Original research · blind format experiment
Best Resume Format: A Blind Test With 20 Recruiters
We hid six resumes in a realistic pile of applicants and asked 20 recruiters to screen it, without telling them any of the resumes were ours. All six candidates were equally qualified for the job — same requirements met, same range of years, same size of budget, same city. Only the way each resume presented that was different. The same qualifications were shortlisted 12 times out of 13 in one format, and 3 times in another.
I ran this experiment myself. I built the pile and the screening tool recruiters worked in, moderated every session, and analyzed the results. Recruiters were recruited through a research panel under a neutral cover, screened against the bar set out in the methodology, and paid for their time. Quotes are verbatim, with recruiters and their employers de-identified.
TL;DR
- Six candidates hidden in the pile were equally qualified for the job. Only the way their resume presented that changed.
- The same qualifications were shortlisted 12 times out of 13 in one format — and 3 times out of 13 in another.
- The candidate with the strongest numbers on paper — the biggest pipeline, the biggest budget — finished last. His resume was a two-column template that scattered them.
- What separated the top from the bottom wasn't looks. It was how quickly a recruiter could find the things the job description asked for.
- These weren't a random panel. Every recruiter was individually screened and paid — most field 200+ applicants per role, and all review resumes by hand, daily or weekly.
- We ran 20 sessions and count the 13 clean ones. Seven are set aside — five for a setup mistake of ours, two where the recruiter screened for a different job than the one described. It's a small blind experiment, not a poll.
20 recruiters · two rounds · June 2026 · blind, within-subject design.
The result
How the resume was written and laid out decided who got shortlisted
Every one of these six candidates met all six requirements on the job description, at a similar level: 5 to 7 years, $2.4M–$3.1M in pipeline, a $500K–$750K budget, HubSpot, a sales team, the same city. A recruiter reading any of them carefully should have concluded “qualified.” How often each was actually shortlisted, across the 13 sessions we count:
Candidate Aours
5 yrs · $2.8M pipeline · $600K budget
12 / 13Shortlisted in 12 of the 13 recruiter sessions we counted. (Seven of the 20 are set aside — see the methodology.)
Qualifications gathered at the top of page one
Candidate B
6 yrs · $2.6M pipeline · $650K budget
11 / 13Shortlisted in 11 of the 13 recruiter sessions we counted. (Seven of the 20 are set aside — see the methodology.)
Self-written, qualifications buried in the bullets
Candidate C
7 yrs · $3.0M pipeline · $700K budget
10 / 13Shortlisted in 10 of the 13 recruiter sessions we counted. (Seven of the 20 are set aside — see the methodology.)
Over-stuffed, three pages
Candidate D
5 yrs · $2.5M pipeline · $550K budget
7 / 13Shortlisted in 7 of the 13 recruiter sessions we counted. (Seven of the 20 are set aside — see the methodology.)
Clean, conventional Word resume
Candidate E
5 yrs · $2.4M pipeline · $500K budget
6 / 13Shortlisted in 6 of the 13 recruiter sessions we counted. (Seven of the 20 are set aside — see the methodology.)
Popular resume-builder template
Candidate F
6 yrs · $3.1M pipeline · $750K budget
3 / 13Shortlisted in 3 of the 13 recruiter sessions we counted. (Seven of the 20 are set aside — see the methodology.)
Two-column builder template with a skills sidebar
The part we did not expect
Candidate F had the best numbers of anyone in the experiment — the biggest pipeline, the biggest budget — and was shortlisted the least of all six. Presentation didn’t just shuffle the ranking. It inverted it.
We built him to be strong. His resume is a polished two-column template with a skills sidebar and a generic professional summary — the qualifications are all in there, scattered. Ten of the thirteen recruiters never shortlisted him. We should be honest that this is the largest effect in the study, and it has nothing to do with our product. It is a finding about what a template does to a good candidate.
Why 13, not 20? We interviewed 20 recruiters and count 13. Seven were set aside under the study’s own rules — not for anything to do with our resume. The methodology explains each one.
The artifacts
The six equal candidates
These six resumes were hidden in a pile of 22. Every one meets all six requirements on the job description, at a similar level — only the presentation differs. That is what makes them the test: same qualifications, six ways of showing them. The recruiters screened all 22 in a custom tool, in random order; the other sixteen were realistic borderline and non-fit applicants.
Fictional candidates built for the experiment. The preview is page one — tap any resume to open the full document. See all 22 resumes in the pile →
The observations
What recruiters actually did
The chart shows the outcomes moved; the recordings show why. Across twenty recruiters thinking out loud while they screened, the same behaviours kept surfacing. This is how they actually read a resume — and it is what the advice below is built on.
The first pass is a matter of seconds
Recruiters told us the opening read is fast — just enough to decide whether a resume earns a proper look. Whatever is going to save you has to be legible at that speed, and near the top of the page.
“The facts say that we look at a resume for seven seconds, and that’s how long it typically is… I’m literally just looking at experience.”They fit-match against the job description, top-down
The most common move we saw: the recruiter arrives holding the job’s requirements in their head and looks for a place to tick them off. A short block at the top that answers each requirement lets them do it in seconds — then they drop into the experience to check it’s real.
“It was nice having it up front, because then I can be like: check, check, check, check, check. And then I’ll go into the rest of the resume.”“It’s literally going down our qualifications on the left-hand side and telling me exactly why she’s qualified.”But about a third skip the top and read experience first
Not everyone. Roughly a third go straight to the work history and treat a block at the top as marketing. For them it adds nothing — the experience underneath has to carry the resume on its own.
“I skip it right away. I’ll get to the experience, education, location, see if that’s all right.”A number is believed only when it has context
Recruiters reach for numbers to break ties, but a bare percentage gets discounted — they can’t tell whether it describes something big or trivial. A figure with a starting point or a scale reads as credible.
““I achieved 50% growth from… 5 to $10 million in revenue”… there’s some context to it… I feel it’s a more believable bullet than “strategically increase sales by 50%.” Well, from a dollar to $50… I don’t have a context to… quantifying that.”Presentation re-ordered candidates who were equally qualified
All six candidates met every requirement, yet their shortlist rates ran from 12 of 13 down to 3. The strongest on paper — the biggest pipeline and the biggest budget — finished last, in a two-column template that scattered his qualifications. What moved the outcome was how quickly each resume could be read.
“I actually even like the fact that the person bolded it for me.”Length is judged fast — and a long resume can give away your age
Past two pages, recruiters lose patience — one rejects outright at five or six. And a long history quietly signals age, which recruiters admitted can bias them.
“The longer your resume is, the further you’re going back to the ’90s… it can create unconscious bias.”They check the commute, look you up, and Google an unknown employer
Around the read, recruiters run three fast checks. Location comes first — but it’s really about the commute: if the role needs time in the office, they check whether you live close enough to get there, and because they often don’t know the area, they pull up a map to see if your city is within reach. They open your LinkedIn to confirm the resume matches. And they Google any employer they don’t recognize. Each is a place a resume can save them time, or cost you.
“I don’t live in Texas so I would need to do some research on that… within a commutable or drivable or reasonable distance.”“Before they go to my hiring managers, then I’m going to double-check with LinkedIn and make sure everything cross-checks…”“Unless you work for an instantly recognizable employer, please indicate the industry of the company… I spend an inordinate amount of time Googling all these companies.”The format
Our resume format, piece by piece
Candidate A — our format — was shortlisted more often than any of the other five, with the same six qualifications the rest of the pile had. What sets it apart is how those qualifications are arranged: surfaced at the top and tied to the job, instead of buried in the experience. Here is how it is built.
Tap to open the full resume.
It is Candidate A on the chart — the same six qualifications every candidate in the pile had, presented so a recruiter finds them fast instead of hunting.
- 1
A one-line header, with location first and in bold
Name, then a single contact line — location, email, phone, LinkedIn. Location leads and it's bold, because it's the first thing recruiters check — whether you're close enough for the role — so the answer is right there before they read anything else. The LinkedIn link is there too — recruiters look candidates up, and a missing profile reads as a small red flag.
- 2
A short “Key Qualifications” block, capped and bolded
Up to about seven lines at the very top that restate the job description's requirements in its own language — with the key terms in bold so a scanning recruiter lands on them in a second. It's deliberately short: too many lines and none stand out. This is the block recruiters fit-match against before they read a word of the experience — "it tells me before I even get to her actual experience that she's already qualified," one said.
- 3
Every bullet is an outcome with a number — and a baseline
Not just "results." Each bullet pairs the impact with a figure and something to measure it against — "lifted MQL→SQL conversion from 18% to 27%," "$2.8M in qualified pipeline in FY25," "a $600K annual paid-media budget… at 4.2x ROAS." A number with a starting point or a scale reads as credible; a bare percentage recruiters wave away.
- 4
The most recent role gets the detail; older roles shrink
Recruiters read your latest job closely and skim the rest, so the depth goes where they look. Here the current role runs six bullets, the one before it three, the one before that two.
- 5
One column, newest first, only what’s relevant
A single column in reverse-chronological order — nothing tucked into a sidebar or a second column, so a recruiter reading top-to-bottom never has to hunt. And only the experience that speaks to this specific job earns a place; the rest is cut.
For your resume
What to put on your resume
Everything the study points to, as a checklist — the format we tested, plus the improvements recruiters told us they wanted. Every item comes from what the recruiters actually said.
- 1
Lead with a tailored key-qualifications block
A short list at the very top that names the job’s requirements and answers each one — in the job’s own words, key terms in bold, about seven lines. It’s the first thing most recruiters read: “it tells me before I even get to her actual experience that she’s already qualified,” one said. Tailor it to the specific role; a generic summary gets skipped.
- 2
Put your total years of experience in plain sight
Recruiters asked for this constantly, and stopped to count dates when it was missing. “Just tell me the years of experience you have,” one said. “Don’t tell me several. Does that mean seven? Does that mean four? Does that mean three?” Say it near the top.
- 3
Put your location at the top — and say if you’ll relocate
Location is the fastest cut recruiters make. State your city — and if you’re open to moving, say so on the resume, because recruiters can’t tell otherwise and tend to assume you aren’t. One told us it settles the question on the spot: “I’ve seen resumes where it says open to relocation… great, answers my question, don’t worry, continue on.”
- 4
Use a job title people recognize
A niche internal title can read as unfamiliar. “If you search just on Google, demand generation manager, it’s not a very standard job title,” one recruiter said. “If she would have called herself a marketing manager it would have been better.” Translate it into what the market actually searches for.
- 5
Make every bullet an outcome with a number — and a baseline
A bullet like “$5M to $10M” is believed; one like “increased sales 50%” isn’t — it could be a dollar to fifty. Pair each figure with a starting point, a budget, or a scale, and write numerals, not spelled-out words.
- 6
Put the most detail on your most recent role
Recruiters read your latest job closely and skim the rest, so put most of your bullets there — about six on the current role, fewer on each older one. Cut the oldest jobs down to a line or two; nobody reads them closely anyway.
- 7
Name your employer’s industry if it isn’t famous — and link it
Recruiters lose real time looking up companies they don’t recognize. Add a clause after the name (“a B2B logistics SaaS”) or hyperlink the company site or LinkedIn, so no one has to leave your resume to find out who you worked for.
- 8
Keep it consistent with your LinkedIn
Recruiters open your profile to check the resume matches — one keeps it “next to the resume to see if there are inconsistencies between the bullets and their LinkedIn profile.” Line the two up, and include the link.
- 9
One column, newest first, no photo, only what’s relevant
A single column in reverse-chronological order, nothing tucked into a sidebar; no photo, at least in the US; and only what speaks to this job. Length isn’t a fixed rule — it scales with seniority, and a one-pager can look thin for a senior candidate. What decides it is whether a recruiter can find the match fast, so keep every page relevant and cut anything that isn’t. As one put it: “The size of the resume itself isn’t necessarily a factor. It’s how much it correlates to the amount of experience of the position that I’m comparing it to.”
- 10
Proofread — a single typo is an instant reject
One misspelling and you’re out. “You spelled experience wrong… I would just automatically disqualify…” one recruiter said. Read it twice, and have someone else read it too.
The interviews
Q&A with the recruiters
Watching recruiters screen is one kind of evidence; asking them point-blank is another. After the blind pass, the sessions ended with the same bank of direct questions — tailoring, AI, cover letters, what they wish candidates knew. The answers are verbatim. (They include the two recruiters whose screening sessions the chart excludes: they screened a different job than the one described, which voids their shortlist data — not their answers.)
“I wish every candidate knew that…”
We ended every interview with the same fill-in-the-blank. All eleven answers are here — the first five, then the rest behind “More answers.”
“That real people look at their resumes.”“Simplicity is king, meaning too much information is not always good… a resume gets you a conversation. It doesn’t need to tell your whole story.”“Recruiting’s like matchmaking. And what we’re trying to do is figure out how your strengths and weaknesses fill into the strengths and weaknesses of the teams we’re building.”“They need to sell themselves as if they’re selling a product, because that’s going to get them more of the job than the person who just puts the basic stuff on their resume.”“Right now, I wish every candidate knew that using AI in your interviews was not a good idea.”She means AI on the live interview call, not on the resume — the helper tools that listen to the interviewer’s questions and feed the candidate suggested answers in real time. She says recruiters can tell, and that candidates caught doing it can be banned from applying again.
More answers (6)
“Human eyes are looking at their resume and not AI.”“Recruiters are people at the end of the day and we’re very limited in capacity and what we can do in eight hours of time… in a workday… it’s a people business and we can’t control everything at the end of the day.”“Unless you work for an instantly recognizable employer, please indicate the industry of the company you’re currently working for. Because I spend an inordinate amount of time Googling all these companies.”“Be prepared and do their research ahead of time for the… screening call… I don’t think people prepare enough and understand what’s going on, or they don’t even know the job description well enough… and we may overlook candidates because there’s a speed level. We have to get through all the people.”The backdrop is the auto-apply wave: tools now blast a resume at hundreds of openings on the candidate’s behalf, so recruiters increasingly reach people who arrive at the screening call not knowing the job — or who never answer at all. As another recruiter put it about contacting them: they’re “always of course the ones that don’t really answer back, because they didn’t really apply to the position in the first place.”
“I wish that every candidate knew what our company was and wanted to work for us… If they’ve done some research up front, that just makes for a better process and success.”“How to present their resume well.”“Does an AI-written resume bother you — or not, as long as it’s accurate?”
We asked this directly, and nobody said yes — one of them uses AI on his own resume. The line every answer drew is accuracy, not authorship: use the tool, but the information has to be true, and you have to be able to back it up on the call.
“Accurate, it’s totally fine. I know I used AI to update and write my resume. I’m not the greatest writer.”“Well, I think everyone is using it. I don’t think there’s people that are not using it in some form or fashion… how can you not use it now?”“Just pulling a job description and pulling your work experience into AI and having it make a resume for you essentially to beat an AI filter on an ATS, I understand that, but you need to put your humanness in it.”More answers (5)
“Does it bother me? As long as all the information’s there? No. But it also makes me think twice and say, like, why? You clearly just left it bare bones… you need to spruce it up a little bit.”“It doesn’t… that comes down to the interview. When I actually talk to them, they don’t back it up to what’s on there. That will show it.”“As long as it’s accurate, it’s okay. But I think that you run the risk then of coming across polished and able to speak well. And then if you get on the phone and you’re not at all that person, it really doesn’t align.”“The only thing I don’t like is when people use AI software to… blanket apply for positions… It’s okay when it’s somebody that’s using it to kind of tailor their resume to the position when it’s already a fit.”“It doesn’t bother me, no.”“Can you tell when AI wrote the resume?”
So this is not “hide that you used AI” — it’s what the one-prompt version looks like to a recruiter. A generic skeleton with your details poured in has a cadence they recognize instantly. The fix is effort, not avoidance: tailor it, put your own voice back in, and read it once yourself before sending.
“It has this really weird cadence… It’s the same cadence over and over again. It’s almost like it has a very particular… skeleton, and then your information is pulled into that skeleton.”“When you copy and paste from ChatGPT, you can just start seeing it. Even the dashes are in there… If they haven’t used it in a way that the technology has learned their style of talking, then it sounds very robotic and it’s a dead giveaway.”“I think I ignore the fluff and I think AI pumps up the fluff. So I haven’t noticed it because I think I ignore it.”More answers (3)
“A lot of it’s the spacing… word usage, word play, spacing… AI tends to throw the little dashes in, for whatever reason. Humans don’t typically use them too often now. And those are things that I look at.”“Everything’s check, check, check. It’s almost too check, check, check that I’m thinking now… did AI write this?”For honesty: she said this while reading one of the pile’s deliberately plain control resumes — no AI involved — and she shortlisted the candidate anyway. The “too perfect” alarm can false-positive; what decides is whether the experience holds up.
“Most of my candidates do not even know how to use an AI… you cannot just use a fake resume and just apply… nobody is going to trust you.”“Is mirroring the job description helpful — or gaming?”
This one is about tailoring: writing your resume in the job description’s own language, so the skills and terms the recruiter is hunting for appear in your resume in the same words. Recruiters leaned “helpful — it speeds up the fit-check,” with one hard line: copy whole sentences word for word and it flips into a red flag.
“It gets my attention enough that I want to talk with you. It’s kind of like if you’re dating… it gets you to the first date, but that doesn’t mean we’re going to hit it off.”“I thought it was helpful. I thought it was like boom, boom, boom. Okay, cool. I see it… the nitty-gritty is gonna come down when you do that phone screen.”“If it was like exactly how the sentence was, then, yeah, I’d have a little red flag for sure.”More answers (4)
“I would somewhat question if it’s AI. I would still warrant a conversation with the candidate if I can… can they speak on it? If they can speak on it and they can speak on their experience, that makes more significance to me than anything else.”“I think it’s yes and no… if someone’s doing it, at least they’re putting effort in to looking at the job description… I think it’s probably more helpful than it’s not helpful. But… if they can’t do something, they’re gonna get exposed.”“For me it’ll open the door, right, because it sounds exactly like the job description. But the interaction with the candidate is more important… if they can’t back up what they’re saying on there in a conversation with me, then… it was a waste.”“I would say it’s helpful. I’m not going to 100 [percent] rely on it, especially if I can tell it’s been done with AI and it’s just basically been copied and pasted. It has to also give specifics.”“Do you still read cover letters?”
Near-unanimous: mostly skipped. The strongest version came from a recruiter who couldn’t even locate them in his own system. Keep in mind who’s answering, though — recruiters screening high-volume piles. At a small company where every application actually gets read, or where a founder or hiring manager does the screening themselves, a cover letter is a different story.
“I honestly wouldn’t know where to find a cover letter in our ATS… I don’t necessarily see the value in them anymore, and I don’t have the time to read all of them.”“Very rarely do they make a significant difference… I would say 8, 9 out of 10 times, they’re probably not as important as having a resume that’s well written.”“I go straight to the resume eight times out of 10… unless it’s captivating and captures my attention, to me, that’s fluff. Let me look at your resume.”More answers (2)
“I never read them. I always told people to write them because the hiring managers like them… I don’t even read them unless I notice somebody has a big gap in their resume… I just look to see if there’s an explanation for any gaps.”“I haven’t looked at a cover letter forever.”“Does hiding white-text keywords in the resume work?”
Don’t do this. Most manual reviewers never even see it — but where it is caught, the consequence can be a formal one: one recruiter at a Fortune-100 tech company described a process that ends with you banned from applying to the company again.
“What I have seen in the past is where people will put keywords of job descriptions and then they’ll put it in white ink so it’ll pop up… That’s as fake as I have seen… It’s an automatic not a fit… You red flag it. It goes to our HR department… put them in our system as a… do not reach out, close them down from applying to other roles. We have a whole group that handles that.”“If it’s in white, I don’t see them. I usually manually review resumes…”“I’ve never highlighted all the text to see if something was highlighted. No.”“Maybe it gets you on the map, maybe it doesn’t. But it only helps you if you are the right fit… to me it’s kind of superfluous.”More answers (2)
“It doesn’t matter to me because I’m going to as it is shortlist with my own knowledge… I will open up that profile… I will go and check their LinkedIn… Then I will have a call with them personally to check what exactly is their experience.”“I’ve actually never seen that. No. Or at least I’ve never noticed that myself.”“How much of the final decision is the resume?”
Every answer landed between roughly 10 and 30 percent — with one exception for deeply technical roles. The resume’s job is to open the door; the screening call decides. That’s the honest size of what this page can help with.
“The resume is probably going to be… 15 to 20% at the most… 80 to 85 of my determination to move a candidate forward… is going to be my screen.”“I think a resume is a tool to get your foot in the door and get you noticed… How you do in those interviews, I think, is most important.”“The resume is really important to get them to the phone screen and then it’s all on them, not necessarily the resume.”More answers (2)
“The resume is kind of the gateway to the other ones… maybe it’s 10, 15, 20%. It’s not a tremendous amount because it’s going to matter how you do in the interviews.”“If I’m looking for a lab tech that’s got to have some very specific technical skills, their resume is going to be probably 70 to 80% of whether they’re moving forward and getting hired. If it’s a business development rep… it could be really 20 to 30%… The more technical the position is, the more their resume counts.”How we ran it
Methodology
20 recruiters screened a realistic pile of applicants for a simulated job opening, in a custom ATS built for this study — without being told that any of the resumes were ours.
Each recruiter qualified through a screener, then joined a paid, hour-long video call. They read a job description — a marketing manager role at a deliberately forgettable mid-market software company in Austin, with six stated requirements — and then worked down a pile of resumes in a purpose-built applicant-tracking system, thinking aloud and setting a status on each one: shortlist, maybe, unqualified, not a fit. We stayed silent during the pass. The shortlist status is the outcome we report — what they actually clicked, not what they said about it afterward.
Who the recruiters were
This was not a random panel. Each recruiter was individually screened against the same high-volume bar as our recruiter-screening study, then paid for the hour. To qualify, every one of them had to be someone who screens real inbound piles for a living:
- 14 of the 20 field 200 or more applicants for a single opening, and nine field 500 or more. Several work at companies fielding 200,000+ applications a year.
- They review resumes daily or weekly — not occasionally — so their read on a format is current and practiced.
- They screen by hand: teams that fully automate, or invite straight to interview without opening a resume, were screened out, because a reaction to a format has to come from real screening.
- They hire for three or more roles at once, most of them more than ten.
- They're in-house recruiters, agency recruiters, or hiring managers who personally do the screening.
Every point above was a hard requirement, checked before anyone joined a call — the reason a sample this small is still worth reading.
What we varied
Six resumes in the pile were built to be equally qualified: all six requirements, 5–7 years, $2.4M–$3.1M in pipeline, a $500K–$750K budget, the same city. What differed was presentation — the layout, and how directly the resume answered the job description. We should be precise about that second part, because it matters: the comparison resumes were deliberately built untailored, with the qualifications present but scattered through the experience rather than surfaced. So the variable is presentation in the broad sense, not typography alone. All 22 resumes are in the appendix — open them and compare.
What we did to keep it honest
- Blind. Recruiters were recruited under a neutral cover story and never told which resume was ours, or that any of them was.
- Within-subject. Every recruiter saw all six candidates, so each person is their own control.
- Position was shuffled. Our resume appeared at different depths in the pile across sessions — as deep as sixteenth — so its result isn't an artifact of landing near the top.
- The pile felt deep. In the second round the system displayed a four-hundred-plus applicant count, so recruiters triaged the way they would under real volume.
Every session we excluded, and why
We ran 20 sessions and report 13. That is a big cut, so here is every exclusion and why. We also give the unscreened number, because two of the three exclusions are sessions where our resume was cut.
- 5 sessions, early in round one — a design error of ours. Our candidate was the only one who lived in the city on the job description, and location turned out to be a hard first gate. That confound ran in our favour, so we fixed it — every candidate moved to the same metro — and threw those sessions away.
- 2 sessions where the recruiter screened for a different job. One hunted for 9–15 years of experience on a role that called for 4–7 and shortlisted a salesperson. The other counted only years served under a “manager” title — the job description asked for 4–7 years of marketing experience, with no such condition — and preferred a generalist over the demand-generation specialist it called for. Both cut all six equally-qualified candidates, not just ours — that is the test we applied, and we applied it before looking at how ours did.
The unscreened round-two number is 8 of 11. Screened, it is 8 of 9. Pooled with the clean round-one sessions it is the 12 of 13 on the chart above. Either number works: the shape of the finding, and the collapse at the bottom of the chart, does not depend on the exclusions.
A second test we are not publishing yet
The experiment had a second arm: a deliberately weaker candidate, below the years bar, shown in our format and in a plain format. On the face of it, it worked — the plain twin was cut and ours was advanced in three sessions, and never the reverse. We are not publishing it as a finding, because we do not believe it. Reading the transcripts, the plain twin's summary opens with "around 3.5 years" and our version states no total-years figure at all — and in all three cases the recruiter arrived at a *different number of years* for two candidates who were supposed to be identical. That is a result about disclosure, not about format. A conclusive version would state the years identically on both resumes; we have not re-run it.
Rigor and limits. This is a small blind experiment, not a poll. Thirteen clean sessions cannot be generalized to a population — the margin of error at this scale is around twenty percentage points, and one flipped decision moves a row. Our resume was shortlisted most often, but only 12 to 11 ahead of the next-best — too close, at this size, to claim it beats a strong resume. The clear, large effect is the gap to the template formats (3 of 13). “Equally qualified” is a design assumption we made, not a measured fact; recruiters’ bars varied. Not every recruiter we spoke to regularly hires for marketing roles, though we chose a deliberately common role so most could screen it comfortably. Read it as a demonstration that presentation can change the outcome for an equally-qualified candidate — which it does, clearly — and not as a measurement of how much, or how often.
20 recruiters · two rounds · blind, within-subject · June 2026 · 412 logged resume decisions.
This study follows “How Recruiters Actually Screen When Hundreds Apply”
This experiment only matters because of what we found first: at a high-volume opening, most resumes are never read by anyone, and the ones that are get seconds. We interviewed 43 recruiters about how that actually works — what makes them stop, and what makes them skip. Read the study →