Bullets that survive a phone call
Improving my CV won’t help
Do you believe that employers are focussing just on your experience and the content of your CV - not the quality of its language?
Put a number on it. Turn duties into achievements. Never write “responsible for”. Does any of this matter?
The most repeated advice in CV writing, we recommend it ourselves, but nobody has ever measured whether it works. No experiment has taken a real CV, held the facts constant, varied only the phrasing, and counted the callbacks. Not for quantification, not for action verbs, not for the number of bullets under each postition held.
The one study worth reading
Wiles, Munyikwa and Horton ran an experiment on the online labour market, published in Management Science in 2025. Nearly half a million new jobseekers; half were offered algorithmic help with their profile writing, half were not. The help covered spelling, grammar, punctuation, word usage, tone and style — it did not change any underlying facts about anyone.
The assisted group were hired about 8% more often, and earned about 8% more. Employers rated them no worse once working.
The authors’ explanation is the interesting part. They call it the clarity view: better writing does not signal a better candidate, it helps an employer work out what you can actually do. The writing was not more impressive. It was more legible. It does not matter if you are amazing, if noone knows it.
Numbers with weight
Would your previous manager agree with the number?
If a number would survive that phone call, use it. If it would not, use a fact that would. Someone is deciding whether to spend forty minutes with you, and every line they cannot verify makes that decision harder.
Numbers where there are facts: “Grew regional revenue 23% by renegotiating supplier terms” is excellent - if it is true and it was yours. Where it is not, the alternative is not vagueness — it is a different kind of specific. What happens when the interviewer asks you to elaborate on how you achieved that?
Does anyone care about my contribution at work?
Most work is not measured. Nursing, teaching, administration, government, care, trades, informal trade, small business. Our industry focusses on roles that look good on a dashboard and leave everyone else in the cold.
Recruiters in the health, education sectors and the public service have read “increased efficiency by 30%” many times, and it persuades nobody who has done the job.
Reach for the facts that are available: scale, frequency, span, duration, constraint, and what did not go wrong.
Caseload of roughly 80 households across three towns, 2019–2022.
About 40 walk-in queries a day, closed same-day except the grant applications.
Sole administrator for six clinicians — ordering, filing and monthly DHIS captures.
Full examination cycle for three grades, 120 learners, no scripts misplaced.
Monthly stats submitted on time for 24 consecutive months.
Ran the branch for eight months with no locum, on paper registers.
Four years on the plant with no time lost to injury in my department.
Your company might not have tracked any of those KPIs. Several are clearly achievements — sole, consecutive, none lost. The last two matter especially for maintenance, safety, compliance and IT work, where the value is obvious. What did not happen/ go wrong is also important.
The formula:
[what you did] [to what] [at what scale] [for whom or how often] [under what constraint]
The constraint is the part people leave out, and it is frequently where the jobs’ value actually was. “Managed the filing” has no impact. “Managed the filing for a clinic with no intern, on paper” - now that’s value managers understand.
Number nonsense
Quantification came out of sales, and became popular in American technology hiring, where the work logs throws off metrics as a byproduct. Applied everywhere, it fails predictably:
Fake precision. “Improved team morale by 20%.” How did you instrument that? Ask someone to test your CV, Does it sound… suspect.
Attribution. You were one of twelve on a project, and you said: “I grew revenue 40%” invites the question: How do you know your impact on that figure?
Confidentiality. Banking, revenue, health, defence. Some figures should not be in a document you email to a stranger.
Metric capture. Whatever sat on a dashboard gets a bullet and whatever did not disappears, which quietly penalises care, teaching, administration and public service.
The wrong reader. In a lot of the South African public service, a panel scores your application against the advert and the job spec. A polished corporate bullet that does not use the language of the circular can lose to a plain duty statement that does. Applying on a Z83 for a provincial or national post? Matching the spec could be critical if you are up against a lot of candidates and their measure is a checklist.
Rands have a date on them
A rand figure from 2015 quoted flat in 2026 misleads, because cumulative inflation over that period runs to roughly two-thirds. It is not a currency collapse, but “R4.2 million” meant materially more then than the same words mean now.
Three honest options: date it — “R4.2m procurement book (2015)”; prefer a unit that does not age as easily, such as headcount, volume, caseload or percentage; or convert “in todays money” and explain that you have.
Do not convert historical rands to dollars on a local CV. It reads as someone dressing up a number.
Titles that get misread by international readers
A misread title wrecks everything underneath it, because it sets the reader’s expectation of scale before they reach a single bullet. Several common South African ones land wrong.
| Title | The problem |
|---|---|
| Deputy Director, Assistant Director | Public-service middle management. A foreign or private-sector reader hears “director” and over-reads it substantially |
| Director, Chief Director | Genuinely senior in the state — but these are senior management grades, not company board seats, and the distinction is invisible abroad |
| CEO or General Manager | Meaningless without the size of the organisation. Chief executive of a six-person NPC needs the six |
| Officer | Junior in one sector, professional in another. Empty without the function |
| Consultant | Unregulated, and on a CV a stretch of it can read as unemployment |
| Administrator, Clerk, Coordinator | Read as junior when the person was running an SME’s entire back office |
| Team leader, Supervisor, Foreman | Meaningless without the span of control |
| Educator, Lecturer | School, TVET and university are different labour markets |
| Candidate attorney, Professional nurse, Artisan | Keep the statutory title exactly, then add a line of scope |
The fix is the same every time: title, then function, then scope — budget, headcount, geography, caseload. And where you hold a registration, put it beside the role rather than at the foot of the page. SANC, HPCSA, ECSA, SACE, the Legal Practice Council: those are worth more than any verb.
The receipts
If you want to know why this article will not promise you a percentage improvement, it is because the genre is full of them and they do not survive inspection.
A figure of “40% more interviews” from quantifying your bullets circulates across career sites with a different parent each time — an unnamed analysis on one, a recruitment platform’s own observational data on another. “3.2× more callbacks” is credited to a “Resume Research Institute” with no discoverable research behind it. The same numbers with different parents is recirculation dressed as replication.
The formula itself — accomplished X, as measured by Y, by doing Z — is Laszlo Bock’s, from a LinkedIn post in 2014 when he ran People Operations at Google. It is one experienced executive’s personal recipe rather than a published Google standard, and the “reviewed 20,000 CVs” line attached to it is his own lifetime estimate. The underlying idea is older than him and comes out of sales.
Beyond the Wiles study, what exists is stated preference: surveys in which hiring managers say they would rather see achievements than duties. Worth knowing, and it is self-report, which is a poor guide to what people do with a stack of applications. There is no rigorous academic evidence for action verbs. The “three to six bullets per role” rule is coaching extrapolation from small commercial eye-tracking work that was about where people’s eyes land, not how many lines to write.
People who quantify tend to have held jobs that were quantified, and stronger candidates write better CVs, in general.
So match the advert’s language, write bullets a sceptical person in that sector would recognise as true, and put in a number only where it is a fact.
A CV is just the gateway to an interview - write it that way. Prepare for it to be challenged.
We do not guarantee that following our advice will improve your chances of employment.