the only thing it’s doing.” (Christian & Griffiths, 2016, p. 124)
Many of you probably recognize in the above description the famous and popular Pomodoro Technique or Timeboxing that both recommend to break down work into intervals, separated by short breaks, and work on a given task until the specified time interval expires. The question is, how long should these intervals be?
In the case of the Pomodoro Technique, they are traditionally 25 minutes in length, but the more appropriate answer would be that “it depends”. When it comes to computers, the interval should not be longer than a time wherein the human user is aware of a delay in responses to her requests. Just think of how frustrating it is when your computer's response is slower than your expectations and requirements. When it comes to humans, the minimal acceptable level of responsiveness will differ by person depending on many factors: from the industry she is working, to her role in the company, right down to the specifics of the project in question. All of these variables (and many more) factor into what you should consider to be the minimal acceptable limit for responsiveness.
But the principle is clear: “[You] should try to stay on a single task as long as possible without decreasing your responsiveness below the minimum acceptable limit. Decide how responsive you need to be - and then, if you want to get things done, be no more responsive than that.” (Christian & Griffiths, 2016, p. 125)
Rely more on batch processing
When dealing with a large amount of tasks and frequent task switching, one can utilize two related concepts coming from computer science: batch processing and interrupt coalescing. In the realm of computers, these concepts refer to the situation when computers “wait until some fixed interval and check everything, instead of context-switching to handle separate, uncoordinated interrupts from their various subcomponents”. (Christian & Griffiths, 2016, p. 125) When transposed into the world of human workers, this design principle can manifest in checking emails every two or three hours instead of continuous handling of all incoming emails. Such arrangement prevents fragmentation of our time and provides us with more focus time we need for deep work and experiencing flow. According to HBR, people check their email inboxes every 37 minutes, 15 times per day. Given that it can take 23 minutes to fully focus again after some distraction, there is apparently a lot of room for improvement.
A similar approach can also be applied to instant messaging, like Slack or MS Teams chat. These have taken over internal communication in many companies, and have consequently caused ubiquitous complaints regarding workplace distractions. The positive effect of batch processing of instant messages on a person's focus time is illustrated on Graphs 1 and 2. On Graph 1, there is a distribution of Slack messaging over one working day of one person and its impact on her estimated focus time. You can see that this person’s overall focus time is relatively low, because she sent messages consistently throughout the whole day. Compare that with the focus time of another person on Graph 2, where the distribution of the very same number of messages was more batched.
Just from visual inspection of the graphs it is clear that the second person has a higher chance to get into the “zone”. To take advantage of this batching effect one can make more use of the option to turn off notifications for some periods of time. Another valuable tool is for teams to make an agreement about their messaging strategy, for example, that one should not send messages continuously, but rather in batches after there are more of them.
Graph 1: The evolution of one person's focus time over the course of one working day as a function of interruptions caused by Slack messaging (57 messages between 8h00 and 16h30). Overall focus time of this person is relatively low, because she sent messages consistently throughout the whole working day. Compare that with the focus time of a person who relied more on batching her messaging displayed on Graph 2.Graph 2: The evolution of one person's focus time over the course of one day as a function of interruptions caused by Slack messaging (57 messages between 8h00 and 16h30). In comparison with the person from Graph 1 this person more batched her messaging and as result she experienced longer periods when she could deeply focus on her work.
Graph 2: The evolution of one person's focus time over the course of one day as a function of interruptions caused by Slack messaging (57 messages between 8h00 and 16h30). In comparison with the person from Graph 1 this person more batched her messaging and as result she experienced longer periods when she could deeply focus on her work.
The ‘positive feedback loop’ trap
Another concept from computer science that may help us better manage the consequences of a responsiveness/throughput trade-off is a positive feedback loop. Its formal definition is that it is a process that occurs in a feedback loop which exacerbates the effects of a small disturbance. You may have experienced this phenomenon recently during your, now so frequent, online meetings, when your ears were "tortured" by unpleasantly loud, squeaky sounds. This is caused by a sound loop between an audio input (a microphone) and an audio output (loudly-amplified loudspeaker). Another well-known example of a positive feedback loop is the Matthew effect, named after the biblical wisdom concisely expressed by aphorism "the rich get richer and the poor get poorer". This effect manifests in many areas of life, from social networks and crowd behavior, to reading acquisition and climate change.
But for the responsiveness/throughput trade-off it would be probably more appropriate to use as an illustration a quote from Laura Numeroff’s book for small children: “If you give a mouse a cookie, he’s gonna ask for a glass of milk.” High or low responsiveness forms a feedback loop - the more responsive I am, the more people expect me to be responsive. This expectation drives the volume of messages/emails I will receive from colleagues, all with the expectation of a fast response. The opposite applies in reverse, also. This way a positive feedback loop makes the trade-off harder to balance. That is why building a careful digital culture for Slack/email usage is important, as it provides more control of the within-company distribution of "talkers" and "zoners" (a word my colleague, Martin Vrany, uses for those who shut away others when in the “zone”).
Shaping and moulding your own digital culture
Not all of us are lucky enough to work in an environment or with such digital tools that would resolve the responsiveness/throughput trade-off for us, or act in your favor. It is usually our responsibility as individuals, teams and organizations, to define and continuously tune our own digital culture. This digital culture should consist of fairly explicit rules, with managed expectations, regarding the use of digital tools in supporting each employee’s ability to achieve their goals, and ideally also their well-being. Below are few examples of such rules and expectations, relevant to the responsiveness/throughput trade-off:
- Do not use emails for instant messaging.
Employees should not be expected to respond to emails within less than 90 minutes, as reacting to email more frequently can break their focus. It should not, however, exceed 4 hours, to ensure some minimal actionability.
- Refine external email responsiveness.
Typical response times to external emails should not exceed 12 hours, so we do not jeopardize trust of our clients or partners.
- Do not abuse the ‘cc’ field.
Emails should be copied only to those people for whom the email is intended or for whom it is important.
- Use threads to organize discussions.
Use Slack threads to enhance readability and ensure that people aren’t disrupted by irrelevant discussion.
- Update your status appropriately.
Inform other Slack users at what times you are available on Slack by using away/active toggle, so that they know when they can engage in conversation with you.
- Block out focus time.
Utilize blockers in your calendar when you need time for focused work. This way others can more easily respect and protect the time you need for accomplishing important and complex tasks.
The specific rules and expectations may differ depending on the industry, company, department and job position. Generally, though, when a sufficient number of employees will follow similar kinds of rules, there is a higher chance that employees will achieve a balance between responsiveness and focus appropriate to their specific conditions.
Measure what you want to manage and improve
When trying to implement a productive digital culture and make its rules and expectations routine, it is always useful to have some tool for answering questions like “Where are we?”, “Where do we want to be?”, and “Are we moving in the right direction?”. Ideally, we should have some hard data in our hands to answer these questions. Fortunately, digital collaboration tools provide enough data that could be anonymized, analyzed and then used for continuous reporting. Some examples of such analytics can be seen on Figures 1-3 with screenshots taken from Time is Ltd. analytics platform.
Fig. 1 shows an alert view that breaks departments into different groups based on their responsiveness to internal emails, and indicates where action might be needed. Given the rule described above, employees may need to increase their responsiveness to colleagues in other departments, or instead place more emphasis on their focus time. In the latter case, they should not check emails as they land in their email inboxes, but instead batch their replies and respond when it is more convenient for them.
Fig. 2 shows an equivalent alert view for responsiveness to external emails. There are situations here where the appropriate external email responsiveness times are at stake, and opportunity knocks for praising good practice or offering support and advice to struggling teams.
Finally, on Fig. 3 you can see departmental statistics and on Fig. 4 the company statistics on the proportion of working time blocked out in employees’ calendars. The shown view informs the user to what extent employees utilize recommended practice of blocking their focus time, to work on more complex tasks in their calendars.
Similarly, other rules and expectations that define digital culture in a given company can be connected through data with actual employees’ behavior. This way, data from digital collaboration platforms provides us with an important “source of truth” in ongoing discussions about how the current digital culture works, and how to improve it.
Rethinking time management as change management
So where does this all leave us? If you are already somewhat informed about the principles and methods of modern time management, you should not be very surprised by recommendations described in this blog post. While it is reassuring that they are supported by rigorous, data-based findings from computer science, that by itself will not make our lives better or easier. Apparently we already have the tools we need to make the difference, but we still need to drive their everyday use.
Maybe the problem we are facing is more about change management than time management. Maybe instead of studying new books about the latest methods of organizing time, we should look for inspiration and advice in books like Reconsidering Change Management, Think Small or Atom Habits. It seems that we have a topic for some of our next blog posts. So stay tuned.
What are the trends in the current workplace that fight for our attention? Read our whitepaper Paying Attention to Attention Pays Off, Protecting Scarce Sources of Focus Time in the World of Digital Exhaustion and Distraction.
At Time is Ltd., we measure digital collaboration and productivity, without ever sacrificing employee privacy. We provide an advanced analytical SaaS platform that delivers a holistic view of an organization collaboration patterns. We measure your team’s digital footprint to improve communication, productivity as well as save precious time. Our approach only aggregates meta-data from a variety of data sources, to show how your teams work with your collaboration tools so you can get them more productive and motivated.
Algorithm. (n.d.). Retrieved November 3, 2020, from https://en.wikipedia.org/wiki/Algorithm
Allen, D. (2015). Getting Things Done: The Art of Stress-Free Productivity. New York: Penguin Books.
Bank run. (n.d.). Retrieved November 10, 2020, from https://en.wikipedia.org/wiki/Bank_run
Bariso, J. (2019, July 3). How Long Should It Take to Respond to an Email? Exactly This Long, Says a Successful CEO (But Is He Right?). Retrieved November 9, 2020, from https://www.inc.com/justin-bariso/how-long-should-it-take-to-respond-to-an-email-exactly-this-long-says-a-successful-ceo-but-is-he-right.html
Batch processing. (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Batch_processing
Christian, B., & Griffiths, T. (2016). Algorithms to live by: the computer science of human decisions. New York: Henry Holt and Co.
Cirillo, F. (2013). The Pomodoro Technique. Berlin: FC Garage GmbH.
Clear, J. (2018). Atomic Habits: An Easy & Proven Way to Build Good Habits & Break Bad Ones. New York: Avery.
Deep Work. (n.d.). Retrieved November 6, 2020, from https://www.calnewport.com/books/deep-work/
First Response Time. (n.d.). Retrieved November 6, 2020, from https://www.geckoboard.com/best-practice/kpi-examples/first-response-time/
Flow (psychology). (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Flow_(psychology)
Glei, J. (2016, October 24). Jocelyn Glei on how email is killing our productivity. Retrieved November 9, 2020, from https://www.theguardian.com/guardian-masterclasses/guardian-masterclass-blog/2016/oct/24/jocelyn-glei-on-how-email-is-killing-our-productivity
ten Have, S., ten Have, W., Huijsmans, A.-B., & Otto, M. (2016). Reconsidering Change Management: Applying Evidence-Based Insights in Change Management Practice. London: Routledge.
Heuristic. (n.d.). Retrieved November 3, 2020, from https://en.wikipedia.org/wiki/Heuristic
Interrupt coalescing. (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Interrupt_coalescing
Job shop scheduling. (n.d.). Retrieved November 9, 2020, from https://en.wikipedia.org/wiki/Job_shop_scheduling
Matthew effect. (n.d.). Retrieved November 10, 2020, from https://en.wikipedia.org/wiki/Matthew_effect
Mean shift. (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Mean_shift
Metawork. (n.d.). Retrieved November 6, 2020, from https://www.urbandictionary.com/define.php?term=metawork
Multitasking: Switching costs. (2006, March 20). Retrieved November 6, 2020, from https://www.apa.org/research/action/multitask
Newport, C. (2016). Deep Work: Rules for Focused Success in a Distracted World. New York: Grand Central Publishing.
Numeroff, L. (2015). If You Give a Mouse a Cookie. New York: HarperCollins Publishers.
Optimal stopping. (n.d.). Retrieved November 9, 2020, from https://en.wikipedia.org/wiki/Optimal_stopping
Plummer, M. (2019, January 19). How to Spend Way Less Time on Email Every Day. Retrieved November 6, 2020, from https://hbr.org/2019/01/how-to-spend-way-less-time-on-email-every-day
Pomodoro Technique. (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Pomodoro_Technique
Positive feedback. (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Positive_feedback
Robison, J. (2006, June 8). Too Many Interruptions at Work? Retrieved November 6, 2020, from https://news.gallup.com/businessjournal/23146/too-many-interruptions-work.aspx
Service, O., & Gallagher, R. (2017). Think Small: The Surprisingly Simple Ways to Reach Big Goals. London: Michael O'Mara Books.
Single source of truth. (n.d.). Retrieved November 9, 2020, from https://en.wikipedia.org/wiki/Single_source_of_truth
Sorting algorithm. (n.d.). Retrieved November 9, 2020, from https://en.wikipedia.org/wiki/Sorting_algorithm
Stafford, T. (2012, June 6). Fundamentals of learning: the exploration-exploitation trade-off. Retrieved November 9, 2020, from https://tomstafford.staff.shef.ac.uk/?p=48
Stanovich, K.E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21, 360-407.
The rich get richer and the poor get poorer. (n.d.). Retrieved November 10, 2020, from https://en.wikipedia.org/wiki/The_rich_get_richer_and_the_poor_get_poorer
The Study of Earth as an Integrated System. (n.d.). Retrieved November 10, 2020, from https://climate.nasa.gov/nasa_science/science/
Thrashing (computer science). (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Thrashing_(computer_science)
Timeboxing. (n.d.). Retrieved November 6, 2020, from https://en.wikipedia.org/wiki/Timeboxing
Udemy (2018). Udemy In Depth. 2018 Workplace Distraction Report. Retrieved November 6, 2020, from https://research.udemy.com/research_report/udemy-depth-2018-workplace-distraction-report/
Viral video. (n.d.). Retrieved November 10, 2020, from https://en.wikipedia.org/wiki/Viral_video
Wallace, B. (2020, February 11). How Email Responsiveness Builds Trust [Infographic]. Retrieved November 6, 2020, from https://www.socialmediatoday.com/news/how-email-responsiveness-builds-trust-infographic/572022/
Whittaker, S., Matthews, T., Cerruti, J., Badenes, H., & Tang, J. (2011, May).Am I wasting my time organizing email? A study of email refinding. Paper presented at Proceedings of the International Conference on Human Factors in Computing Systems, Vancouver, BC, Canada. Paper retrieved November 6, 2020, from https://drive.google.com/file/d/1LSi-ZIAvWGjLxx0sxCMp1bwPtAfVw-nd/view