.png)
Running Someone Else's Inbox: The Research on Triage, Voice and the Line You Agree in Advance.
It is 7.12 on a Monday and you open her inbox before you open your own. Two hundred and fourteen things arrived overnight. Four say urgent in the subject line and two of those are newsletters. A supplier wants a decision by noon and has helpfully put the deadline in bold. The chair's assistant has sent one mild sentence that anyone who has done this job for a fortnight reads as a small fire. And somewhere in the middle of it sits the message that will decide how Thursday goes, written by somebody who did not think it worth flagging.
You have thirty-eight minutes before the first call and your own work has not started. The thing nobody outside this profession quite understands is that you are not reading these messages; you are deciding, two hundred and fourteen times, what each one is going to cost the person you support.
Running someone else's inbox is one of the highest-trust tasks in any organisation and one of the least examined. There is a serious body of research on email, attention, urgency and how people read tone in writing, and almost none of it was written with you in mind. This post puts it where it belongs, in the hands of the person holding the keys.
What you will get out of this
What the glance actually costs, and the finding that interrupted people work faster and pay for it somewhere else
Why the advice on batching keeps contradicting itself, and the four studies that explain when it works
What five days without email did to thirteen people's heart rate variability
What people mean when they write urgent, from the research on urgency, deadlines and how long a silence can last before it becomes a message
A triage framework built on the email research rather than productivity folklore, including the filing habit the evidence says is not worth your time
What delegated access does to trust and privacy, and the four rules worth agreeing before you ever open somebody's mail
Why tone does not survive the journey, from five experiments where sarcasm over email landed no better than a coin toss
The handover rituals that stop an inbox quietly becoming a second job
The inbox check, a tool that reads the inbox you run, and the Inbox Pages, three pages for the triage, the drafting standard and the agreement with your executive
What the glance actually costs
Start with what has been measured, because the numbers are more interesting than the folklore. In 2005 three researchers sat with twenty-four information workers and recorded what they did, minute by minute, at their own desks. Work turned out to be highly fragmented: people spent very little time in a working sphere before switching, 57 per cent of those spheres were interrupted, and although most interrupted work was picked up again the same day, more than two other activities happened in between. That is a study about what a working day looks like when anything can arrive at any moment, which is a fair description of the job.
Then the finding that changes how you think about the subject. In 2008 a laboratory study asked whether the content of an interruption mattered. It did not. What the researchers found instead was that people completed interrupted tasks in less time than uninterrupted ones, with no difference in quality. We compensate, and the price showed in the other measures: more stress, higher frustration, more time pressure and more effort. The work gets done. Something else absorbs the cost.
The interruption rarely shows up in the output; it shows up in the person.
Part of what carries that cost has a name. Attention residue is the part of your mind that stays with the first task after you have moved to the second, and a set of studies found it reduced performance on whatever you turned to next. Later experiments added the detail that matters most in this job: residue was worst when people expected to come back to the interrupted work under time pressure, and a small intervention fixed it. Before you switch, take a few seconds to note where you were and what you will do first when you return. They called it being ready to resume, and it is close to free.
The email-specific numbers are older and still recognisable. A study at a UK company in 2003 found most employees had their software checking for new messages every five minutes and responded to a message arriving within six seconds. Six. There was a recovery period afterwards, shorter than for a telephone call but real, and a later study at a car rental firm found a typical task took about a third longer when email interrupted it.
The part I find most useful is about perception rather than performance. A study comparing what monitoring software recorded with what the same people said in a survey found that users drastically underestimate the disruptive effect of email on their own work. And laboratory studies found that a notification on its own, with no reply and no unlocking of the phone, disrupted an attention-demanding task by an amount comparable to actually using the device. You do not have to answer the notification for it to cost you; you only have to hear it.
Two honest counterweights belong here. Studies of self-interruption find that much of the switching is initiated by the person, from habit rather than from an alert, so not all of this is done to us. And a daily survey of 174 employees across 811 days found that perceived interruptions predicted negative affect through time pressure while being responsive to those same messages predicted positive affect and a sense of having accomplished something. The inbox gives as well as takes.
Which brings us to the specific thing your profession does. You are doing all of this twice, in two mailboxes, and in one of them every decision is made on somebody else's behalf. The literature has no word for it. I would call it carrying two attentions.
Try this. For one week, give every switch a two-second bookmark. Before you leave your own work for theirs, write one line: where you were, and the first thing you will do when you come back. That is the ready-to-resume finding, and it is the cheapest intervention in this post.
The research behind this section
Mark, Gonzalez and Harris (2005) observed twenty-four information workers in detail and found work to be highly fragmented, with people averaging little time in a working sphere before switching and 57 per cent of working spheres interrupted; most interrupted work was resumed the same day, but more than two intervening activities occurred first. Mark, Gudith and Klocke (2008) found that the context of an interruption made no difference, that people completed interrupted tasks in less time with no difference in quality, and that they appeared to compensate by working faster at the price of more stress, higher frustration, time pressure and effort. Leroy (2009) introduced attention residue, showing that switching between tasks leaves part of attention on the previous task and reduces performance on the next one. Leroy and Glomb (2018) found across four studies that anticipating time pressure on the resumption of an interrupted task made it harder to switch attention and produced attention residue and lower performance on the interrupting task, and that a brief ready-to-resume intervention, in which people reflect on and plan their return, reduced the residue. Monsell (2003) reviewed task switching and reported that responses are substantially slower and usually more error-prone immediately after a switch, and that this switch cost is reduced but not eliminated by an opportunity to prepare. Jackson, Dawson and Wilson (2003) monitored employees at a UK company and found that most had their email software checking for incoming messages every five minutes and responded to the arrival of a message within six seconds, with a recovery time that existed but was shorter than published recovery times for a telephone interrupt. Marulanda-Carter and Jackson (2012) observed seven employees at a large UK car rental company and found that a typical task took one third longer when it was interrupted by email. Renaud, Ramsay and Hair (2006) compared electronic monitoring data with an online survey of email users and found that users drastically underestimate the disruptive effects of email. Stothart, Mitchum and Yehnert (2015) found that mobile phone notifications alone significantly disrupted performance on an attention-demanding task even when participants did not interact with the device, and that the magnitude was comparable to actively using the phone for calls or texts. Dabbish, Mark and González (2011) examined why people interrupt themselves, and the role of environment and habit. Wajcman and Rose (2011) argued from fine-grained data on knowledge workers that frequent, short mediated interactions should not be understood as discrete interruptions, because communication technologies are a constant and constitutive part of the work itself. Sonnentag, Reinecke, Mata and Vorderer (2018) surveyed 174 employees across 811 day-level observations and found that perceived interruptions by online messages predicted negative affect through time pressure, and that responsiveness to those messages predicted positive affect and perceived task accomplishment.
Batching, and why the advice keeps contradicting itself
Every article about email tells you to check it at set times. The research on that advice is an interesting mess, and knowing the shape of the mess is more useful than picking a side. The most quoted study is from 2015. Over two weeks, 124 adults were asked to limit email checking to three times a day for one week and to check as much as they liked in the other. In the limited week they checked about five times a day rather than about twelve, and they reported significantly lower daily stress, which in turn predicted better outcomes across a range of measures. Each person acted as their own comparison, which makes it stronger than most of what gets written about email.
A 2019 field experiment with 237 people tested three conditions: notifications as usual, batched three times a day, and switched off entirely. The batched group felt more attentive, in a better mood and more in control of their phones, and reported lower stress and fewer interruptions. The group with notifications off got very few of those benefits and reported higher anxiety and more fear of missing out. That deserves its own line in your head, because switching everything off is what people reach for when batching sounds too complicated, and it was the worst of the three.
Now the disagreement. A 2016 study tracked forty information workers for twelve workdays using computer logging, biosensors and daily surveys, so stress was measured on the body rather than reported on a form. The longer the daily time on email, the lower the perceived productivity and the higher the measured stress, and batching was associated with higher rated productivity when email duration was long. On the central claim the authors are blunt: despite widespread claims, they found no evidence that batching email leads to lower stress.
A 2022 quasi-experiment inside a Dutch financial services organisation helps reconcile the two. The batching group did have fewer email interruptions, and emotional exhaustion fell immediately after the intervention, most of all for people with high volumes and people who believed instant responses were not expected. The effects had worn off two weeks later, and the authors conclude that batching should be encouraged only where it fits the person's tasks and their organisation's expectations about response times.
One counterweight speaks directly to this job. An experience-sampling study followed fifty-two email users through 376 real emails and found that resisting an email interruption depleted wellbeing without improving task goal achievement. The self-control you spend not looking is a real cost, and in that study it bought nothing.
My reading of all four together is this. Batching helps when the volume is high and when nobody is standing behind you expecting an answer in six seconds. It does very little when the expectation of speed is the real load, because then the checking is not what costs you; the anticipating is. That is exactly the situation of somebody running an executive's inbox, which is why the next two sections matter more than this one.
Try this. Do not batch the whole inbox; batch a lane. Pick one category that reliably does not need you within the hour, such as suppliers, newsletters or anything sent to a distribution list, and give it two fixed windows a day, while the few senders who really do need a fast answer keep arriving as they always did. You get most of the benefit and none of the anxiety the notifications-off group reported.
The research behind this section
Kushlev and Dunn (2015) used a within-subjects design over two weeks in which 124 adults were randomly assigned to limit email checking to three times a day during one week and to check without limit during the other, and found significantly lower daily stress during the limited week, with lower stress in turn predicting higher wellbeing across a range of outcomes; participants checked on average 4.70 times a day when limited and 12.54 times when unlimited. Fitz, Kushlev, Jagannathan, Lewis, Paliwal and Ariely (2019) ran a randomised field experiment with 237 participants and found that those whose notifications were batched three times a day felt more attentive, productive, in a better mood and more in control of their phones, and reported lower stress and fewer interruptions, while those who received no notifications at all reaped few of those benefits and experienced higher anxiety and fear of missing out. Mark, Iqbal, Czerwinski and Johns (2016) tracked forty information workers for twelve workdays with computer logging, biosensors and daily surveys, and found that longer daily time on email was associated with lower perceived productivity and higher measured stress, that batching was associated with higher rated productivity when email duration was long, and that despite widespread claims there was no evidence that batching email leads to lower stress. Wijngaards, Pronk and Burger (2022) used survey data from a top-down batching intervention in a Dutch financial services organisation and found fewer email interruptions in the intervention group and reduced emotional exhaustion immediately afterwards, especially for workers with high email volumes and those who believed instantaneous response was not expected, with effects wearing off after two weeks and no significant effect on work engagement. Russell, Woods and Banks (2017) ran an experience-sampling field study with fifty-two email users across 376 emails and found that deploying conscientiousness to resist an email interruption depleted affective wellbeing without a positive effect on task goal achievement.
The five days without email, and what they proved
There is one study people always mention and usually get slightly wrong, so it is worth telling properly. In 2012 researchers cut off email entirely for five working days for thirteen information workers inside one organisation, measuring what happened with computer logs, ethnographic observation and wearable heart rate monitors. Without email, people multitasked less and held a longer focus, shifting between windows less often and spending longer in each one, and stress measured through heart rate variability was lower. The interviews matched: people said they could concentrate.
It is a small study and its authors do not pretend otherwise. What it establishes is worth having anyway: the effect of an inbox on attention is measurable on the body rather than only in what people report on a questionnaire.
Set beside it a larger piece of work that complicates the picture beautifully. A study combining survey and interview data found that the more time people spent handling email the more overloaded they felt, and also that the more email they actually processed, the greater their sense of being able to cope. Its central argument is the one I would put on a poster: email had become a cultural symbol of overload, and by serving as that symbol it distracted people from noticing the other sources of overload in their working lives. The inbox is a very convenient thing to blame.
A person-environment fit study of 118 workers sharpens it further: both too much email and too little, measured against what the individual wanted, were appraised as stressful. There is no correct volume, only a gap between what you have and what suits you, and the gap is the stressor in either direction.
Then the effect on the person whose inbox it is, which is the half that gets forgotten. An experience-sampling study of forty-eight managers across ten consecutive workdays found that on days with heavier email demands they perceived less progress towards their goals, and in response did less of the ordinary work of leading: less setting of structure and direction, and less of the inspiring, developing sort of behaviour. When your executive spends the morning in their inbox, the organisation loses some leadership that day. That is the business case for what you do, and it beats anything in a job description.
Try this. Ask your executive one question and write down the answer: on a day when you are in your inbox a lot, what does not get done? That answer is the thing your triage protects, and it beats any list of email rules the two of you could write together.
The research behind this section
Mark, Voida and Cardello (2012) cut off email for five workdays for thirteen information workers and found, using computer logs, ethnographic methods and wearable heart rate monitors, that without email people multitasked less, held a longer task focus and had lower stress as measured by heart rate variability. Barley, Meyerson and Grodal (2011) found that the more time people spent handling email the greater their sense of being overloaded, that the more email they processed the greater their perceived ability to cope, and that email served as a cultural symbol of overload which distracted people from recognising other sources of overload in their work. Stich, Tarafdar, Stacey and Cooper (2019) surveyed 118 working individuals and found that both more email and less email than the individual desired were appraised as stress-creating, framing the problem as email misfit rather than email volume. Rosen, Simon, Gajendran, Johnson, Lee and Lin (2019) ran an experience-sampling study of forty-eight managers across ten consecutive workdays and found that daily email demands were associated with a perceived lack of goal progress, to which leaders responded by reducing both initiating structure and transformational behaviours, with the effect on goal progress strongest when email was perceived as less central to the job.
What people actually mean by urgent
Here is the sentence I would like every person who has ever run an inbox to have on a card: the word urgent in a subject line is information about the sender, not about the work.
The research on urgency is unusually clean. Studies published in 2018 found what the authors called the mere urgency effect: people chose tasks carrying a sense of expiry over tasks with objectively better payoffs, and the effect was stronger among people who saw themselves as busy. A companion paper found something similar about deadlines, where a more generous deadline could make a goal less likely to be pursued. Urgency hijacks selection, it does not track importance, and the people most vulnerable to it are the ones with the most on.
The scarcity research adds the deeper version. When a resource is scarce, attention is captured by the pressing problem and neglects what lies outside it, and anyone working through a Monday inbox is operating under scarcity of time, so the tunnel is forming before the first message is opened.
So what does urgent actually encode? Mostly, a sender's expectation about how long they will wait, and the research on that expectation is more precise than you might think.
Chronemics is the study of time as a communication signal, and in writing the strongest signal is how long a reply takes. A study asked managers to evaluate job candidates whose email response latency varied between one day, two weeks and silence for more than a month, and latency changed judgements of credibility and attractiveness, by an amount that depended on how much the manager wanted that candidate in the first place. Silence, in that literature, is a message rather than an absence of one, and its meaning is supplied by the reader. Field studies of email rhythms make the same point more kindly: people form expectations about your response rhythm largely from your own past behaviour, and a departure from the rhythm is what triggers concern rather than any particular number of hours. The expectation is something you set rather than something imposed on you.
Which leads to the concept that names the feeling. Workplace telepressure was defined in 2015 as the preoccupation with and urge to respond quickly to work messages, and its validation study found it predicted burnout, absenteeism, poorer sleep quality and more email responding over and above job involvement and general work demands. A follow-up found the route ran through psychological detachment: telepressure made it harder to switch off, and the failure to switch off did the damage. A later paper argued that the word pressure is slightly wrong, because much of the urge is generated inside the person, which is uncomfortable and hopeful at once, since an internal habit is something you can work on.
Two further pieces belong here. Three studies found that organisational expectations to monitor work email out of hours produced anxiety that affected the health and relationship satisfaction of employees and, remarkably, of their partners. And a study of knowledge professionals using mobile email named the autonomy paradox, in which flexibility escalated collective expectations of availability until the choice to work anywhere became the practice of working everywhere.
Make it specific to this job. When you run somebody else's inbox you carry their telepressure as well as your own, and theirs arrives with the added weight that the consequences of getting it wrong land on somebody else. That is a structural feature of the role rather than a character flaw, and the fix for it is an agreement rather than more willpower.
Try this. Replace the word urgent with a time, every time. Decide what it means in hours before anything else: within the hour, before end of day, before Thursday, or no fixed time at all. Write the answer at the top of the message. Four categories, one decision, and the word loses most of its power over your morning.
The research behind this section
Zhu, Yang and Hsee (2018) found across a set of studies a mere urgency effect, in which people were more likely to perform tasks carrying an illusion of expiration over tasks with objectively better payoffs, and that the effect was more pronounced among individuals who perceived themselves as busy. Zhu, Bagchi and Hock (2019) found a mere deadline effect, in which a longer deadline could increase the perceived difficulty of a task and reduce goal pursuit. Shah, Mullainathan and Shafir (2012) showed in laboratory experiments that scarcity itself biases cognitive processes, capturing attention with the pressing problem and producing neglect elsewhere. Kalman and Rafaeli (2011) had managers evaluate job candidates whose email response latency varied between one day, two weeks and silence for more than a month, and found that candidate reward valence moderated the effect of latency on evaluation, credibility and attractiveness, and set out a norms-based definition of online silence as a chronemic cue. Tyler and Tang (2003) studied rhythms in email usage and how people form and interpret expectations about response times. Dabbish, Kraut, Fussell and Kiesler (2005) examined what predicts whether a recipient acts on a message. Dabbish and Kraut (2006) analysed the factors associated with email strain at work. Barber and Santuzzi (2015) defined and validated workplace telepressure as the preoccupation with and urge for responding quickly to work messages, and found it predicted physical and cognitive burnout, absenteeism, sleep quality and email responding beyond job involvement, affective commitment and general work demands. Santuzzi and Barber (2018) found that psychological detachment intervened between workplace telepressure and worker wellbeing. Grawitch, Werth, Palmer, Erb and Lavigne (2018) examined telepressure as a unique contributor to workplace wellbeing beyond perceived demands and individual differences, and questioned the word pressure on the grounds that the urge may be internally generated. Barber, Conlin and Santuzzi (2019) found across two studies with 254 and 409 employees that telepressure was negatively related to satisfaction with work-life balance, explained by psychological detachment and control over leisure time. Cambier and Vlerick (2020) examined how workplace telepressure relates to email communication behaviour. Becker, Belkin, Conroy and Tuskey (2021) ran three studies, including experience sampling with 108 working adults and a dyadic study of 138 employees and their significant others, and found that organisational expectations to monitor email during non-work hours elicited anxiety which negatively affected the health and relationship satisfaction of employees and crossed over to their partners. Mazmanian, Orlikowski and Yates (2013) found that knowledge professionals using mobile email devices gained flexibility and control in the short term while intensifying collective expectations of their availability, an autonomy paradox in which working anywhere became working everywhere.
A triage that holds up on a Monday
The founding study is from 1996 and it named the problem. Email had stopped being a communication medium and become a task manager, a filing cabinet and a contact book at once, and the inbox was doing all three badly. A 2014 replication separated the two meanings of overload that people muddle: overload of volume, too many messages, and overload of status, too many kinds of pending thing in one list. Work email, they found, is particularly overloaded by status, which is why sorting by what a message is for beats sorting by who sent it.
Then the finding that will save you time this week. A study of email refinding compared people who filed diligently into folders with people who left everything in one place and searched. Filing took real, measurable effort and did not make refinding faster or more successful; scanning and searching did about as well. There are good reasons to keep a few folders, and preparing for an audit is one of them. Building a beautiful taxonomy so that you can find things later is not one the evidence supports.
What does a triage decision actually turn on? A study of what predicts whether somebody acts on a message found the answer sits in three features: who it is from, whether it explicitly requests an action, and how important the recipient judges it to be. Those are the things worth reading for on the first pass.
There is one more idea, from 1986, and it is the best thing in this section. Organisations process information for two different reasons. Sometimes there is uncertainty, where somebody needs more data to answer a question, and that is solved by sending the number. Sometimes there is equivocality, where the data exists and people read it differently, and that is solved only by a conversation in which people build a shared reading. A good deal of what sits unanswered in an executive's inbox is equivocal, and every reply you draft for it will fail, because what is needed is twenty minutes and two humans. Telling the two apart is a superpower in this job.
So here is the framework. Four lanes, decided in one pass, in this order.
Mine. I can answer this in their voice without asking. This lane is the point of the arrangement and should be the biggest one; if it is small, the agreement in the last section is the thing to fix.
Theirs. Only they can decide, and the decision has a real deadline. This lane gets a time attached and a one-line summary from you, and because the research says the sender, the ask and the importance are what get read, give all three in that line.
Somebody else's. This belongs to a colleague and you are going to route it. Say so to the sender, by name, so the thread does not come back.
Neither. It looks like a request and it is not one: information, courtesy copies, things already handled. Filed unread past the first line, and if a category keeps landing here, that is a filter waiting to be written.
Then the pile that is not a lane, because it lives on a different page: Waiting. Anything you have sent on and expect a reply to. Give it a date and a home outside the inbox.
And one rule that makes the lanes work: when something is equivocal, do not draft a reply at all. Put twenty minutes in the diary and say why.
Try this. Run the four lanes once, with a timer, on the next fifty messages. No replies, no drafting, no reading past the first screen. Count how many landed in Mine. That number is the best measure of how well the delegation is working, and it is the one to take into the conversation in the last section.
The research behind this section
Whittaker and Sidner (1996) studied twenty office workers and named email overload, showing that the inbox had become a task manager, a personal archive and a contact manager at the same time, and that folder strategies imposed substantial costs and often failed. Whittaker, Bellotti and Gwizdka (2006) reviewed email in personal information management and described how the inbox, folders, search and sort support task management, personal archiving and contact management. Grevet, Choi, Kumar and Gilbert (2014) replicated and extended the 1996 work in the context of Gmail and found that overload persists in two distinct forms, volume overload and status overload, with work email particularly status-overloaded and personal email also type-overloaded. Whittaker, Matthews, Cerruti, Badenes and Tang (2011) studied email refinding and found that although filing was effortful, it did not improve the speed or success of refinding compared with scanning and searching. Dabbish, Kraut, Fussell and Kiesler (2005) found that whether a recipient acted on a message was predicted by the sender, by whether the message requested an action, and by the recipient's judgement of its importance. Dabbish and Kraut (2006) analysed factors associated with email strain. Daft and Lengel (1986) distinguished uncertainty, the absence of information, from equivocality, the existence of multiple conflicting interpretations, and argued that rich media and direct interaction are needed to resolve equivocality while more data resolves uncertainty.
The keys to someone else's house
Delegated access to a mailbox is one of the largest grants of trust in ordinary working life, and it is usually made by clicking a permissions box on a Tuesday. The research on delegation says the grant matters. Studies of managers delegating to their people find it associated with better outcomes through the sense of being trusted, and a meta-analysis of psychological empowerment found it related to job satisfaction, commitment, task and contextual performance, and lower strain and turnover intention. The model of trust most of this work rests on has three parts: ability, benevolence and integrity. An executive handing over their inbox is judging all three at once, and the last two are doing most of the work. They are not mainly betting that you can write. They are betting that you will not use what you see.
Which is where the privacy research earns its place. Communication privacy management theory treats private information as something people own and then co-own once it is shared, and says trouble arises when the rules for the co-owned information were never agreed. The word it uses for the mess that follows is turbulence, which is a precise description of what happens when an assistant reads something unexpected and nobody has ever discussed what comes next. The wider privacy literature makes the same point in another register: privacy is contextual, and what feels appropriate depends on the setting and its expectations rather than on the content itself.
The research on secrets adds a kind note. Studies of more than 800 people holding over 10,000 real secrets found the burden sits less in concealment and more in how often the mind returns to the thing, and that confiding predicted higher wellbeing through a greater sense of being able to cope. If you hold things for your executive, and you do, the answer is one agreed person with whom the fact of holding can be discussed in general terms.
Then the part nobody says out loud, which is that the access runs in the other direction too. Everybody writing to your executive knows, or should know, that you will read it, and the research on electronic monitoring is worth borrowing here. A major review concluded that the effects of monitoring are contradictory and depend on its characteristics: what is monitored, how openly, and for what purpose. A study across two call centres found more frequent supervisory monitoring associated with better task performance and more citizenship behaviour rather than less. And a study of 582 managers and professionals found that controls can raise trust in the organisation when they are seen as procedurally fair, while controls that are inconsistent or overly rigid undermine it. The lesson transfers exactly: access that is openly declared and consistently used builds trust, and access that people discover by accident does the opposite.
Delegated access is less a permission setting than an agreement about what you will do with what you see.
So: four rules, agreed out loud, and certainly before the first awkward message.
Who else knows you read it. Whether their reports, the board and their family know that mail sent to them is read by you. Theirs to decide, yours to ask.
What you never open. Anything marked personal, anything from named individuals, anything about your own pay or performance. Write the list.
What happens when something lands that you should not have seen. The rule people usually want is the simplest: you say so, once, in one line, and it is never mentioned again. Agreeing it in advance turns a horrible moment into a procedure.
How a confidence gets flagged. A word your executive can put on a thread that means do not action, do not summarise, do not mention. Ten seconds to agree, and it prevents the thing everyone dreads.
Try this. Put those four rules in one email, send it to your executive and ask them to reply with the word yes. Five minutes of their time, and it converts an assumption into an agreement.
The research behind this section
Leana (1986) examined the predictors and consequences of delegation by managers. Chen and Aryee (2007) studied delegation and employee work outcomes and the processes that mediate them. Seibert, Wang and Courtright (2011) meta-analysed psychological and team empowerment and found empowerment positively associated with job satisfaction, organisational commitment, task and contextual performance, and negatively associated with strain and turnover intentions, with team empowerment positively related to team performance. Mayer, Davis and Schoorman (1995) set out the integrative model of organisational trust in which trustworthiness is composed of ability, benevolence and integrity. Zhao, Wayne, Glibkowski and Bravo (2007) meta-analysed psychological contract breach and found it related to every work-related outcome examined except actual turnover, with affect mediating the effect on attitudes and effectiveness. Petronio (2010) set out communication privacy management theory, in which private information is co-owned once disclosed and requires negotiated rules, with boundary turbulence following when those rules are unclear or violated. Smith, Dinev and Xu (2011) reviewed information privacy research across disciplines and emphasised the contextual nature of privacy. Slepian, Chun and Mason (2017) examined the experience of secrecy across a large set of real secrets and found that mind-wandering to a secret, rather than active concealment during conversation, accounted for much of its harm. Slepian and Moulton-Tetlock (2019) examined more than 800 participants and over 10,000 secrets and found that confiding a secret predicted higher wellbeing through perceived coping efficacy and less frequent mind-wandering to the secret, without reducing instances of concealment. Ravid, Tomczak, White and Behrend (2020) reviewed electronic performance monitoring and concluded that contradictory findings in the literature reflect the multidimensional nature of monitoring, with contextual and psychological variables determining its effects. Bhave (2014) found across two field studies that more frequent supervisory use of electronic performance monitoring was associated with better task performance and more organisational citizenship behaviour, and was not significantly related to counterproductive work behaviour. Weibel, Den Hartog, Gillespie, Searle, Six and Skinner (2016) surveyed 582 European managers and professionals and found that organisational control related positively to trust in the organisation through procedural fairness and organisational prestige, while interviews showed that inconsistent, overly rigid or badly incentivised controls undermined trust. Truss, Alfes, Shantz and Rosewarne (2013) surveyed 1,011 secretaries in the UK and found that although secretaries were better qualified and generally well paid, with a minority undertaking complex managerial tasks, most continued to perform traditional tasks and career prospects remained poor.
Writing in their voice, and when to say it is you
Now the craft, which is where this job stops being administration and becomes something closer to translation. Begin with the bad news, because everything else follows from it. Five experiments published in 2005 asked people to communicate sarcastic and non-sarcastic statements by email and to predict how well they would be understood. In the study that compared media directly, listeners decoded sarcasm about three-quarters of the time and email readers performed at a level indistinguishable from chance, and senders did not anticipate the difference. In a later study in the same paper, people expected about seventy-two per cent accuracy and achieved about sixty-three, and when they were made to experience a statement the way they had intended it, their expectation rose to eighty-eight per cent while their actual accuracy did not move at all. The researchers attribute it to egocentrism: you hear your own sentence in the tone you meant, and it is very hard to hear it any other way.
A review of emotion in email gives the error its shape. With no gesture, emphasis or intonation to work with, a positive message tends to be read as more neutral than intended and a mildly negative one as more negative. A pleasant note reads as flat; a slightly terse one reads as annoyed. And the cost is documented: studies of email incivility find rude messages damage mood, energy, engagement and performance well beyond the moment. When you write in somebody else's name, a curt line does not cost you your reputation. It costs them theirs, and they will not know why.
There is a surprising piece of good news to set against that. A laboratory experiment with 117 participants had people deliver consequential positive or negative information by computer, telephone or face-to-face, and they distorted the negative information least over the computer, being more accurate and reporting more comfort. The medium that flattens your tone also makes it easier to say the true thing.
So how do you write as somebody else? The language research is encouraging, because voice turns out to be made of small, countable things.
Authorship attribution studies show that people can be identified from their writing largely through function words, the small grammatical words nobody notices using. Text analysis research makes the same point from the other side: the informative part of language sits in pronouns, articles and prepositions rather than in the vocabulary people think of as style. Five studies of pronouns and rank found a consistent pattern, including in a naturalistic study of real incoming and outgoing email: people of higher status used fewer first-person singular pronouns and more first-person plural and second-person ones. Your executive almost certainly writes fewer I's than you do, and more we's and you's. That is less a personality trait than a position in a room, rendered in grammar.
A humbling finding comes attached. Language style matching, the tendency of two people's function-word use to converge, happens automatically and predicts how cohesive a group is, and when researchers asked participants to increase their matching deliberately, they could not. You cannot will your way into somebody's voice by trying harder. You get there by reading enough of their writing for the pattern to become ambient, and by writing down the parts that are conscious. A separate set of studies found matching unrelated to how good participants and judges thought the interaction was, so treat it as a marker of engagement rather than a guarantee of quality.
The conventions around greetings and sign-offs are more arbitrary than anyone admits. A study of workplace email found that greetings and closings varied enormously between organisations and did not map neatly onto how close two people were, and an earlier study found people signed differently depending on whether they were writing upwards, downwards or sideways in the hierarchy. There is no correct sign-off. There is only your executive's sign-off, which is why it belongs on a sheet rather than in your judgement each time. Politeness is similarly learnable: email requests come out more polite than the spoken equivalent, people accommodate to the politeness level of whatever arrives, and emoticons in workplace email function largely as hedges softening a request or a criticism rather than as expressions of feeling.
So: the drafting standard, six lines, written once and kept where you can see it.
The greeting and the sign-off, verbatim. Both versions, internal and external. No thinking required, ever again.
Three sentences they would never write. Take them from their sent items. The ones that make you wince are the most useful.
The pronoun default. Whether replies go out as I or as we, and when that switches. Given the status research, this is the highest-value line on the sheet.
The hedge and the refusal. How they say no, and how they say not yet, in their exact words.
The length ceiling. Most executives write shorter than their assistants do, because brevity is a privilege of position. Agree a number of lines above which you check.
The attribution rule. When the message goes out as them, and when it goes out as you.
That last line deserves its own paragraph, because it is the one people agonise over. My rule is that the signature follows the decision rather than the keyboard. If the message carries a decision that belongs to your executive, it goes out in their name, because pretending otherwise makes the reader chase them for confirmation. If it carries a decision that belongs to you, or is logistics, scheduling, chasing or information, it goes out as you, because that is true and because it builds your standing with people who will deal with you directly one day. And when a message is theirs in substance but yours in the writing, the phrase on behalf of costs you nothing and buys the reader an accurate picture of who to reply to.
There is an honest cost to all of this, and the emotional labour research names it. Writing as somebody else, in their register, managing their relationships and absorbing the reactions, is surface acting by any reasonable definition, and meta-analytic work finds surface acting associated with emotional exhaustion and reduced wellbeing. That is an argument for the agreement in the next section, and for the lane called Mine being truly yours rather than a performance of being them.
Try this. Open their sent items, take the twelve most recent messages they wrote themselves and fill in the six lines above. Twenty minutes. Then send it to them with one question: have I got you right? Almost every executive enjoys that conversation, and almost none has ever been asked.
The research behind this section
Kruger, Epley, Parker and Ng (2005) ran five experiments and found that people believe they can communicate emotion and tone over email more effectively than they actually can; in the study comparing media, participants who heard sarcastic statements decoded nearly three-quarters of them while those who read them on email performed at a level indistinguishable from chance, and senders failed to anticipate the difference; in a later study anticipated accuracy was 72.3 per cent against actual accuracy of 62.8 per cent, and participants led to experience a statement as they had intended it believed they had been understood 88.2 per cent of the time while actual accuracy was unchanged. The authors attribute the overconfidence to egocentrism. Byron (2008) reviewed the communication and miscommunication of emotion by email and argued that the absence of non-verbal cues produces a neutrality effect, in which positive emotion is read as more neutral than intended, and a negativity effect, in which slightly negative emotion is read as more negative than intended. Cheshin, Rafaeli and Bos (2011) found that emotion conveyed through text and behaviour influenced others' affect in virtual teams in the absence of non-verbal cues. Lim and Teo (2009) found that cyber incivility in email affected employees' work attitudes and behaviour. Giumetti, Hatfield, Scisco, Schroeder, Muth and Kowalski (2013) compared incivility with support in an online context and found differential effects on mood, energy, engagement and performance. Sussman and Sproull (1999) had 117 participants deliver positive or negative personally consequential information by computer-mediated communication, telephone or face-to-face, and found that participants distorted negative information less, and were more accurate and honest, over computer-mediated communication, while reporting higher satisfaction and comfort. Koppel, Schler and Argamon (2009) reviewed computational authorship attribution, including verification and profiling problems, in which authors are identified largely from function-word usage. Tausczik and Pennebaker (2010) described text analysis methods and the psychological meaning of word categories including pronouns and function words. Kacewicz, Pennebaker, Davis, Jeon and Graesser (2014) found across five studies, including a naturalistic study of the incoming and outgoing email of nine participants, that people with higher status consistently used fewer first-person singular pronouns and more first-person plural and second-person pronouns. Gonzales, Hancock and Pennebaker (2010) applied a language style matching algorithm to seventy groups totalling 324 individuals and found that matching predicted group cohesiveness in both face-to-face and computer-mediated settings. Ireland and Pennebaker (2010) found across samples totalling 1,744 students that people automatically matched the language style of the texts they responded to, that the measure was internally consistent and reliable, and that participants were not able to increase style matching intentionally. Niederhoffer and Pennebaker (2002) found across three studies that linguistic style matching occurred at both conversation and turn level but was unrelated to participants' and judges' ratings of the quality of the interaction. Waldvogel (2007) studied greetings and closings in workplace email across two organisations and found substantial variation between workplaces that was not explained by relationship distance alone. Sherblom (1988) found that direction within the organisational hierarchy influenced how people signed their electronic mail. Duthler (2006) compared the politeness of requests made by email and by voicemail and found email requests more polite. Bunz and Campbell (2004) found that people accommodated to the politeness markers in the email they received. Skovholt, Grønning and Kankaanranta (2014) analysed emoticons in workplace email and found they functioned largely as hedges and politeness markers attached to requests and criticism rather than as expressions of emotion. Hülsheger and Schewe (2011) meta-analysed three decades of emotional labour research and found surface acting associated with impaired wellbeing and emotional exhaustion.
The rituals that stop it becoming a second job
An inbox is the easiest thing in the world to be handed and the hardest to hand back, because nobody ever announces that it has become yours. The research on how that happens is sharper than most people expect. Illegitimate tasks are those that violate norms about what a person can reasonably be expected to do, because they seem unnecessary or unreasonable for somebody in that role. Across several studies they predicted low self-esteem, resentment and burnout over and above role conflict and perceived injustice, and across two months they predicted strain rather than being predicted by it. What matters for an inbox is that the same task can be legitimate or illegitimate depending only on whether it was agreed. Drafting replies for your executive is skilled professional work when it is part of the arrangement, and an imposition when it arrived by accident.
That distinction sits in the definition rather than in the task, and definitions differ. A classic study found that employees vary in whether they consider a behaviour part of their job or a discretionary extra, and that the broader somebody's definition of their role, the more of those behaviours they perform. The costs of that generosity are documented: individual initiative relates to role overload, job stress and work-family conflict, and a well-known analysis describes the paradox in which citizenship behaviour helps the organisation and harms the career of the person performing it. A study of tasks everybody would rather somebody else did found that women volunteered more, were asked more often and accepted more often.
Invisible is the right word here, and it has a literature of its own. Work becomes invisible through cultural, legal and spatial mechanisms, and once invisible it gets economically devalued. Running an inbox is almost perfectly designed to disappear: it happens inside somebody else's account, it produces messages signed by somebody else, and its main output is the absence of problems. Nobody has ever written a performance review line that says nothing went wrong in April because of her.
So the answer is ritual, and the most useful research comes from somewhere else entirely. Medicine has spent twenty years studying handover, because in hospitals the consequences of a bad one are measured in harm. Reviews found handovers highly variable and frequently implicated in communication failures, and a critical incident study of sign-outs found those failures led to uncertainty about care decisions, inefficiency and delay. A multi-site study of a structured handover programme, with a standard format, a mnemonic and training, reported a significant reduction in medical errors and preventable adverse events without lengthening the handover itself. The same logic sits behind the surgical safety checklist, a short list read aloud before an operation that was associated with markedly fewer complications across eight hospitals.
An inbox handover is not surgery, and the principle transfers anyway: what makes a handover work is that it has a fixed shape and happens whether or not anybody feels it is needed.
An arrangement that is never reviewed stops being an arrangement and quietly becomes an accumulation.
Four rituals, none of them long.
Ten minutes a week. About the lanes rather than the messages: how many landed in Mine, how many in Theirs, and which category keeps arriving that neither of you has a rule for yet.
A fortnight of counting, twice a year. Count the messages you handle in their inbox and the hours it takes, and write the number down. The invisible work research says the number is the whole battle, because you cannot discuss what nobody can see.
The holiday handover, written once. One page with the lanes, the four access rules, the drafting standard and the people whose messages always get a reply. This is the page you hand to whoever covers, and writing it is what converts an accumulation into a defined piece of work.
The annual re-agreement. Once a year, out loud: is this still the right arrangement? Their role changes, and so does yours.
Try this. Do the fortnight of counting before anything else in this section, not to complain and not to build a case. Do it because most people who run an executive's inbox have no idea how much of their week it takes, and neither does their executive, and you cannot have a sensible conversation about an arrangement neither of you can measure.
The research behind this section
Semmer, Tschan, Meier, Facchin and Jacobshagen (2010) showed that illegitimate tasks, meaning tasks that violate norms about what can reasonably be expected of a person and are experienced as an offence to professional identity, related to counterproductive work behaviour while controlling for effort-reward imbalance, personality and organisational justice. Semmer, Jacobshagen, Meier, Elfering, Beehr, Kälin and Tschan (2015) reported three studies in which illegitimate tasks predicted low self-esteem, resentment towards the organisation and burnout while controlling for role conflict and justice perceptions, and in which illegitimate tasks predicted strain across two months rather than being predicted by it. Björk, Bejerot, Jacobshagen and Härenstam (2013) examined illegitimate tasks as a stressor in relation to organisational control and resource deficits. Morrison (1994) found that employees differ in whether they define a given behaviour as in-role or extra-role, and that broader role definitions were associated with performing more citizenship behaviours. Bolino and Turnley (2005) found that individual initiative was related to role overload, job stress and work-family conflict. Bergeron (2007) set out the paradox in which citizenship behaviour benefits the organisation while potentially harming the individual's career through the allocation of time. Babcock, Recalde, Vesterlund and Weingart (2017) examined the allocation of tasks with low promotability and found that women volunteered more, were asked to volunteer more and accepted requests more often than men, with beliefs that women are more likely to say yes driving the difference. Daniels (1987) introduced invisible work as a concept for culturally and economically devalued labour. Hatton (2017) reconstructed invisible work as an analytical concept and identified cultural, legal and spatial mechanisms through which work is made invisible and therefore devalued. Truss, Alfes, Shantz and Rosewarne (2013) surveyed 1,011 secretaries and found continuity in traditional task content and limited career prospects. Cohen and Hilligoss (2010) reviewed the published literature on hospital handoffs and identified substantial deficiencies in both practice and evidence. Arora (2005) analysed communication failures in patient sign-out through critical incidents and found failures of content and of process that led to uncertainty and inefficiency. Starmer, Spector, Srivastava, West, Rosenbluth and colleagues (2014) implemented a structured handoff programme across nine hospitals and reported a significant reduction in medical errors and preventable adverse events with no increase in handoff duration. Haynes, Weiser, Berry and colleagues (2009) found that a nineteen-item surgical safety checklist used across eight hospitals was associated with a marked reduction in complications.
The inbox check
Everything above comes down to one situation: an inbox that is not yours, a person whose name goes on the replies, and an arrangement between you that was probably never written down. So here is a tool for that. Put in whose inbox it is, how much of it comes to you, what you are allowed to answer, how fast people expect replies and what happens when something goes wrong. It will tell you where the load is coming from, what the research suggests you change first, what to check in a month, and it gives you an agreement to copy and send.
Where to start
If you take one thing from all of this, make it the agreement: what you send, what you hold and what you answer. Everything else here is technique, and technique applied to an undefined arrangement produces a very efficient version of a problem. The research on illegitimate tasks says the same work can be a skilled contribution or a quiet insult depending only on whether it was agreed, and the research on privacy says the same access can be trust or turbulence depending only on whether the rules were named. The agreement is the hinge.
After that, there are three moves worth making.
Replace the word urgent with a time, every time, before you decide anything else about a message. Sort by what a message is for rather than who sent it, and stop filing for retrieval, because the evidence says search does the job. And write the drafting standard from their sent items rather than from your instincts, because voice is made of small countable things and your instincts are working from your own ear.
I have put the three pages I would use into a download: a triage page with the four lanes and the urgency test, a drafting standard you fill in from their sent items, and the inbox agreement itself, laid out to be filled in during one conversation and signed off in one reply.
What I hope comes across is how much skill the job contains. Deciding, two hundred times before lunch, what deserves another person's attention is a judgement task of a high order, performed at speed, on incomplete information, with somebody else's reputation attached to the outcome. The research is finally catching up with that, and reading it made me prouder of the work rather than more tired by it. I hope it does the same for you.
The part of this that shows up in how you are seen at work sits inside what the Connected Leader Assessment calls Judgement and Decision-Making: how you decide what matters, how you decide it quickly, and how well those decisions hold up when somebody looks at them afterwards. An inbox is that capability performed several hundred times a day, in public, on behalf of somebody else.
The assessment is free, takes about twelve minutes and gives you a read across all five dimensions. If you want to go further on this one, the Skill Sprints are short, focused programmes on a single dimension, and the Irreplaceable Audit is the place to look at how much of what you do in that inbox is judgement nobody could hand to a filter, which is the part most worth making visible.
Stay savvy.
Meg ✌️
Free tools
The inbox check, the interactive tool in this post, free
The Inbox Pages, the download in this post, free
The Connected Leader Assessment, twelve minutes, free, with a read on Judgement and Decision-Making, the dimension this post sits in
The Irreplaceable Audit, forty-eight activities across four kinds of work, free, for separating the judgement in your inbox work from the parts a rule could do
P.S. Want to Go Deeper on Running an Inbox?
If this resonated and you want to go further, here is the evidence the piece is built on. These are the papers that shaped my thinking on what interruptions cost, what batching does and does not do, how people read urgency and silence, how tone survives the journey into writing, and what turns a delegated task into an accumulated one, and they are well worth your time.
Every DOI below has been checked. Where I could not confirm a figure from the paper itself, I have left the figure out rather than guess.
📃 Mark, Gonzalez and Harris (2005) Twenty-four information workers observed at their desks. Work was highly fragmented and 57% of working spheres were interrupted. https://doi.org/10.1145/1054972.1055017
📃 Mark, Gudith and Klocke (2008) Interrupted tasks were finished faster with no loss of quality, at the price of more stress, frustration, time pressure and effort. https://doi.org/10.1145/1357054.1357072
📃 Leroy (2009) Attention residue: part of your mind stays with the task you left, and performance on the next one suffers. https://doi.org/10.1016/j.obhdp.2009.04.002
📃 Leroy and Glomb (2018) Four studies. Expecting to resume under time pressure made residue worse; a brief ready-to-resume plan reduced it. https://doi.org/10.1287/orsc.2017.1184
📃 Monsell (2003) Review of task switching. Switch costs shrink with preparation but do not disappear. https://doi.org/10.1016/s1364-6613(03)00028-7
📃 Jackson, Dawson and Wilson (2003) Employees at a UK company checked mail every five minutes and answered an arriving message within six seconds. https://doi.org/10.1016/s0268-4012(02)00068-3
📃 Mark and colleagues (2016) Forty workers, twelve days, biosensors. Longer email time meant more measured stress, and batching showed no stress benefit. https://doi.org/10.1145/2858036.2858262
📃 Dabbish, Mark and González (2011) Much of the switching in a working day is started by the person, from habit and environment rather than an alert. https://doi.org/10.1145/1978942.1979405
📃 Stothart, Mitchum and Yehnert (2015) A phone notification on its own disrupted an attention task about as much as using the phone. https://doi.org/10.1037/xhp0000100
📃 Wajcman and Rose (2011) A counterweight: for knowledge workers the short mediated exchanges are the work, not breaks from it. https://doi.org/10.1177/0170840611410829
📃 Kushlev and Dunn (2015) 124 adults over two weeks. Limiting email to three times a day significantly lowered daily stress. https://doi.org/10.1016/j.chb.2014.11.005
📃 Fitz and colleagues (2019) Field experiment, 237 people. Batching three times a day helped; switching notifications off entirely raised anxiety. https://doi.org/10.1016/j.chb.2019.07.016
📃 Wijngaards, Pronk and Burger (2022) A batching intervention in a financial services firm. Exhaustion fell for some workers, and the effect had gone in two weeks. https://doi.org/10.1016/j.invent.2022.100494
📃 Sonnentag and colleagues (2018) 174 employees, 811 days. Interruptions predicted negative affect through time pressure; being responsive predicted positive affect. https://doi.org/10.1002/job.2239
📃 Russell, Woods and Banks (2017) 52 email users, 376 emails. Resisting an email interruption depleted wellbeing without helping the task. https://doi.org/10.1111/joop.12177
📃 Mark, Voida and Cardello (2012) Five days with no email for thirteen workers. Less multitasking, longer focus, lower stress on heart rate variability. https://doi.org/10.1145/2207676.2207754
📃 Barley, Meyerson and Grodal (2011) More time on email meant more overload, more email processed meant more coping, and email became the symbol people blamed. https://doi.org/10.1287/orsc.1100.0573
📃 Becker and colleagues (2021) Expectations to monitor email out of hours produced anxiety that reached employees' health and their partners. https://doi.org/10.1177/0149206319890655
📃 Barber and Santuzzi (2015) Workplace telepressure defined and validated. It predicted burnout, absenteeism, poorer sleep and more email responding. https://doi.org/10.1037/a0038278
📃 Santuzzi and Barber (2018) The route from telepressure to poorer wellbeing runs through failing to detach from work. https://doi.org/10.1007/s41542-018-0022-8
📃 Grawitch and colleagues (2018) Telepressure explained wellbeing beyond job demands, and much of the urge is generated inside the person. https://doi.org/10.1002/smi.2792
📃 Cambier and Vlerick (2020) How workplace telepressure shows up in actual email behaviour. https://doi.org/10.1007/s10111-019-00592-1
📃 Barber, Conlin and Santuzzi (2019) Two studies, 254 and 409 employees. Telepressure went with lower satisfaction with work-life balance. https://doi.org/10.1002/smi.2864
📃 Stich and colleagues (2019) 118 workers. Both more email and less email than a person wanted were appraised as stressful. https://doi.org/10.17705/1jais.00531
📃 Rosen and colleagues (2019) 48 managers, 10 days. Heavy email days meant less perceived goal progress and less leading. https://doi.org/10.1037/apl0000343
📃 Mazmanian, Orlikowski and Yates (2013) The autonomy paradox. Flexibility raised collective expectations until working anywhere became working everywhere. https://doi.org/10.1287/orsc.1120.0806
📃 Marulanda-Carter and Jackson (2012) Observation in a car rental firm. A typical task took about a third longer when email interrupted it. https://doi.org/10.1108/13287261211221146
📃 Kalman and Rafaeli (2011) Response latency as an expectancy violation. One day, two weeks and a month of silence changed how a person was judged. https://doi.org/10.1177/0093650210378229
📃 Tyler and Tang (2003) Rhythms in email use, and how people form expectations about when a reply is due. https://doi.org/10.1007/978-94-010-0068-0_13
📃 Dabbish and colleagues (2005) Whether somebody acts on a message is predicted by the sender, the ask, and how important they judge it. https://doi.org/10.1145/1054972.1055068
📃 Dabbish and Kraut (2006) The factors associated with email strain at work. https://doi.org/10.1145/1180875.1180941
📃 Zhu, Yang and Hsee (2018) The mere urgency effect. People chose tasks that felt like expiring over tasks that paid better. https://doi.org/10.1093/jcr/ucy008
📃 Zhu, Bagchi and Hock (2019) The mere deadline effect. A longer deadline could make a goal feel harder and less likely to be pursued. https://doi.org/10.1093/jcr/ucy030
📃 Shah, Mullainathan and Shafir (2012) Scarcity captures attention with the pressing problem and produces neglect elsewhere. https://doi.org/10.1126/science.1222426
📃 Whittaker and Sidner (1996) Email overload named. The inbox had become a task manager, an archive and a contact book at once. https://doi.org/10.1145/238386.238530
📃 Whittaker, Bellotti and Gwizdka (2006) How the inbox, folders, search and sort carry task management, archiving and contact management. https://doi.org/10.1145/1107458.1107494
📃 Whittaker and colleagues (2011) Filing email into folders was effortful and did not make refinding faster or more successful than searching. https://doi.org/10.1145/1978942.1979457
📃 Grevet and colleagues (2014) Overload comes in two forms: too many messages, and too many kinds of pending thing in one list. https://doi.org/10.1145/2556288.2557013
📃 Renaud, Ramsay and Hair (2006) Monitoring data against self-report. People drastically underestimate how disruptive their own email is. https://doi.org/10.1207/s15327590ijhc2103_3
📃 Leana (1986) What predicts delegation by managers, and what follows from it. https://doi.org/10.2307/255943
📃 Chen and Aryee (2007) Delegation and employee work outcomes, and the processes that carry the effect. https://doi.org/10.5465/amj.2007.24162389
📃 Seibert, Wang and Courtright (2011) Meta-analysis. Empowerment related to satisfaction, commitment and performance, and to less strain and turnover intention. https://doi.org/10.1037/a0022676
📃 Mayer, Davis and Schoorman (1995) The model of trust: ability, benevolence and integrity. https://doi.org/10.5465/amr.1995.9508080335
📃 Zhao and colleagues (2007) Meta-analysis. Psychological contract breach affected every outcome examined except actual turnover. https://doi.org/10.1111/j.1744-6570.2007.00087.x
📃 Ravid and colleagues (2020) Review of electronic monitoring. Its effects depend on what is watched, how openly, and why. https://doi.org/10.1177/0149206319869435
📃 Bhave (2014) Two field studies. More frequent supervisory monitoring went with better performance and more citizenship behaviour. https://doi.org/10.1111/peps.12046
📃 Weibel and colleagues (2016) 582 managers and professionals. Controls raised trust when seen as fair, and undermined it when inconsistent or rigid. https://doi.org/10.1002/hrm.21733
📃 Smith, Dinev and Xu (2011) Interdisciplinary review of information privacy, and why privacy is contextual. https://doi.org/10.2307/41409970
📃 Petronio (2010) Communication privacy management: shared private information needs agreed rules, or turbulence follows. https://doi.org/10.1111/j.1756-2589.2010.00052.x
📃 Slepian, Chun and Mason (2017) The burden of a secret comes mostly from how often the mind returns to it. https://doi.org/10.1037/pspa0000085
📃 Slepian and Moulton-Tetlock (2019) More than 800 people, over 10,000 secrets. Confiding predicted wellbeing through a greater sense of coping. https://doi.org/10.1177/1948550618765069
📃 Kruger and colleagues (2005) Five experiments. Sarcasm over email was read at chance level, and senders had no idea. https://doi.org/10.1037/0022-3514.89.6.925
📃 Byron (2008) Why email flattens positive feeling and sharpens mild negative feeling in the reader. https://doi.org/10.5465/amr.2008.31193163
📃 Gonzales, Hancock and Pennebaker (2010) 70 groups, 324 people. Matching of function words predicted how cohesive a group was. https://doi.org/10.1177/0093650209351468
📃 Ireland and Pennebaker (2010) 1,744 students. Language style matching happened automatically, and people could not increase it on purpose. https://doi.org/10.1037/a0020386
📃 Niederhoffer and Pennebaker (2002) A counterweight: style matching was unrelated to how good participants and judges thought the interaction was. https://doi.org/10.1177/026192702237953
📃 Duthler (2006) Requests made by email came out more polite than the same requests left as voicemail. https://doi.org/10.1111/j.1083-6101.2006.00024.x
📃 Cheshin, Rafaeli and Bos (2011) Emotion carried in text alone shifted how other people in a virtual team felt. https://doi.org/10.1016/j.obhdp.2011.06.002
📃 Sussman and Sproull (1999) 117 participants. People distorted bad news least when they had to write it down. https://doi.org/10.1287/isre.10.2.150
📃 Waldvogel (2007) Greetings and closings in workplace email varied hugely between organisations rather than by closeness. https://doi.org/10.1111/j.1083-6101.2007.00333.x
📃 Sherblom (1988) People signed their electronic mail differently depending on whether they wrote up, down or across. https://doi.org/10.1177/002194368802500403
📃 Bunz and Campbell (2004) People accommodate to the politeness markers of whatever email arrives. https://doi.org/10.1080/08824090409359963
📃 Skovholt, Grønning and Kankaanranta (2014) Emoticons in workplace email work as hedges and politeness markers, not as feelings. https://doi.org/10.1111/jcc4.12063
📃 Koppel, Schler and Argamon (2009) Authorship attribution. Writers are identifiable largely from the small function words they never notice. https://doi.org/10.1002/asi.20961
📃 Tausczik and Pennebaker (2010) Text analysis and why pronouns and function words carry more about a person than vocabulary does. https://doi.org/10.1177/0261927x09351676
📃 Kacewicz and colleagues (2014) Five studies, including real email. Higher status meant fewer I's and more we's and you's. https://doi.org/10.1177/0261927x13502654
📃 Hülsheger and Schewe (2011) Meta-analysis of three decades of emotional labour. Surface acting went with exhaustion and poorer wellbeing. https://doi.org/10.1037/a0022876
📃 Lim and Teo (2009) Cyber incivility in email and what it did to employees' attitudes and behaviour. https://doi.org/10.1016/j.im.2009.06.006
📃 Giumetti and colleagues (2013) Rude email damaged mood, energy, engagement and performance compared with supportive email. https://doi.org/10.1037/a0032851
📃 Daft and Lengel (1986) Uncertainty is solved by more data; equivocality is solved only by people talking. https://doi.org/10.1287/mnsc.32.5.554
📃 Cohen and Hilligoss (2010) Review of hospital handoffs. Highly variable, thinly studied, and often behind communication failures. https://doi.org/10.1136/qshc.2009.033480
📃 Starmer and colleagues (2014) A structured handover programme across nine hospitals cut medical errors without making handover longer. https://doi.org/10.1056/nejmsa1405556
📃 Arora (2005) Critical incident study of patient sign-out. Communication failures produced uncertainty, inefficiency and delay. https://doi.org/10.1136/qshc.2005.015107
📃 Haynes and colleagues (2009) A nineteen-item checklist read aloud before surgery, associated with fewer deaths and complications in eight hospitals. https://doi.org/10.1056/nejmsa0810119
📃 Semmer and colleagues (2010) Illegitimate tasks, meaning work that violates what can reasonably be asked of you, related to counterproductive behaviour. https://doi.org/10.1111/j.1464-0597.2009.00416.x
📃 Semmer and colleagues (2015) Three studies. Illegitimate tasks predicted low self-esteem, resentment and burnout, and predicted strain over two months. https://doi.org/10.1080/02678373.2014.1003996
📃 Björk and colleagues (2013) Illegitimate tasks as a stressor, alongside organisational control and missing resources. https://doi.org/10.1080/02678373.2013.818291
📃 Morrison (1994) Employees differ in what they count as part of the job, and the wider the definition the more they do. https://doi.org/10.2307/256798
📃 Bolino and Turnley (2005) Individual initiative related to role overload, job stress and work-family conflict. https://doi.org/10.1037/0021-9010.90.4.740
📃 Bergeron (2007) The paradox where citizenship behaviour helps the organisation and costs the individual's career. https://doi.org/10.5465/amr.2007.26585791
📃 Babcock and colleagues (2017) Tasks nobody wants. Women volunteered more, were asked more and accepted more. https://doi.org/10.1257/aer.20141734
📃 Daniels (1987) The original account of invisible work and why it gets devalued. https://doi.org/10.1525/sp.1987.34.5.03a00020
📃 Hatton (2017) Three mechanisms, cultural, legal and spatial, that make work invisible and therefore cheap. https://doi.org/10.1177/0950017016674894
📃 Truss and colleagues (2013) 1,011 secretaries surveyed. Better qualified and better paid, with task content and prospects largely unchanged. https://doi.org/10.1111/j.1468-0432.2012.00587.x
P.P.S. Why Do I Even Have the Nerve to Write This?
Fair question.
I am someone who has spent years working alongside remarkable executives, navigating chaos, translating vision into structure and figuring things out, sometimes beautifully and sometimes the very hard way. I have had the privilege of being the steady right hand to leaders who move fast, take risks and expect a lot. That environment has shaped me more than any classroom ever could.
I also happen to have an MBA, and I am studying psychology because people fascinate me. How we work. Why we disconnect. What actually holds us together when the pressure rises.
But really, none of that is the point.
I am here because I have lived the erosion and rebuilt from it. I have seen the cost of carrying too much and the relief that arrives when you finally stop disappearing inside your own competence.
So I share what I have learned in case it helps someone else. Take what is useful. Leave the rest. And remember this one truth that professionals like you often forget:
You are probably doing far better than you give yourself credit for.
Comments