Does an ATS score tell you whether your CV will get you interviews?
Rarely. Jobscan recommends a match rate of 75% on one page and 80% on another, and a checker vendor writes that the same CV run through three tools often comes back 15 to 20 points apart. Greenhouse says its AI doesn't score or rank applications, and wherever an employer's system does rank people, an external checker can't see it. Real parsing trouble comes from columns, tables, text boxes and contact details in headers. A field experiment with nearly half a million jobseekers found that people whose CVs became clearer, with fewer errors, were hired 8% more often at 10% higher wages, with no keywords or scores involved.
Dear MoreThanCareer.de community,
Welcome back to the newsletter that most CV scanners would grade at 41%, for containing too few keywords and a dangerously high concentration of opinions.
Save this one now. You'll need it the evening a friend sends you a screenshot of a red dial reading "ATS score: 58%, weak match", in the tone usually reserved for a biopsy result. You'll need it again the morning a stranger in your LinkedIn inbox offers to resuscitate that dial for €349 by Friday, payment up front, because apparently the patient is crashing.
A confession before the first body is wheeled in. This article is partly an advert, and I'd rather confess now than be exhumed later. I rewrite CVs for a living. Everything I do is at www.morethancareer.de: career coaching with job search strategy, interview preparation and salary negotiation training with tailored scripts, reverse recruitment, and CV and LinkedIn rewrites and rebranding. So read this the way you'd read a tobacco company's study on the respiratory benefits of menthol. Then check my footnotes. By the standards of this trade they're unusually well behaved, since every one of them has a return address.
A note for readers outside the German-speaking world. Most examples start in DACH, because that's where I work most. Nearly everything here applies equally if you write "CV" in Manchester, "CV zgodne z ATS" in Kraków or "résumé" in Chicago. I'll show you why halfway down, with a French law that died without ever being born.
The ATS score: a pulse reading taken from a mannequin
We begin with the most successful product in the CV business. It's a percentage, and it's thriving.
You upload your CV and paste in a job advert, and a dial swings to 62%. Red. You add "stakeholder management" four more times and it rises to 81%. Green. Your qualifications didn't change in those ninety seconds. Your blood pressure did, and nobody is scoring that.
Jobscan more or less invented this category. On its tutorial page it recommends a match rate of at least 75%1. On a blog page it recommends 80%2. Same company, same metric, five points of disagreement with itself, which in a GMP lab would earn somebody a CAPA and a very quiet lunch. Its homepage then warns that a resume can be over-optimised, and concedes that scoring above 75% may not be possible without stuffing in keywords. The same page promises three times more interviews.3 It's a diet pill with a leaflet warning that, taken as directed, you may disappear.
Then ask for a second opinion. A company that builds one of these checkers writes that the same CV and the same advert, run through three tools, often come back 15 to 20 points apart, and describes that as normal rather than a bug.4 Enhancv tested its own sector's templates. Some Microsoft Word templates scored 100%, one scored 36%, and Canva designs were scattered anywhere between 52% and 92%.5 Three pathologists, one body, three causes of death, and all three invoices payable within fourteen days. In German we call this Kaffeesatzleserei (reading coffee grounds, our word for elaborate guesswork in a lab coat). This version comes with a progress bar and auto-renewal.
Then there's the employer's side, where the joke turns properly dark. Greenhouse, one of the larger applicant tracking systems, tells candidates on its own blog that its AI doesn't score or rank applications, and doesn't decide who moves forward.6 A newer matching feature does sort applicants so that stronger matches get reviewed first, according to Jobscan's own guide, but it doesn't advance or reject anybody.7 Wherever an employer's system does rank people, the employer configured it and keeps it inside its own walls, where an external checker can't see it. So there are people paying €49.95 a month to raise a score the employer's software isn't computing. It's like laying flowers every Sunday on a grave that turns out to be somebody else's.
German engineers have a saying: Wer misst, misst Mist (whoever measures, measures rubbish). The point is that a measurement is only as good as its instrument. The ATS score is Scheingenauigkeit (spurious precision) in its purest form. It puts a decimal place on a guess, a subscription on the decimal place, and a cancellation button three menus deep behind the subscription.
ATS-konform (ATS-compliant): the German word for one column
In the German-speaking market the same fear has a sturdier name: der ATS-konforme Lebenslauf (the ATS-compliant CV, meaning one formatted so that applicant tracking software can read it). It sounds like something the TÜV certifies (the German inspection body that decides whether your car may stay on the road), and providers sell it with the gravity of a heart valve.
Here's what an applicant tracking system actually struggles with, according to the vendor's own help centre. Greenhouse lists the causes of a failed parse: files over 2.5 MB, graphics, photos or word art, a CV uploaded as an image, tables, headers and footers, contact details in a header, footer or text box, columned layouts, sections without clear structure, company names without identifiers such as Inc. or Ltd, and shortened job titles such as "Sr. Account Exec" instead of "Senior Account Executive".8
That's the whole secret, the one people pay €300 to be told. Use one column. Put your name and phone number in the body of the page. Use ordinary headings and write your job title out in full. A competent person can fix all of it before their coffee goes cold, and a slow one before it grows a skin.
And the cause of death, when formatting goes wrong, is less dramatic than advertised. When a parse fails, Greenhouse says the CV is simply attached to the candidate's profile and someone enters the details by hand.8 That's the tragedy. There's no cremation and no robot pressing a red button. A recruiter types your phone number, sighs, and spills a little coffee on a keyboard that has seen worse. I've been that recruiter. Your family will get through it. The one real risk is that a mangled profile turns up less often when someone runs a keyword search across the database. That's worth ten minutes of your time. It isn't worth €300 of your money.
The German market then adds its own grave goods: the DIN-konformer Lebenslauf (a CV that complies with the German DIN standard). Some providers advertise CVs laid out "according to the current DIN norm". DIN 5008 is Germany's standard for writing and layout rules in business correspondence. Guides to the norm point out that it applies to the cover letter, that no standard format was ever set for the CV, and that following DIN standards is voluntary anyway.9 A DIN-compliant CV is about as meaningful as a CE-marked horoscope. The marking is genuine. It belongs to something else. In German that's Etikettenschwindel (label fraud: packaging that promises what the box doesn't contain).
Earlier this month I did the autopsy on the rest of the automation myth, including the knockout questions on application forms. That's the stage where an application can end automatically, triggered by an answer you gave and not by your CV.10 I won't reopen the coffin. The short version: the gate that actually closes is the form, and no dial measures the form. The dial is busy measuring your adjectives.
Same corpse, different passports
Before readers outside DACH close the tab with a smug smile: none of this is a German eccentricity. The software is the same product in Munich, Warsaw, Lyon, Manchester and Toronto. When I checked seventeen pharma and CRO careers portals for my last article, they ran on Workday, SAP SuccessFactors, Eightfold, Avature, SmartRecruiters and Phenom.10 A Workday parser doesn't develop a new personality when it crosses the Oder. What the Greenhouse help centre says about columns and headers applies to a CV written in Polish, French or Texan. Only the undertakers' vocabulary changes.
In Poland the product is called CV zgodne z ATS (an ATS-compliant CV), and the 75% myth has emigrated there in excellent health. One Polish provider tells readers that ATS systems automatically reject up to 75% of applications.11 Poland also has its own national talisman, the klauzula RODO (the GDPR consent clause that generations of Polish candidates have typed at the bottom of their CVs like a short prayer for the dead). The Polish digital rights foundation Panoptykon explains that when the data in a CV stays within the catalogue set out in the Labour Code, no such clause is needed to process it.12 That catalogue in Article 22¹ was rewritten with effect from 4th May 2019.13 Hundreds of thousands of CVs still end in a legal incantation addressed to nobody in particular. Nie mój cyrk, nie moje małpy (not my circus, not my monkeys), as the Poles say. The trouble is that the candidates keep buying tickets to the circus.
France offers the most elegant death of all. A 2006 law made anonymous CVs compulsory at companies with more than fifty employees. The implementing decree was never issued. The Conseil d'État, France's highest administrative court, ordered the government to issue it. In August 2015 parliament solved the problem by making the anonymous CV optional.14 France anonymised its own law. The CV survived. The obligation went into an unmarked grave.
In the UK, the Advertising Standards Authority upheld a complaint in December 2024 against an ad headed "Trainee Health & Safety Advisor". It promised a job guarantee within 20 miles of your location, and turned out to be a fee-paying training course.15 The job was real in exactly one sense: somebody was getting paid.
North American readers get the only unambiguously good news in this article, and it's Canadian. Since 1st January 2026, Ontario employers with 25 or more employees must state in publicly advertised job postings whether they use artificial intelligence to screen, assess or select applicants.16 Candidates in Toronto can now read in the advert itself what candidates in Frankfurt pay €49.95 a month to guess.
So here's the rule of thumb wherever you're reading. What's true of CVs on the German market is almost entirely true in Warsaw, Lyon, Manchester, Boston and Toronto. The labels differ. The layout rules come from the same handful of vendors. The fear is identical, and so, give or take the currency, is the invoice.
The one large experiment worth your evening
There is serious research on whether better CVs get people hired, and, refreshingly, it wasn't funded by anyone selling CVs.
Emma Wiles, Zanele Munyikwa and John Horton ran a field experiment on an online labour market with nearly half a million jobseekers. The treated group got non-generative algorithmic help with the writing on their CVs: grammar, spelling and readability, with nothing invented. They were hired 8% more often, at 10% higher wages, and the researchers found no sign that employers were less satisfied afterwards. The treated CVs had fewer errors and were easier to read.17
An 8% efficacy gain, a 10% higher dose of salary, and no rise in reported regret. Few products in our industry get a label like that without a boxed warning. The authors' reading is that clearer writing helps employers recognise ability that was there all along. That's the whole honest case for a CV rewrite, and what's missing from it is telling: there are no keywords, no scores and no dials. The effect came from clarity. The human on the other side understood the candidate faster, and hired them before they gave up and retrained as a sommelier.
One caveat, since this newsletter sells neither miracles nor holy water. The experiment ran on a freelance platform with mostly new entrants. It won't transfer one to one to a Director of Medical Affairs in Basel. The mechanism transfers better than the percentage.
Now the darker trial. Jiannan Xu, Gujie Li and Jane Yi Jiang tested language models as CV screeners. They found the models prefer CVs they generated themselves over human-written ones, even with content quality controlled. Self-preference ranged from 68% to 88% across major models. In simulated hiring pipelines, candidates who used the same model as the screener were 23% to 60% more likely to be shortlisted than equally qualified people who wrote their own.18 Narcissus has been given an API key and a seat on the hiring panel, and he's voting for his reflection by a comfortable margin. In most industries that's called a conflict of interest. In HR tech it goes into the release notes.
Before you let the machine write your whole life: you don't know which model an employer uses, or whether it uses one at all. In Germany, a Bitkom survey put AI pre-selection of applications at 1% of companies.10 And the person interviewing you has read the CV you didn't write. They'll ask about sentences you're seeing for the first time, live, on camera, in front of three witnesses. It's the closest most of us will get to hearing our own eulogy while we can still blush.
The marketplace: from the price of a Döner to the price of a used Polo
At the bottom of the market, life is cheap. On Fiverr, one seller offers an ATS-friendly resume for $5.19 A single Fiverr category page for ATS resume services listed more than 4,100 of them.20 Five dollars buys you less than a Döner in Berlin, and the Döner at least lets you choose the sauce. At that price and volume, the CV comes out nach Schema F (by scheme F, the German phrase for work done from a fixed template without a pulse). Wer billig kauft, kauft zweimal (buy cheap, buy twice). Here you buy once, send it forty times and get buried forty times, which is at least an efficient use of the purchase.
In the middle sits the German Bewerbungsservice (application-writing service). StepStone puts a basic package of cover letter and CV at typically €100 to €300.21 One German provider advertises customer satisfaction above 99%, a result the old East German Volkskammer would have found a touch vulgar. The same provider says that if a recruiter does notice an application was ghost-written, customers report it goes down well that someone spent money to get the job.22 Speaking as a recruiter: nothing moves me like the faint scent of a card transaction. I once wept at a Mastercard receipt.
Then there are the generators, where the humour turns fully noir. The Verbraucherzentrale Hessen (the Hesse consumer protection agency) described a man who paid €1.99 once to download a CV from an online generator. A few days later €29.99 left his account, and the following month it left again. He had signed up for a subscription without realising it.23 It's the only CV product I know of that keeps working long after you've stopped using it, the way some fungi do.
In the English-speaking market the free review is a classic of the genre. TopResume says its free review usually takes around 48 hours and lands in your inbox.24 Sites that sell competing services describe that review as a sales funnel, built to surface weaknesses and route you into a paid package, and note that the interview guarantee applies to a rewrite rather than a refund.25 In fairness, those critics have shop windows of their own to fill. The guarantee model is common. A comparison site lists another large provider's 60-day interview guarantee as a free rewrite if no interviews come in.26 If the parachute fails to open, the manufacturer will happily send you a second one from the same batch, delivered to the address on file.
The reviews that sell all this now have a regulator, at least in the United States. Since 21st October 2024, an FTC rule prohibits selling or buying fake reviews and testimonials, buying positive or negative reviews, and running company-controlled review sites that pose as independent.27 Violations can mean civil penalties of up to $53,088 each.28 Do the maths on a testimonial wall with forty invented, radiantly grateful clients: a potential liability of $2,123,520. Fiction has rarely paid this badly, and literary fiction has been trying for centuries.
And the credentials, which travel well beyond the US? The best known is the CPRW, the Certified Professional Resume Writer. The exam asks you to write a CV and cover letter for a mock client within 72 hours.29 O*NET lists the certification as having no minimum education level and no minimum work-experience requirement.30 That's not nothing. Plenty of good writers hold it, and writing well is half the job. But it certifies the writing. It says nothing about whether the holder has ever defended a shortlist to a hiring manager who had already decided to hire his cousin.
In 1964 German television began a series warning viewers about Nepper, Schlepper, Bauernfänger (swindlers, touts and yokel-catchers). The title outlived the programme because the cast keeps regenerating, like a liver. Most people in this trade are honest. The dishonest ones are simply better at SEO, in every language, and their conversion rates would make an oncologist weep.
A Beipackzettel (package leaflet) for CV services
Every medicine in the EU comes with a Beipackzettel (the patient information leaflet folded into the box, listing side effects nobody reads until they develop them). The EU even standardises the frequencies: very common means more than 1 in 10 users, rare means up to 1 in 1,000. CV services ship without a leaflet in any language, so here's one. The frequencies are my own clinical estimate, dosed with irony. Run it on anyone before you pay, including me.
- They sell you a score (very common). The commodity pitch is "we'll get you to 90+". The test: ask what score their last client reached inside the employer's actual system. They can't know, because they've never been inside it. I don't hand out ATS scores, and in my experience the serious writers rarely do. The only score that counts is the rate at which applications turn into interviews, and that one arrives by email, usually on a Friday at 16:47.
- The free review finds you guilty (very common). Send them the CV of a friend who was hired last month. If it comes back "weak, urgently needs professional help", you've learned what the review is for, and your friend has learned they're apparently unemployable despite starting on Monday.
- Nobody interviews you (common). A good CV is an extraction of evidence, and extraction needs questions: budgets, headcount, batch sizes, submissions, audit findings closed, the number that moved because you were there. If a provider sends a template and a questionnaire and never asks a single follow-up question, you're buying die Katze im Sack (a cat in a sack, the German version of a pig in a poke). Sometimes the cat is fine. Sometimes it's been in the sack a while.
- They've never sat on the hiring side (common). Ask how a recruiter finds a candidate inside an applicant tracking system. Someone who has done the work will describe keyword searches across parsed profiles, knockout questions on the form, and a hiring manager who reads page one properly and page two in the lift. They'll say it in boring detail, with the thousand-yard stare of someone remembering a requisition that was reopened four times. Someone who hasn't will talk about "beating the algorithm", like a man who has read about war on Wikipedia.
- The guarantee hands you more of the same (common). Ask, in writing, what happens if nothing changes after sixty days. If the answer is another rewrite by the same process, you've been offered a second helping of the dish that made you ill.
- The reviews can't be checked (uncommon, but fatal). Look for reviews on a platform the provider doesn't control, written by people with names and faces. Then glance at the payment page for a subscription hiding behind a €1.99 button, the way a lawyer hides behind a grieving widow.
- They offer to improve the facts (uncommon, and terminal). A title rounded up, a skill you don't have, dates stretched across a gap. The interview usually finds it within twenty minutes, and a reference check finds the rest. If you're over 50, this is where good and bad writers separate most clearly. A good one trims the 1990s to a single line and puts your recent, high-stakes work on page one. A bad one quietly edits your decade of birth and leaves you to explain the arithmetic to a panel of 34-year-olds with calculators.
- The product name does the selling (very common). "ATS-proof", "DIN-konform (DIN-compliant)", "CV zgodne z ATS (ATS-compliant CV)", "guaranteed to get past the robots". That's Augenwischerei (eyewash). Da liegt der Hund begraben (that's where the dog is buried, meaning that's the heart of it, and in this trade occasionally the dog as well): when the label is the product, the contents are usually a template with your name pasted in.
What good looks like from the recruiter's chair
I've read CVs that were technically alive but showed no signs of consciousness. A strong one is different on sight. The target role is obvious before I scroll. Each bullet carries one result and, where possible, one number. Job titles read the way the industry says them, not the way a 2019 reorganisation named them before it was itself reorganised. Page one answers the question I'm actually asking: can this person do this job, here, soon, without me having to explain them to a hiring manager who's already late for the next meeting?
A good writer also knows which funeral they're dressing you for. A photo is still common on CVs in Germany, Austria, Switzerland and Poland, and widely advised against in the UK, the US and Canada. A two-page CV is normal for experienced people in DACH. In North America, a CV (the long, academic kind) and a résumé (the short, industry kind) are different documents altogether. A provider who sells one "international" template for every market is dressing every corpse in the same suit.
For entry-level candidates, a good writer turns a master's thesis into evidence of the skills a GxP environment needs, instead of listing modules like an obituary of every seminar attended. For international candidates who need sponsorship, a good writer states work authorisation plainly: an EU Blue Card, the Chancenkarte (Germany's points-based opportunity card for jobseekers), or "no sponsorship required" if that's true. Leaving it to guesswork turns a knockout question into a guillotine. For executives, the CV is read by a headhunter first and a board member second. It needs the P&L, the headcount, and the decision you took that others were too frightened to take. Adjectives about vision won't do that job. Visionaries are, after all, the people most often seen walking into walls.
None of this needs a dial. It needs someone who has been on the receiving end pi mal Daumen (pi times thumb, meaning roughly) fifty thousand times, and has the mild scarring to prove it.
The advert, as promised
I've spent 17 years in recruitment: the last nine at Bayer and the five before that at the global CROs ICON and Syneos. I've recruited in most EU countries and in the United States, up to SVP level. Today I work with candidates in DACH, across Europe including Poland, and in North America, with especially deep knowledge of the German-speaking market.
Two results I'm proud of. Two senior communications directors rated my rewrite of their CVs as very strong. These are people who have spent their careers improving other people's words, which makes them the most dangerous audience a CV writer can face, roughly like cooking for a food critic who also owns a knife collection. The other is a Clinical Trial Manager who came to me with zero interviews. In reverse recruitment, where I write and submit the applications, five applications produced four interviews.
That last one is exceptional, and I'd be misleading you if I sold it as the norm. Most clients get clearly more interviews with my version of their CV than with their own. Few go from zero to eight interviews out of ten applications. When the numbers don't move, the cause is usually targeting, timing or the market, and we work on those instead of polishing a dial. I don't guarantee a job, and I won't give you a score. If you need a service provider with a guaranteed delivery rate, I can recommend a funeral home. They almost never miss a deadline. My reviews are Google reviews with real names attached, on a platform I can't edit, which I've come to regard as a mercy.
Your turn
Likes are pleasant, the way a nicely arranged wreath is pleasant. A comment helps the next reader far more, and a repost reaches the friend who's about to pay €349 to resurrect a green dial.
- What's the highest ATS score you ever reached, and how many interviews did it buy you?
- If you paid for a CV rewrite, what changed in actual interviews, not in percentages?
- Recruiters and hiring managers, wherever you sit: can you spot a purchased CV, and does it count against the candidate?
- Polish readers: do you still add the klauzula RODO (GDPR consent clause), and has anyone ever read it?
- For readers over 50: did a CV writer ever suggest editing your dates, and what happened at the interview?
- Which German phrase should enter the English language next, now that Schadenfreude has been naturalised?
Subscribe to the newsletter if you'd rather get the next issue in your own feed than via a friend's screenshot at 23:40. And save this one for the day the dial turns red.
The full picture of what I do is at www.morethancareer.de: career coaching (job search strategies, interview preparation and salary negotiation training with tailored scripts), reverse recruitment, and CV and LinkedIn rewrites and rebranding. Subscribers to my email Inner Circle get 5% off coaching and reverse recruitment programmes, with exclusive career tips for the list coming in future: morethancareer.de/#newsletter
Prost. Und lasst euch keinen Score andrehen. (Cheers. And don't let anyone palm a score off on you.)
Sources and references
[1] Jobscan, Jobscan tutorial (match rate of at least 75%): jobscan.co
[2] Jobscan, What Jobscan match rate should I aim for? (80%): jobscan.co
[3] Jobscan homepage (75% recommendation, over-optimisation warning, "3X more interviews" claim): jobscan.co
[4] SimpleCVBuilder, ATS resume checker: what score do you actually need?, 21st April 2026 (vendor source): simplecvbuilder.com
[5] Enhancv, Busting ATS myths: testing of popular resume builders (vendor source, own test): enhancv.com
[6] Greenhouse, What really happens after you apply for a job (candidate blog): my.greenhouse.com
[7] Jobscan, Greenhouse ATS: what job seekers need to know: jobscan.co
[8] Greenhouse Support, Unsuccessful resume parse, last updated 2nd March 2026: support.greenhouse.io
[9] Bewerbungen.de, DIN 5008 Bewerbung: diese Regeln gelten: bewerbungen.de
[10] Andreas Schulz, The 75% auto-rejection myth: what is actually reading your CV, September 2026 (includes the careers portal check and the Bitkom survey source): morethancareer.de
[11] Vanner, Jak napisać CV pod systemy ATS? (Polish provider repeating the 75% claim), February 2026: vanner.pl
[12] Fundacja Panoptykon, Od klauzuli w CV do rozmowy z algorytmem, czyli RODO w rekrutacji: panoptykon.org
[13] Infor.pl, Zmiany w Kodeksie pracy od 4 maja 2019 r. (RODO): infor.pl
[14] Conseil d'État, CV anonyme, 2015: conseil-etat.fr
[15] The Register, Watchdog deep-sixes job ad that was actually pay-to-play training course, 19th December 2024 (ASA ruling on IT Career Change Ltd): theregister.com
[16] Government of Ontario, Your guide to the Employment Standards Act: requirements related to publicly advertised job postings: ontario.ca
[17] Wiles, E., Munyikwa, Z. and Horton, J. J. (2025), Algorithmic writing assistance on jobseekers' resumes increases hires, Management Science 71(12): pubsonline.informs.org (working paper: nber.org)
[18] Xu, J., Li, G. and Jiang, J. Y., AI self-preferencing in algorithmic hiring: empirical evidence and insights, arXiv 2509.00462, revised 9th February 2026: arxiv.org
[19] Fiverr, example gig offering an ATS-friendly resume for $5: fiverr.com
[20] Fiverr, ATS resume services category: block.fiverr.com
[21] StepStone, Bewerbung schreiben lassen: Kosten, Anbieter, Alternativen: stepstone.de
[22] Provider website quoted for its own satisfaction and detection claims: die-bewerbungsschreiber.de
[23] Verbraucherzentrale Hessen, Versteckte Abofallen bei Online-Lebenslauf-Diensten, 4th April 2025: verbraucherzentrale-hessen.de
[24] TopResume, Free resume review: topresume.com
[25] Resume Optimizer Pro, TopResume reviews 2026 (competitor source, commercially interested): resumeoptimizerpro.com
[26] IGotAnOffer, Best resume writing services 2026: igotanoffer.com
[27] Federal Register, Trade regulation rule on the use of consumer reviews and testimonials, effective 21st October 2024: federalregister.gov (FTC announcement: ftc.gov)
[28] Crowell & Moring, Keeping it real: FTC targets fake reviews, December 2025: crowell.com
[29] PARWCC, How to become a Certified Professional Résumé Writer: parwcc.com
[30] O*NET OnLine, Certification: Certified Professional Resume Writer (CPRW): onetonline.org
© 23rd September 2026 Andreas Schulz. All rights reserved.
Ready to take control of your pharma career?
Book a free getting-to-know call. I will tell you honestly where you stand, what the DACH pharma market looks like for your profile, and what your next move should be.
Book a Free Getting-to-Know Call