.png)
The Two Questions That Decide It: How to Tell Whether a Chief of Staff Role Is Actually Yours
Two questions arrive every single time, usually in the same breath, and usually from somebody who has been turning this over quietly for months. What should I go and get. And where does it lead.
I used to have a tidy answer ready for both. I do not any more, because the evidence does not support the tidy answer, and what it supports instead turns out to be far more useful than the thing I was saying.
This is the second half of a longer piece. The first half is about what these two jobs actually are, and if you have not read it the short version is this: the Chief of Staff role has no fixed shape, because it is made in the pairing between one executive and one particular person, which is why nobody can tell you where it sits or what it leads to from the title alone. Ten live adverts for the role, read and counted, agreed on only two things out of twenty.
Read part one first EA vs Chief of Staff: two jobs, not two rungs. What the research says about each of them, including the ten live adverts and what they did and did not agree on.
This half is the part you can run this week.
What you will get out of this
What to go and get, and where this leads, with the honest answer to both and the thing that actually predicts advancement in the evidence that exists
What a fit question really is, and why it is four questions rather than one
What a check like mine can and cannot tell you, with the numbers drawn rather than asserted
Why fit is built rather than discovered, and how quickly it happens
The two questions that decide whether one of these roles will work, and the one thing worth negotiating before you say yes
A free interactive check that gives you four readings rather than a score, and printable pages to take into the conversation
---
---
What to go and get, and where it leads
These are the two questions everybody asks next, and I have had both of them put to me in the same breath more times than I can count. The evidence answers them in a way that is useful, and not at all the way the market around this role would answer them.
Take where it leads first. Nobody knows.
I am not being coy. The corporate study I keep coming back to lists the long-term career trajectories of chiefs of staff as one of four things still to be researched. That is the authors of the only substantial study of the role saying, in 2024, that where people go afterwards is an open question. So when somebody tells you confidently that this job is the route to a chief operating officer seat, or to the board, they are not citing anything. They may well turn out to be right. There is no evidence either way, and you should price that into any decision that rests on it. The political literature does have findings about where chiefs of staff go next, and those are features of political appointment systems rather than facts about corporate careers, so I am leaving them out.
What is evidenced, and evidenced well, is something a good deal more useful than a destination. Across a large American firm, compensation, positive performance evaluations, promotions and good ideas were disproportionately held by the people whose networks spanned the gaps between groups. A separate study found that position in the workflow and communication networks was strongly related to how influential both supervisors and non-supervisors thought somebody was, and to promotion into supervisory roles, with contacts beyond the normal requirements of the job mattering particularly. A later experiment settled the direction of it: access to those gaps causes a person to be seen as a leader, rather than the other way round.
What predicts advancement, in the evidence that exists, is where you sit. Not what you are called.
Now the other question, and this is the one where I have changed my own mind.
The 2012 survey of a thousand and eleven people in assistant roles found that they were, by then, better qualified and generally well paid. It also found that career prospects remained poor. Sit with the order of those two findings. Qualification levels rose. The route out did not open. That is one occupation at one point in time and I would not build a law on it. It is also the only large-sample evidence available, and it points away from the certificate as the lever.
The literature that matters most here is about what happens when you hold more than the job uses, because that is where a great many people reading this already are.
Read those together, because the picture they make is precise. People doing this work without the title are performing perfectly well, feeling considerably worse, and quietly looking elsewhere. A follow-up study found the effect bites hardest in the people for whom the career matters most, which is to say the people most likely to have read this far.
Qualification levels rose. The route out did not open. The certificate is not the lever, and the evidence on what is has been sitting in the network research for forty years.
So the honest answer to "what should I go and get" is that another credential is unlikely to be the thing that moves this, and the evidence on what does move it points at the work you do and the position you occupy between groups.
I want to be careful with that, because it is not an argument against learning anything, and I would be arguing against my own life's work if it were. Learning that changes what you do on a Monday is exactly what the crafting evidence a few sections from now supports. The distinction worth holding is between a credential, which is a signal you hand to somebody else, and a capability, which changes the work in your hands. The evidence is unkind to the first as a lever for this particular move, and supportive of the second.
Try this. Before you book anything, write down what you would do differently on the Monday after it. If you cannot finish that sentence, what you are buying is a credential rather than a capability, and on this evidence the credential is not what moves this.
The research behind this section
Zhou, Hiller, Zaccaro, Campbell, McCauley, Parris and Klimoski (2024) list the long-term career trajectories of chiefs of staff as one of four domains in their future research agenda, meaning no peer-reviewed evidence currently supports any claim about where corporate chiefs of staff go next. Burt (2004), with over 3,800 citations, studied managers in a large American electronics company and found that compensation, positive performance evaluations, promotions and good ideas are disproportionately in the hands of people whose networks span structural holes, and that brokers between groups were more likely to express ideas, less likely to have ideas dismissed and more likely to have ideas evaluated as valuable. Brass (1984) analysed positions in workflow, communication and friendship networks and found them strongly related to perceptions of influence by both supervisors and non-supervisors and to promotions to the supervisory level, with contacts beyond the normal work requirements particularly important for acquiring influence. Burt, Reagans and Volvovsky (2021) established experimentally that access to structural holes causes a person to be perceived as a leader, while noting that in small teams this depended on one person emerging as the sole broker. Kwon, Rondi, Levin, De Massis and Brass (2020) review the field and distinguish brokerage structure from brokerage behaviour, a distinction that matters because somebody may occupy the position without engaging in the behaviour or the reverse. Lee, Quintane and colleagues (2024) supply the counterweight, finding that brokering is associated with burnout and with abusive behaviour towards others. Truss, Alfes, Shantz and Rosewarne (2012) surveyed 1,011 secretaries and found them better qualified and generally well paid while career prospects for all remained bleak. Harari, Manapragada and Viswesvaran (2017) meta-analysed twenty-five years of research on perceived overqualification across 61 studies and found it associated with job satisfaction at rho = -0.41, organisational commitment at -0.35, turnover intentions at 0.37, job search behaviours at 0.30 and psychological wellbeing at -0.26, while its association with task performance was 0.04 and with creative and innovative performance 0.04; it correlated with objective overqualification at 0.40 and with education at only 0.08. Liao, Zhang, Carnevale, Huang and Wang (2024) provide the current synthesis across 704 effect sizes from 251 independent samples and 87,229 people, and treat perceived overqualification as having both an upside and a downside depending on conditions. Erdogan, Tomás and colleagues (2018) followed 143 new graduates across two time periods and found perceived overqualification interacted with career centrality to predict relative deprivation, so the effect was strongest for those to whom the career mattered most. Erdogan and Bauer (2021) review the literature. Merluzzi and Phillips (2016) found a specialist discount among elite MBAs, and the wider question of whether breadth or depth pays better remains disputed across at least five studies that reach different conclusions, with Byun and Raffiee (2023) suggesting breadth wins when a job separation is involuntary and depth when it is voluntary.
---
What a fit question actually is
"Is this role for me" sounds like one question. I have asked it of myself in exactly that shape, and it is the shape that makes it unanswerable.
In the research it is four questions, and the four do not move together. Fit with the job. Fit with the organisation. Fit with the immediate group. Fit with the supervisor. The organising meta-analysis covers a hundred and seventy-two studies and eight hundred and thirty-six effect sizes, and nearly all of its credibility intervals excluded zero, which in plain terms means these relationships hold up across settings rather than appearing in some studies and quietly vanishing in others.
Here is the number that should change how you think about your own situation, and here is what it actually looks like.
Apply that to yourself and it stops being abstract. You can want the Chief of Staff work and not want it in this company. You can be entirely suited to the job and badly suited to the executive. These are not the same finding about you, and a single blended score tells you nothing you can act on.
Inside fit with the job there is a further split, and it matters more than any other distinction in this post. One half asks whether you can do what the job demands. The other asks whether the job supplies what you want. Demands and abilities on one side, needs and supplies on the other. Almost everybody I have talked to about this move spends their worry on the first half and almost none of it on the second, and the evidence suggests the second half is doing a great deal of the work. In one two-year study, the half about whether the job supplied what people needed predicted who was still there later. The half about whether they were able to do it predicted something stranger, which I will come back to.
You can be entirely suited to the job and badly suited to the executive. In the research those are two different findings, and only one of them is about you.
There is a live disagreement about which kinds of fit matter most, and it is worth naming rather than tidying away. A study of five hundred and seventy-one French employees found fit with the job and fit with the organisation more strongly related to satisfaction, burnout and intention to leave than fit with the group or the supervisor. A meta-analysis across ninety-six studies and three world regions found that the relative weight of those two families varies systematically by culture. The probable resolution is that both are right about their own setting. The practical reading, for a role defined by a pairing with one person, is that you should weigh the supervisor question more heavily than a general model would tell you to.
Try this. Split your hesitation in two on paper. On the left, what you are not sure you could do. On the right, what you are not sure the job would give you. Most people find the right-hand column is shorter and more decisive, and it is the one they have been ignoring.
The research behind this section
Kristof-Brown, Zimmerman and Johnson (2005) meta-analysed 172 studies yielding 836 effect sizes on person-job, person-organisation, person-group and person-supervisor fit against pre-entry and post-entry criteria including attitudes, performance, withdrawal behaviours, strain and tenure, and reported that nearly all credibility intervals excluded zero, indicating broad generalisability across situations; the per-fit-type point estimates sit in the article's tables and could not be retrieved, so none is quoted here. Lauver and Kristof-Brown (2001) surveyed 231 employees of a national trucking company, 104 office personnel and 127 drivers, and found a correlation of r = .18 between self-reported person-job and person-organisation fit, with both having a unique effect on job satisfaction and intent to quit, person-organisation fit a better predictor of intentions to quit, and no relationship between perceived person-job fit and task performance. Andela and van der Doef (2018) studied 571 employees in France across four fit types and found person-job and person-organisation fit more strongly correlated with job satisfaction, burnout and turnover intention than person-group and person-supervisor fit, with job satisfaction fully mediating several of those relationships. Oh, Guay and colleagues (2014) meta-analysed 96 studies and 110 independent samples across East Asia, Europe and North America and found the effects of rational fit, meaning person-organisation and person-job fit, relatively stronger in North America and to a lesser extent Europe, while the effects of relational fit, meaning person-group and person-supervisor fit, were relatively stronger in East Asia, concluding that fit matters everywhere while its composition varies. Verquer, Beehr and Wagner (2003) meta-analysed 21 studies and found mean effect sizes from -.18 for intent to turnover to .28 for organisational commitment, with subjective fit measures and value congruence producing larger effects. Boon and Biron (2016) followed 160 employees in a Dutch elderly care organisation across two survey waves with turnover data two years later, and found higher needs-supplies fit associated with lower turnover, while among employees in high-quality leader-member exchange relationships the demands-abilities dimension of person-job fit was associated with higher turnover.
---
What a questionnaire like mine can and cannot tell you
Before you use the check in this post I want to tell you what it is, because this is the section I would want to read first if somebody handed me a career quiz.
Start with how accurately any of us assess our own abilities. A review of twenty-two meta-analyses found the average correlation between how people rate their own ability and how they actually perform is .29. Not nothing, and not a lot. The same review found the relationship is stronger when the self-evaluation is about a specific area rather than a broad one, and when the task is objective or familiar. A separate analysis put the figure at .33 and found it improved when people were asked to rate themselves against a named comparison group rather than in the abstract.
That is a design instruction rather than a disappointment, and it is why the check below asks what you did last month rather than whether you are strategic. A narrow, concrete, comparative question gets a more accurate answer out of a person than a broad one about their qualities. The same logic shows up in a meta-analysis of proactivity, which found that scales asking about proactive behaviour predicted performance beyond personality, while scales asking about proactive personality added very little.
There is a related finding worth having. Research on self-estimated ability suggests these estimates reflect personality more than they reflect measured ability, and that professional interests track the self-estimates rather than the abilities. So a questionnaire about your interests is substantially a questionnaire about your self-concept, which is a real and useful thing to measure as long as nobody pretends it is a measurement of your capability.
Now the limits of this entire class of instrument, because you deserve them before you spend ten minutes on mine.
Across a century of evidence, interest inventories pick a person's eventual career choice about half the time. Fifty point eight per cent, to be exact about somebody else's number. On satisfaction, a meta-analysis of a hundred and five studies and nearly forty thousand people put the relationship between interest fit and job satisfaction at .19, which is lower than the people selling these instruments tend to imply, and notably stronger for how people rate their overall career path than for how they feel about a given Tuesday. A separate team argues that single-index measures of this kind misrepresent the shape of the relationship entirely. That dispute is unresolved, and it affects how any score of this kind should be read.
And on the question people actually want answered, which is whether fit predicts doing well, the figure is sobering. A meta-analysis assessing fit as a selection tool found it predicted job performance at .15, with the lower bound of its credibility interval including zero, against a stronger .31 for its relationship with attitudes. Fit tells you something fairly reliable about how you will feel, and something much weaker about how you will perform.
One last correction, because a post about self-assessment that quotes the famous version of this without the update is behind the evidence. The well-known finding that the least competent people most overrate themselves has been re-examined with a sample of nine hundred and twenty-nine people, using tests designed for the purpose, and the authors found an essentially linear relationship and no significant pattern of the kind the original proposed. So the standard warning that you cannot trust your own read on yourself, because incompetence is invisible to itself, is weaker than it used to be. Your read on yourself is moderately good. Treat it as evidence rather than as proof.
A questionnaire about your interests is substantially a questionnaire about your self-concept. That is worth measuring. It is not a measurement of what you can do, and anybody who tells you otherwise is selling something.
There is also no validated instrument for this specific role. None exists. The check in this post is assembled from validated general measures and applied to this job: perceived fit items, the role ambiguity and role conflict items from a 1970 paper that is still the standard, the split of ambiguity into knowing the goals and knowing the processes, and a job crafting scale. I have said so where each one appears, and it gives you four separate readings rather than one number, for the reason the picture in the last section showed you.
Try this. Before you open the check, write down the answer you expect it to give you. People usually know. What the tool is for is showing you which of the four readings your instinct was built on, and which one you had not considered at all.
The research behind this section
Zell and Krizan (2014) reviewed 22 meta-analyses and found a mean correlation of .29 between ability self-evaluations and objective performance, with the relationship stronger when self-evaluations were specific to a domain rather than broad, and when tasks were objective, familiar or low in complexity. Freund and Kasten (2012) meta-analysed self-estimates of cognitive ability and reported r = .33, improving when relative scales with specified comparison groups were used. Neubauer and Hofer (2021) found self-estimates of ability reflect personality more than measured ability, and that professional interests are mostly a function of self-estimates rather than measured abilities. Tornau and Frese (2012) meta-analysed proactivity constructs and found proactive behaviour scales predicted performance beyond personality while proactive personality scales showed very low to no incremental validity. Hanna and Rounds (2020) reviewed a century of evidence on interest inventories and found they correspond to the eventual career choice 50.8 per cent of the time. Hoff, Song and colleagues (2020) systematically reviewed 105 studies spanning 65 years, k = 194 and N = 39,602, and found interest fit related to overall job satisfaction at rho = 0.19 with a 95 per cent confidence interval of 0.16 to 0.21, with the relation notably stronger for satisfaction facets capturing how people evaluate their career choice in general, and concluded that the applied importance of vocational interests needs reconceptualising. Wiegand, Drasgow and Rounds (2021) used polynomial regression and response surface methodology across two large samples and argued that misfit is asymmetric, that single-index fit measures misrepresent the relationship, that the form of asymmetry changes across fit comparisons and across interest types, and that satisfaction sometimes increases for fit at high levels of both person and environment, concluding interests follow a complementary needs-supplies model rather than a supplementary one. Nye, Su, Rounds and Drasgow (2012, 2017) analysed 60 years of research and subsequently 92 studies and 1,858 correlations and found congruence indices a stronger predictor of performance than interest scores alone, with baseline correlations of 0.32 and 0.16 respectively, against a competing meta-analysis that had reached a different conclusion. Arthur, Bell and colleagues (2006) meta-analysed person-organisation fit as a selection predictor and found estimated true criterion-related validities of .15 for job performance across 36 studies and 5,377 people and .24 for turnover across 8 studies and 2,476 people, compared with .31 for work attitudes across 109 studies and 108,328 people, with the lower 95 per cent credibility values for performance and turnover including zero. Gignac and Zajenkowski (2020) tested the Dunning-Kruger hypothesis with n = 929 using the Glejser test and nonlinear regression and found no significant heteroscedasticity and an essentially linear relationship, both contrary to the hypothesis, concluding the effect may be much smaller than previously reported. The instruments the check draws on are Chuang, Shen and Judge (2016) on perceived person-environment fit, Rizzo, House and Lirtzman (1970) on role ambiguity and role conflict, Sawyer (1992) on the two-factor split of ambiguity into knowing the goals and knowing the processes, Lopper, Horstmann and Hoppe (2023) on approach and avoidance job crafting, and Parker, Williams and Turner (2006) on role breadth self-efficacy. No validated chief of staff selection or fit instrument exists.
---
Fit is built, and it is built faster than you would think
Everything above treats fit as something to be discovered, as though there were a fact about you sitting in a drawer waiting for a good enough questionnaire to find it. The evidence says that is only half of it, and the other half is the half with something you can do in it.
Over a much longer horizon, a twelve-year study found that fit between people's interests and their work improved over time, and that the mechanism was selection rather than socialisation. People did not gradually come round to liking what they were doing. They moved towards what suited them by making moves, repeatedly, over years.
And in a study of managers in newly created roles, the structure of the role and the discretion inside it developed together. The authors put it plainly: new managerial roles do not have to be static.
Put those three together and the check in this post changes meaning. A low reading is information about now. It is not a verdict about you, and the thing it most often means is that you have not yet done the version of this work that would show you the answer.
This is also the honest thing to say to the person already doing Chief of Staff work without the title. You are not waiting for permission to find out whether this suits you. You are already running the experiment.
Two cautions, because I am not going to sell you crafting as a free lunch.
The first is a real trade-off. The meta-analysis most often cited for crafting found that taking on more challenging demands was associated with how others rated your performance at a correlation of .42, which is substantial. A later study found that crafting demands, both the challenging kind and the obstructive kind, increased job strain by increasing job demands, and its authors concluded that people may have difficulty crafting demands in a way that produces only the good outcomes. Another found a reciprocal relationship over time between the defensive sort of crafting and burnout. Taking on more of this work will raise how you are seen and raise what it costs you. Both halves are true, and you should plan for both rather than be surprised by the second.
The second is about the strength of the evidence itself. The association between crafting and good outcomes rests on a hundred and twenty-two samples and over thirty-five thousand workers. The evidence that running a crafting intervention produces good outcomes rests on fourteen studies, with one confidence interval that nearly touches zero. So "people who shape their jobs tend to do better" is much better supported than "doing a shaping exercise will make you do better", and I would rather give you the first than sell you the second.
And the strange finding I promised you. In that two-year study, among people who had a good relationship with their manager, being well able to do what the job demanded predicted a higher chance of leaving. Not lower. The most likely reading is that capability plus a manager who notices it produces options. Which is worth knowing before you decide that becoming very good at this job is a way of staying in it.
Try this. Pick one piece of Chief of Staff work you are not currently asked to do, and do it once this month without announcing it. A pre-read before the leadership meeting. A one-page options note instead of a verbal update. Then notice two things: what it changed in the room, and what it cost you in the week. You now have data from the only experiment that counts.
The research behind this section
Tims, Derks and Bakker (2016) ran a weekly study and found that employees who crafted their job resources and demands reported higher person-job fit the following week, that demands-abilities fit then related to greater meaningfulness by the final week, and found no support for the reverse causal models. Hanna, Briley and colleagues (2021) followed interest fit across 12 years and found fit improved over time, with the mechanism being selection rather than socialisation. Sandhu and Kulik (2019) studied managers in newly created roles and found that role structure and discretion co-evolved, stating that new managerial roles do not have to be static, and identified semi-structured formalisation combined with a decentralised programme as the configuration in which new-role managers could actually use their discretion. Rudolph, Katz, Lavigne and Zacher (2017) meta-analysed job crafting across 122 independent samples and 35,670 workers and found increasing challenging job demands associated with other-rated performance at rc = 0.422. Holman and colleagues (2023) found that both challenge and hindrance demand crafting increased job strain via increases in job demands and concluded that employees may have difficulty crafting job demands in ways that produce beneficial outcomes. Lichtenthaler and Fischbach (2018) found a reciprocal positive relationship over time between prevention-focused job crafting and burnout. Oprea and colleagues (2019) meta-analysed job crafting interventions and found only 14 eligible studies with effects ranging from g = 0.19 to g = 0.44, including a confidence interval on contextual performance of 0.01 to 0.78. Boon and Biron (2016) found that among employees in high-quality leader-member exchange relationships, demands-abilities fit at the second wave was associated with higher turnover two years later, as above.
---
---
The two questions that decide it, and the one thing to negotiate
If you take one practical thing from this post, take this section.
A Chief of Staff role that goes wrong usually does not go wrong because the wrong person took it. It goes wrong because nobody defined it, and the research on that particular failure is old, large and clear.
A meta-analysis separated two things that get confused constantly. Role ambiguity is not knowing what you are supposed to be doing. Role conflict is being pulled between demands you understand perfectly well. Against job performance, ambiguity correlated at minus .21 and conflict at minus .07. Read those two numbers next to each other, because the difference between them is the most useful thing in this post. Being pulled in three directions by three executives who all want different things is uncomfortable, and it barely touches how well you do the work. Not knowing what the job is costs you three times as much. Both were associated with worse wellbeing across thirty-three studies and nearly twenty thousand people, so neither is harmless. They are not equally dangerous, though, and the one everybody complains about is the lesser one.
Then a 1992 paper split ambiguity itself into two things that turn out to be separate: knowing the goals, and knowing the processes by which you are meant to reach them. Two factors, fitting the data better than one. Which gives you your two questions, and they are the two I would ask in any conversation about a Chief of Staff role before I asked about money.
What is this role for. Not the duties. What would be different in a year if it worked.
And how am I meant to achieve it. What access, what authority to convene, whose agreement do I need and whose do I not.
If the answer to the first is a list of tasks, the role has no goal, and you will spend a year inventing one while being measured against somebody's unspoken version. If the answer to the second is "you will work that out", you have been handed the hardest part of the job and told it is a perk.
Now the thing to negotiate, which comes from that study of managers in new roles. The configuration in which those managers could actually use their discretion was neither a fully specified role nor an open one. It was semi-structured: a defined remit with a decentralised way of working inside it. Not a job description with forty bullets. Not a blank page. A clear statement of what the role is for, and room to decide how.
So the ask is specific. One sentence on what the role is for that you and the executive both sign up to, and explicit freedom on method. That is a reasonable thing to request, and it is the single highest-value thing you can get before you start, including the salary.
Not knowing what the job is costs three times what being pulled in several directions costs. The failure mode everybody worries about is the smaller one.
Try this. Ask the two questions out loud, of the person who would be your principal, before you decide anything. Write their answers down in their own words. If you cannot get a sentence for the first one, that is the answer to the whole question this post is about, and it has saved you a year.
The research behind this section
Rizzo, House and Lirtzman (1970) developed the role conflict and role ambiguity scales that remain the standard measures. Tubre and Collins (2000) meta-analysed role ambiguity, role conflict and job performance and found role ambiguity correlated with job performance at r = -.21 and role conflict at r = -.07. Sawyer (1992) distinguished knowing the goals from knowing the processes as two separate components of role ambiguity and found a two-factor model fitted better than a single factor. Schmidt and colleagues (2014) meta-analysed 33 studies covering 19,926 people and found role ambiguity and role conflict both associated with worse mental wellbeing, at r = .279 and r = .318 respectively. Jackson and Schuler (1985) meta-analysed 29 correlates of role ambiguity and role conflict across four categories; the magnitudes were not retrievable and none is quoted. Sandhu and Kulik (2019) identified semi-structured formalisation with a decentralised programme as the configuration that allowed managers in newly created roles to exercise discretion, as above. Peij, Bezemer and Maassen (2015) studied role ambiguity and role conflict specifically among company secretaries on two-tier boards, the closest available application of these constructs to an administrative governance role. Rousseau, Ho and Greenberg (2006) and the subsequent idiosyncratic deals literature cover the negotiation of individual role arrangements.
---
The check
The tool below gives you four readings rather than a score, for the reason the cloud of dots showed you earlier.
The first reading is what you are already doing, asked as what you did in the last month rather than as what sort of person you are. The second is what you want the work to give you, which is the half of the fit question most people skip. The third is whether the role in front of you would exist properly, built from the two questions. The fourth is whose, because the role is contingent on the pairing and the research is unambiguous about that.
It will tell you which of the four is thin, what to do about that one, and what to ask for. It will not tell you whether you would be good at it, because the evidence says an instrument like this predicts how you will feel far better than it predicts how you will perform, and I would rather build you something honest than something impressive.
At the end it writes you a note you can copy. Take it into the conversation, or bring it to me and we will go through it together.
---
Where to start
The one thing, if you do nothing else: ask the two questions, of the specific person, and write down their actual words. Everything else in this post is preparation for that conversation.
Then three moves. Run the check and look only at your thinnest reading, because that is where the next month of effort belongs. Do one piece of Chief of Staff work this month that nobody asked you for, and notice both what it changed and what it cost. And write the sentence, the one about what the role would be for, in your own words first, so that you are not standing about waiting for somebody else to supply it.
The Two Jobs Pages below put the useful parts on paper: the two questions with room for their answers, the four readings with what each one means, and a one-page comparison of the two roles as the evidence actually describes them rather than as the job adverts do. Free, and no email wall.
I have spent time in both of these seats, and the thing I wish somebody had told me is that the title arrives last, if it arrives at all, and that waiting for it is the only truly wasted move available to you. The work is yours to start. Whether you ever get the word on your email signature is a separate question, decided by other people, and considerably less interesting than what you are actually able to do on a Tuesday.
Working out which version of this role you are already doing, and whether to build towards it, is a question about whether you shape what happens or only run it, which is what the Connected Leader Assessment measures under Strategic Contribution. It is free, it takes about twelve minutes, and it gives you a read across all five dimensions rather than the one.
And if you want a second pair of eyes on what the check gives you, that is what the Mentoring Packages are for. Bring the note it writes you, bring the two answers your executive gave you, and we will work out together what they actually mean. The Irreplaceable Audit is the other thing worth your time here, because it asks what the shape of your role is worth rather than what it is called.
Stay savvy.
Meg ✌️
Free tools
The two jobs check, the interactive tool in this post, free
The Two Jobs Pages, the download in this post, free
The Connected Leader Assessment, twelve minutes, free, with a read on Strategic Contribution, the dimension this post sits in
The Irreplaceable Audit, free, for what the shape of your role is worth rather than what it is called
P.S. Want to Go Deeper on Fit and the Two Questions?
If this resonated and you want to go further, here is the evidence this half is built on. These are the papers behind what the brokerage research actually shows about advancement, what holding more than the job uses does to a person, what a fit question really consists of, what a self-assessment can fairly claim, how quickly fit is built, and why not knowing what a job is costs so much more than being pulled between things you understand. Part one carries its own list for the two roles themselves.
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, and where I could not read a paper at all I have said so in the body rather than quote it.
📃 Zhou and colleagues (2024) Mixed-methods study of 2,500 LinkedIn profiles, 108 surveys and 13 interviews framing the corporate chief of staff as quasi-executive, influencing the chief executive, top team and wider organisation, with four domains still unresearched. https://doi.org/10.1177/15480518241267078
📃 Truss and colleagues (2012) Survey of 1,011 UK secretaries finding they are better qualified and generally well paid, that a minority undertake complex managerial tasks, and that career prospects for all remain poor. https://doi.org/10.1111/j.1468-0432.2012.00587.x
📃 Truss, Goffee and Jones (1995) Questionnaire and interview study of secretaries in England, France and Germany, plus their principals, introducing the secretary as an agent actively shaping her own work identity. https://doi.org/10.1177/001872679504801105
📃 Burt (2004) Study of managers in a large American electronics company finding that compensation, positive evaluations, promotions and good ideas sit disproportionately with people whose networks span between groups. https://doi.org/10.1086/421787
📃 Brass (1984) Analysis of workflow, communication and friendship networks finding position strongly related to perceived influence and to promotion into supervisory roles, with contacts beyond the job mattering most. https://doi.org/10.2307/2392937
📃 Burt, Reagans and Volvovsky (2021) Experimental study establishing that access to the gaps between groups causes a person to be perceived as a leader, rather than the reverse. https://doi.org/10.1016/j.socnet.2020.09.002
📃 Kwon and colleagues (2020) Review of network brokerage distinguishing the structural position from the behaviour, so a person can hold one without the other. https://doi.org/10.1177/0149206320914694
📃 Lee and colleagues (2024) The counterweight on brokerage, finding that spanning between groups is associated with burnout and with abusive behaviour towards others. https://doi.org/10.1287/orsc.2023.1664
📃 Harari, Manapragada and Viswesvaran (2017) Meta-analysis of 61 studies over 25 years: feeling overqualified related to satisfaction at -.41, commitment -.35, intention to leave .37, and to task performance at .04. https://doi.org/10.1016/j.jvb.2017.06.002
📃 Liao and colleagues (2024) Current synthesis across 704 effect sizes, 251 samples and 87,229 people, treating overqualification as having both an upside and a downside depending on conditions. https://doi.org/10.1177/01492063241239298
📃 Erdogan and colleagues (2018) 143 graduates across two time periods, finding the cost of feeling overqualified was greatest for those to whom the career mattered most. https://doi.org/10.1016/j.jvb.2018.05.003
📃 Erdogan and Bauer (2021) Annual Review synthesis of the overqualification literature. https://doi.org/10.1146/annurev-orgpsych-012420-055831
📃 Merluzzi and Phillips (2016) Two cohorts of elite MBAs, where focused specialist profiles drew fewer offers and smaller starting bonuses, concluding the specialist advantage may be overstated. https://doi.org/10.1177/0001839215610365
📃 Byun and Raffiee (2023) Study of career specialisation and job displacement finding breadth pays when a separation is involuntary and depth when it is voluntary. https://doi.org/10.1177/00018392221143762
📃 Kristof-Brown, Zimmerman and Johnson (2005) Meta-analysis of 172 studies and 836 effect sizes across person-job, person-organisation, person-group and person-supervisor fit, in which nearly all credibility intervals excluded zero. https://doi.org/10.1111/j.1744-6570.2005.00672.x
📃 Lauver and Kristof-Brown (2001) Survey of 231 employees finding a correlation of only r = .18 between perceived person-job and person-organisation fit, with each having a distinct effect on satisfaction and intent to quit. https://doi.org/10.1006/jvbe.2001.1807
📃 Andela and van der Doef (2018) Study of 571 French employees across all four fit types, finding person-job and person-organisation fit more strongly related to satisfaction, burnout and turnover intention than person-group or person-supervisor fit. https://doi.org/10.1177/0894845318789512
📃 Oh and colleagues (2014) Meta-analysis of 96 studies and 110 samples across three world regions, finding rational fit stronger in North America and Europe and relational fit stronger in East Asia. https://doi.org/10.1111/peps.12026
📃 Verquer, Beehr and Wagner (2003) Meta-analysis of 21 studies with effects from -.18 for turnover intent to .28 for commitment, finding subjective fit measures produce larger effect sizes than other approaches. https://doi.org/10.1016/s0001-8791(02)00036-2
📃 Boon and Biron (2016) Study of 160 employees over two years finding higher needs-supplies fit predicted lower turnover, while for those with strong manager relationships demands-abilities fit predicted higher turnover. https://doi.org/10.1177/0018726716636945
📃 Zell and Krizan (2014) Metasynthesis of 22 meta-analyses finding a mean correlation of .29 between ability self-evaluations and objective performance, stronger when self-evaluations were specific to a domain rather than broad. https://doi.org/10.1177/1745691613518075
📃 Freund and Kasten (2012) Meta-analysis of 154 effect sizes in 41 studies estimating the relationship between self-estimated and measured cognitive ability at r = .33, improving when scales specified a comparison group. https://doi.org/10.1037/a0026556
📃 Neubauer and Hofer (2021) Two studies finding self-estimates of ability reflect personality more than measured ability, and that professional interests are mostly a function of those self-estimates rather than of actual abilities. https://doi.org/10.1016/j.paid.2020.109850
📃 Tornau and Frese (2012) Meta-analysis of 163 samples and 36,079 people finding proactive behaviour scales predicted job performance beyond personality, while proactive personality scales showed very low to no incremental validity. https://doi.org/10.1111/j.1464-0597.2012.00514.x
📃 Hanna and Rounds (2020) Meta-analysis spanning almost a century of research finding interest inventories attain an estimated overall hit rate of 50.8% for predicting career choice. https://doi.org/10.1037/bul0000269
📃 Hoff and colleagues (2020) Meta-analysis of 105 studies over 65 years, N = 39,602, finding interest fit relates to overall job satisfaction at rho = 0.19 and more strongly to how people evaluate their career choice. https://doi.org/10.1016/j.jvb.2020.103503
📃 Wiegand, Drasgow and Rounds (2021) Polynomial regression and response surface analysis of two large samples finding the interest fit to satisfaction relationship is often asymmetric and better described by a needs-supplies than a matching model. https://doi.org/10.1016/j.jvb.2020.103524
📃 Nye and colleagues (2012) Meta-analysis of 60 studies and roughly 568 correlations finding vocational interests related to performance and persistence, with congruence indices more predictive than interest scores alone. https://doi.org/10.1177/1745691612449021
📃 Nye and colleagues (2017) Meta-analysis of 92 studies and 1,858 correlations finding interest congruence a stronger predictor of performance than interest scores alone, at baseline correlations of 0.32 and 0.16 respectively. https://doi.org/10.1016/j.jvb.2016.11.002
📃 Arthur and colleagues (2006) Meta-analysis estimating person-organisation fit validity at .15 for job performance and .24 for turnover, with lower 95% credibility values including zero, against .31 for work attitudes. https://doi.org/10.1037/0021-9010.91.4.786
📃 Gignac and Zajenkowski (2020) Study of 929 adults finding no significant heteroscedasticity and an essentially linear link between measured and self-assessed intelligence, both contrary to the Dunning-Kruger hypothesis. https://doi.org/10.1016/j.intell.2020.101449
📃 Tims, Derks and Bakker (2016) Three-wave weekly study of 114 employees over 430 occasions finding those who crafted their resources and demands reported higher person-job fit the following week, with no support for reverse models. https://doi.org/10.1016/j.jvb.2015.11.007
📃 Hanna and colleagues (2021) Twelve-year longitudinal study using four analytic methods finding interest fit improved over time and that selection, not socialisation, was the main mechanism by which people achieved better fit. https://doi.org/10.1177/08902070211014022
📃 Sandhu and Kulik (2019) Interview study across 21 organisations finding new managerial roles took three structural configurations, with semi-structured formalisation and a decentralised programme best enabling the incumbent to use discretion. https://doi.org/10.1177/0001839218778018
📃 Rudolph and colleagues (2017) Meta-analysis of 122 samples and 35,670 workers finding overall job crafting associated with work engagement at rc = 0.450, and increasing challenging demands with other-rated performance at rc = 0.422. https://doi.org/10.1016/j.jvb.2017.05.008
📃 Holman and colleagues (2023) Meta-analytic modelling of 58 samples and 20,347 employees finding resource crafting improved well-being through job resources, while challenge and hindrance demand crafting raised strain by raising demands. https://doi.org/10.1111/joop.12450
📃 Lichtenthaler and Fischbach (2018) Meta-analysis finding promotion-focused crafting related positively to engagement and negatively to burnout, with prevention-focused crafting showing the reverse and a reciprocal relationship with burnout over time. https://doi.org/10.1080/1359432x.2018.1527767
📃 Oprea and colleagues (2019) Meta-analysis of 14 intervention studies finding small to moderate effects on crafting (g = 0.26) and work engagement (g = 0.31), with plans covering both organisational and personal objectives working best. https://doi.org/10.1080/1359432x.2019.1646728
📃 Rizzo, House and Lirtzman (1970) Development and testing of the standard questionnaire measures of role conflict and role ambiguity, showing the two to be factorially identifiable and independent in samples of managers. https://doi.org/10.2307/2391486
📃 Tubre and Collins (2000) Meta-analysis finding role ambiguity related to job performance at r = -.21, moderated by job type and rating source, while role conflict showed a negligible r = -.07. https://doi.org/10.1177/014920630002600104
📃 Sawyer (1992) Study of 402 mental health workers splitting role ambiguity into two constructs, knowing the goals and knowing the processes, with a two-factor model fitting better than a single factor. https://doi.org/10.1037/0021-9010.77.2.130
📃 Schmidt and colleagues (2014) Meta-analysis of 33 studies and 19,926 participants finding role ambiguity related to depressive symptoms at r = .279 and role conflict at r = .318, and supporting the two as distinct. https://doi.org/10.1080/1359432x.2012.711523
📃 Jackson and Schuler (1985) Meta-analysis examining 29 correlates of role ambiguity and role conflict across organisational context, individual characteristics, affective reactions and behavioural reactions, and critiquing the field's conceptual basis. https://doi.org/10.1016/0749-5978(85)90020-2
📃 Peij, Bezemer and Maassen (2015) Exploratory study of one hundred company secretaries in the Dutch two-tier board system, finding they significantly influence the governance framework while facing practical challenges in fulfilling diverse roles. https://doi.org/10.22495/cocv12i3p10
📃 Rousseau, Ho and Greenberg (2006) Conceptual paper distinguishing functional from dysfunctional idiosyncratic deals and developing propositions on how such deals are formed and how they affect both the holder and their coworkers. https://doi.org/10.5465/amr.2006.22527470
📃 Chuang, Shen and Judge (2016) Development and validation of the Perceived Person-Environment Fit Scale, a multidimensional instrument measuring person-job, person-organisation, person-group and person-supervisor fit separately. https://doi.org/10.1111/apps.12036
📃 Lopper, Horstmann and Hoppe (2023) Development and validation of the Approach-Avoidance Job Crafting Scale, measuring the hierarchical structure of job crafting including the approach and avoidance orientations. https://doi.org/10.1111/apps.12466
📃 Parker, Williams and Turner (2006) Study of 282 UK manufacturing workers finding proactive personality and job autonomy predicted proactive behaviour through role breadth self-efficacy and flexible role orientation. https://doi.org/10.1037/0021-9010.91.3.636
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