Skip to content

How much speaking does each job demand?

We compared 70 roles with all 905 occupations in the U.S. government's O*NET database on five measures of spoken communication. 45 of the 70 score above the typical occupation.

Compared
70 roles with 905 occupations
Calculated
November 1, 2026

Most of these roles ask for more speaking than the typical occupation. What separates them is the kind: calls, conversation in the room, or speaking to a group.

Where selected roles fall among 905 occupations

Each bar counts the occupations in a 5-point band of the speaking-demand score. Each marker is a role and its label says what share of the 905 occupations it scores above. Eight in ten occupations score between 60.8 and 83.8.

View as a table
Score bandOccupations
40 to under 451
45 to under 504
50 to under 5511
55 to under 6059
60 to under 65123
65 to under 70113
70 to under 75118
75 to under 80169
80 to under 85243
85 to under 9052
90 to under 9510
95 to under 1002

How to read the score

The score is not a grade. O*NET rates each measure from 1 to 5 and we rescale that to 0 to 100, where 100 is the top rating. A job would need the top rating on every part of the score to reach 100 and none does: the median occupation scores 76 and the highest scores 95.7. So a score in the 80s is a very high score here. The percentile says the same thing in plainer words: it is the share of the 905 occupations a role scores above.

The score averages how much talking matters to the job (Speaking and Oral expression) with how often the job involves its most frequent kind of talking. A customer service representative scores 100 on telephone conversations and only 16 on public speaking, so the phone counts and the missing stage does not: 83, above 86% of occupations. Teachers get there on public speaking (85) and face-to-face discussion (95) and score 85, above 93%.

This definition replaced an earlier plain average of all five measures, which put customer service at 66, above 45%, because most jobs score low on public speaking. Both numbers are in the dataset.

Four findings

  1. 45 of 70

    Most of these roles ask for more speaking than the typical occupation

    The median occupation scores 76 and the median of these roles 80.8. Of the 70, 34 are in the top quarter of all occupations and 3 in the bottom quarter.

  2. 10th percentile

    Software Engineer is the lowest of the 70 roles

    Software Engineer scores 60.9, higher than 10% of all 905 occupations. The other roles in the bottom quarter are Full Stack Developer and UX Designer.

  3. 27 of 70

    The phone is the most common standout

    27 roles stand out on telephone conversations, 17 on face-to-face discussion and 9 on public speaking. 17 have no standout and they include QA Engineer, Medical Assistant, Data Analyst and Sound Designer.

  4. 14.3%

    Public speaking is the rarest of the three demands

    Only 14.3% of all occupations score 50 or more on public speaking, against 85.9% for telephone conversations and 99.8% for face-to-face discussion. Teacher stands out among the roles.

Four kinds of speaking

Each role is placed by the one demand that stands out most against all occupations. Roles are listed from the strongest standout down and each name jumps to its row in the table.

Using it for your own preparation

  • Find your role

    Search the table below. Each role links to its interview guide, where one is written and the table shows the five measures for every job.

  • Look at the kind of speaking

    A role that stands out on the phone spends much of its day on calls. One that stands out in the room is mostly in conversation with people and one that stands out on stage addresses groups. Practice answers in the setting your job works in.

  • Do not read it as interview difficulty

    The score describes the job. It says nothing about how an interview for that job is run.

Practice an answer on cameraHow to practice for a video interviewInterview tips for your role

All 70 roles

Sorted by percentile: the share of all 905 occupations a role scores above. The columns on the right are the five individual measures, each 0 to 100. In the bar, the dot is the score and the shaded span is the 95% interval.

Level
Kind of speaking

Showing 70 of 70 roles, highest percentile first.

Roles compared with all occupations on spoken-communication demand
RolePercentileScore (95% interval)Kind of speakingSpeakOralFacePhonePublic
Entertainment Lawyerclose match99
94.991 to 99
On the phone9197869743
Lawyer99
94.991 to 99
On the phone9197869743
Executive98
88.985 to 93
On stage8188989852
HR Manager97
88.384 to 93
On the phone8188959644
Social Worker97
87.886 to 90
In the room8183999742
Film Director95
85.880 to 91
On the phone7588909543
Line Producerclose match95
85.880 to 91
On the phone7588909543
Showrunnerclose match95
85.880 to 91
On the phone7588909543
Speech Language Pathologist94
85.483 to 88
In the room75811007027
Teacherclose match93
85.183 to 87
On stage7783956385
Journalist92
84.477 to 92
On stage8185808845
Nurse Practitioner91
84.382 to 86
In the room757810010031
Recruiterclose match91
84.382 to 86
On stage78759810058
Talent Agent91
84.382 to 86
On the phone75788910025
Financial Manager89
83.680 to 88
In the room7578989041
Occupational Therapist89
83.681 to 86
In the room7578988132
Digital Marketing Managerclose match88
83.581 to 86
On the phone7578859842
Marketing Manager88
83.581 to 86
On the phone7578859842
Product Marketingclose match88
83.581 to 86
On the phone7578859842
Business Analystclose match88
83.481 to 86
On the phone7578949744
Customer Service Representative86
83.380 to 86
On the phone72786510016
Physician Assistant86
83.383 to 83
In the room75751009531
Pharmacist84
83.082 to 84
In the room7575999731
Architect83
82.981 to 85
On the phone7578909643
Administrative Assistant82
82.882 to 83
On the phone7575919815
Nurse82
82.882 to 84
In the room7575999933
Casting Director81
82.580 to 85
On the phone7578869537
Executive Assistant79
82.380 to 84
On the phone72759110010
Interior Designer79
82.381 to 83
On the phone7575949742
Supply Chainclose match79
82.380 to 84
On the phone72759010036
Operations Manager79
82.281 to 83
In the room7575979437
Real Estate Agent76
81.780 to 84
On the phone7575789522
Civil Engineer74
81.579 to 84
On the phone7275949833
Physical Therapist74
81.580 to 83
In the room7575958729
Restaurant Manager71
80.878 to 84
In the room7272989140
Financial Advisor66
80.078 to 82
On the phone7575789025
QA Engineer65
79.877 to 83
No standout7275927736
IT Manager62
79.377 to 82
On the phone7275918727
Sales Account Executiveclose match62
79.372 to 86
On the phone7269869721
Cardiac Medical Tech62
79.176 to 82
In the room6975948413
Medical Assistant54
77.172 to 83
No standout7275828427
Dental Hygienist52
76.973 to 81
In the room636999602
AI Engineerclose match52
76.673 to 80
On stage6375928553
Machine Learning Engineerclose match52
76.673 to 80
On stage6375928553
Paralegal52
76.672 to 81
On the phone6672928919
Data Analystclose match50
75.873 to 79
No standout7275808040
Production Designerclose match50
75.772 to 80
On stage6969908249
Sound Designerclose match49
75.471 to 80
No standout6675868122
Performance Marketerclose match46
74.370 to 78
No standout7272637935
Social Media Managerclose match46
74.370 to 78
No standout7272637935
Coloristclose match45
74.070 to 78
In the room5669978730
Dental Assistant45
74.069 to 79
On stage6075868855
Film Editor45
74.070 to 78
In the room5669978730
Video Editor45
74.070 to 78
In the room5669978730
Accountant44
73.870 to 78
On the phone6069879328
Animator44
73.870 to 78
No standout5675907724
Cybersecurityclose match44
73.869 to 79
On the phone6669878727
Cinematographerclose match42
73.569 to 78
In the room5666997019
Mechanical Engineer42
73.470 to 77
On stage6075858648
Project Manager41
72.868 to 78
No standout6372818441
Scrum Masterclose match41
72.868 to 78
No standout6372818441
Game Designer37
71.468 to 75
No standout5369935834
Copywriter36
71.066 to 76
No standout6972527225
Screenwriterclose match36
71.066 to 76
No standout6972527225
TV Writerclose match36
71.066 to 76
No standout6972527225
Data Scientist36
70.766 to 76
No standout6666816933
Graphic Designer34
69.665 to 74
On the phone6063828741
Full Stack Developerclose match23
65.360 to 71
No standout5656837924
UX Designer21
64.660 to 70
No standout5360738120
Software Engineer10
60.958 to 64
No standout5369506116

Where each measure peaks and bottoms out

Highest and lowest roles on each measure, with the all-occupation average
MeasureHighest roleLowest roleAll occupations
SpeakingSkill91 Entertainment Lawyer and Lawyer53 Game Designer, UX Designer and Software Engineer63 on average89.1% score 50 or more
Oral expressionAbility97 Entertainment Lawyer and Lawyer56 Full Stack Developer68 on average95.4% score 50 or more
Face-to-face discussionsWork context100 Speech Language Pathologist, Nurse Practitioner and Physician Assistant50 Software Engineer89 on average99.8% score 50 or more
Telephone conversationsWork context100 Nurse Practitioner, Recruiter and Talent Agent and 2 more58 Game Designer76 on average85.9% score 50 or more
Public speakingWork context85 Teacher2 Dental Hygienist33 on average14.3% score 50 or more

Common questions

Which jobs demand the most speaking?

Among the 70 roles, the highest scores go to Entertainment Lawyer (94.9), Lawyer (94.9), Executive (88.9), HR Manager (88.3) and Social Worker (87.8). The lowest are Software Engineer (60.9), UX Designer (64.6) and Full Stack Developer (65.3). Across all 905 occupations, the highest are Education Administrators, Broadcast Announcers and Radio Disc Jockeys and Lawyers.

What does the percentile mean?

A role at the 10th percentile scores higher than 10% of the 905 O*NET occupations. Each occupation counts once, however many people work in it, so the percentile describes occupations, not workers.

What are the kinds of speaking?

We compared each role with all occupations on three measures: telephone conversations, face-to-face discussions and public speaking. The highest of the three is the role's standout if it is at least 0.5 standard deviations above the typical occupation. 27 roles stand out on the phone, 17 in the room, 9 on stage and 17 have no standout mode.

Does a high score mean a harder interview?

No. The score describes the job, not the interview. We did not measure how interviews are run, so the study makes no claim about interview difficulty or style.

Where does the data come from?

The O*NET 31.0 Database, August 2026 Release, published by the U.S. Department of Labor under a CC BY 4.0 license. It holds survey ratings of how important skills and abilities are to each occupation and how often the work involves things like telephone calls and public speaking. Open the O*NET database page.

Why is my role missing?

19 of the 89 roles on this site have no defensible match in the O*NET database, or O*NET has no ratings for the matching occupation. Each is listed with its reason below and none was given an estimated score.

Method

  1. 1.Source

    The O*NET 31.0 Database (August 2026 Release), which holds survey ratings from workers and occupational analysts for each U.S. occupation. It is published by the U.S. Department of Labor and licensed CC BY 4.0. Download it at www.onetcenter.org/database.html.

  2. 2.Measures

    Five ratings that describe spoken communication: speaking (skill), oral expression (ability), face-to-face discussions (work context), telephone conversations (work context) and public speaking (work context). Each is on a 1 to 5 scale and was rescaled to 0 to 100.

  3. 3.Index

    The index is the mean of three numbers: Speaking, Oral expression and the strongest of Face-to-face discussions, Telephone conversations and Public speaking. The first two say how much talking matters to the job. The third says how often the job involves its most frequent kind of talking. This replaced an equal-weight mean of all five measures on 2026-11-01, after the first results showed that it scored phone-heavy jobs such as customer service at about the typical occupation, because almost every job scores low on public speaking. The earlier average is in the dataset as five_measure_average. The strongest of the three alone was tried and rejected, because most occupations score above 90 on it and it cannot tell jobs apart.

  4. 4.Baseline

    Every O*NET occupation that has all five ratings is the baseline: 905 of the 1016 in the database. The percentile is the share of those that score lower. Each occupation counts once, not weighted by how many people work in it.

  5. 5.Level

    A role is in the top quarter if it scores 81.5 or more, the bottom quarter if it scores 66 or less and the middle half otherwise. These cut-offs are the quartiles of all 905 occupations.

  6. 6.Kind of speaking

    For each role we took its z-score on telephone conversations, face-to-face discussions and public speaking against all occupations. The highest of the three is the standout if it is at least 0.5 standard deviations above the typical occupation; otherwise the role has no standout. Both rules were fixed before the results were computed.

  7. 7.Mapping

    Each role page was matched to the O*NET occupation that describes it. 46 roles match an occupation directly and 24 to the closest one, labelled "close match" in the table. Where a role spans several occupations the scores are averaged. Roles matched to the same occupation share a score, which is why some rows are identical. Roles with no defensible match were excluded, not forced.

  8. 8.Precision

    O*NET publishes a standard error for each rating. For the three work-context measures it does not, so we computed it from the published answer distribution (the spread of answers divided by the square root of the respondents). The three components are treated as independent and the choice of the strongest context is ignored, which are approximations. The result is a 95% interval for each score.

Roles we did not score

AI Content Creator
No O*NET occupation matches the role.
Costume Designer
The nearest occupations (fashion designers, costume attendants) describe different work.
Customer Success Manager
No O*NET occupation matches; sales managers and customer service managers differ.
Devops Engineer
No O*NET occupation matches; software developers and systems administrators each cover only part of the role.
Financial Analyst
O*NET 31.0 has no ratings for the matching occupation (Financial and Investment Analysts). Nothing was imputed.
Investment Banker
The nearest O*NET occupation (Financial and Investment Analysts) has no ratings in release 31.0. Nothing was imputed.
Risk Manager
The matching O*NET occupation (Financial Risk Specialists) has no ratings in release 31.0. Nothing was imputed.
Engineer
Too general. The role covers many O*NET engineering occupations.
Location Manager
The nearest O*NET occupation (entertainment and recreation managers) is a different job.
Product Manager
O*NET 31.0 has no product manager occupation. Marketing managers and project management specialists differ too much to stand in.
Production Assistant
No single O*NET occupation matches.
Production Coordinator
No single O*NET occupation matches.
Prompt Engineer
No O*NET occupation matches the role.
Recent Graduate
Not an occupation.
Scientist
Too general. The role covers many O*NET science occupations.
Script Supervisor
No O*NET occupation matches the role.
Story Producer
No single O*NET occupation matches.
Studio Executive
Could be chief executives, producers or operations managers; no single match.
VFX Supervisor
The nearest occupation (special effects artists) is a non-supervisory role.

What this study cannot tell you

  • It describes the job, not the interview

    O*NET asks how much the work involves speaking. It does not ask how interviews for that work are run, so a high score does not mean a harder interview.

  • Occupations count once each

    The baseline is not weighted by employment. A percentile describes where a role sits among the 905 occupations in the database, not among workers.

  • The role-to-occupation match is a judgment

    24 of 70 roles are matched to the closest occupation, not an exact one. The match and its O*NET title are in the dataset, so you can disagree with any of them.

  • The ratings are survey estimates

    They come from small samples of workers and analysts, collected over several years and a few look surprising for the job. That is why each score carries an interval.

  • The definition and the 0.5 threshold are choices

    Another definition would reorder roles that sit close together and another threshold would move roles between standout groups and none. The threshold was set before the results. The index definition was changed once, after the first results, for the reason given in the method; the earlier average is published for comparison.

  • Face-to-face discussion barely separates occupations

    99.8% of occupations score 50 or more on it, so it adds little to the overall score and a role's standout on it is a small absolute difference.

  • The scope is the roles on this site

    70 of the 89 roles are covered. The 19 excluded roles are listed below with the reason.

Data and corrections

Use the data

Download the dataset (CSV), licensed CC BY 4.0. It has every score, interval, percentile, level, standout kind and the O*NET occupation each role was matched to.

Report a correction

If you think a role is matched to the wrong occupation, write to support@zoevera.com. Corrections are applied to the dataset and listed here with the date.

Corrections so far: none.

Cite this study

ZoeVera (2026). How much speaking does each job demand? November 1, 2026. https://prepare.zoevera.com/role-speaking-index

This page includes information from the O*NET 31.0 Database by the U.S. Department of Labor, Employment and Training Administration (USDOL/ETA), used under the CC BY 4.0 license. O*NET® is a trademark of USDOL/ETA. ZoeVera has modified all or some of this information. USDOL/ETA has not approved, endorsed, or tested these modifications.

Practice now