Do Employers Actually Read Cover Letters? The Evidence, Checked
83%, 94%, 77%, 81%: search for whether employers read cover letters and you will find all four, each quoted as settled fact, though they measure different things. The honest answer is that nobody knows how many UK employers read cover letters, and almost every figure you will see was published by a company that sells CV or cover letter tools. So do we: we write cover letters for £6.99, which is worth keeping in mind as you read this.
We sell cover letters
We gain if you believe cover letters matter, because we charge £6.99 to write one. That puts us in the same position as every company whose survey we question below, so we have tried to hold ourselves to the same standard: every figure here links to where it came from, and where we couldn't check something, we say so.
Where the numbers come from
These are the figures that turn up most often, traced back as far as we could take them:
- 83% "always" or "frequently" read cover letters, 94% find them "influential" when deciding who to interview, and 67% call them "important". All three come from one Resume Genius survey of 625 US hiring managers, run through the online panel Pollfish in 2023. Resume Genius offers a cover letter builder.
- 81% have rejected a candidate based solely on their cover letter, and 89% expect one. From Zety's survey of 753 American recruiters and HR staff, run online on 4 and 5 June 2024. Zety sells a resume builder.
- 77% give preference to candidates who send a cover letter, even when it's optional. From a ResumeLab survey of 200 recruiters, HR specialists and hiring managers, carried out by the polling firm OnePoll and reported by HR Dive in February 2020. ResumeLab runs resume and cover letter builders. Its survey page wouldn't load for us, so we couldn't confirm the exact wording or which country the respondents were in.
- 57% say cover letters are "influential or extremely influential" in their hiring decisions. From CV Genius, April 2024: 625 people who had hired in the UK or Ireland in the past year. It is the only UK survey we found. CV Genius sells a CV builder with a cover letter builder alongside it.
Every one of these is a survey by a company that sells CV or cover letter tools, asking people whether something matters. Two of the four surveyed Americans, one surveyed the UK and Ireland, and for one we couldn't confirm. The one UK figure is far lower than the American 94% in the Resume Genius survey, but the question was worded differently, so the gap may say more about the wording than about the country.
The Harvard Business Review figure that isn't HBR's
The 83% is often credited to Harvard Business Review. iCover, a UK CV writing service, gives HBR as its source. HBR did publish the figure, in a March 2025 article by Joel Schwartzberg, but it didn't run the research. The article says plainly that the 83% comes from Resume Genius's 2023 survey and the 81% from Zety's, and it describes Zety as "a career service that helps people improve their resumes and cover letters".
HBR named its sources. The problem starts one step later, when a vendor's survey is passed on under a famous publisher's name and the vendor drops out of the sentence. A figure that would raise an eyebrow as "a CV builder says" sounds authoritative as "Harvard Business Review says".
Where "67% of UK recruiters" seems to come from
You may have seen 67% quoted as the share of UK recruiters who think cover letters matter. We couldn't find a UK survey that produced it. The closest match is the Resume Genius survey above, where 67% of US hiring managers called cover letters "important" when deciding who to interview. If that is the source, it is an American figure with a British label. If it comes from somewhere else, we couldn't find where.
What a survey can and can't tell you
A survey asks hiring managers what they do. It can't watch them do it. Someone asked "do you read cover letters?" has every reason to say yes: it sounds conscientious, and the question itself hints at the answer.
The wording matters too. "Always or frequently read", "influential", "important", "expect" and "have rejected someone because of" are five different questions. An employer could truthfully answer yes to all five while skimming most letters for a few seconds and never opening the rest.
None of this makes the figures false. It means they measure opinions about cover letters, gathered by companies with an interest in the answer, and they can't tell you whether yours will be read.
The one experiment, and who ran it
The closest thing to hard evidence is a field experiment by ResumeGo, which describes itself as "a company that offers resume writing services". Between July 2019 and January 2020 it sent 7,287 fictitious applications to real job adverts on job boards such as ZipRecruiter, Glassdoor and Indeed. Some had no cover letter, some had a generic one, and some had one written for the job. Applications with a tailored letter got a 53% higher callback rate than those with none.
This is better evidence than any survey, because it measures what employers did rather than what they said. It still has limits: it was run by a company that sells the service, it hasn't been peer reviewed, the write-up doesn't say which countries the jobs were in, and it was finished before ChatGPT existed. That last point matters more now than it did then.
What happened when AI started writing them
Two studies have looked at this. Both are by economists, both use data from one freelance platform, Freelancer.com, and neither has been peer reviewed yet. On Freelancer.com, workers bid for short paid jobs, such as a few days of coding, with a written pitch that works as a cover letter. The first study covers only PHP coding and internet marketing jobs. That is a long way from applying for a salaried job in the UK, so read what follows as a clue, not a verdict.
Source: Cui, Dias and Ye, Signaling in the Age of AI: Evidence from Cover Letters, Yale University, working paper revised November 2025. 5.5 million cover letters for PHP and internet marketing jobs on Freelancer.com, January to September 2023. Not peer reviewed.
The fading is the part most often retold, and the part most often stretched. The paper doesn't say cover letters stopped working after two months. It says the boost from using the AI tool faded after two months, even though the tool kept producing letters just as closely matched to the adverts. The authors suggest employers may have got better at recognising AI-written letters, and say their data couldn't capture that. Employers also leaned more on workers' past work history. Workers who spent longer editing the AI draft did better, but the authors say this "should not be interpreted as evidence of a causal relationship".
A second paper, by Galdin and Silbert at Dartmouth and Princeton, used Freelancer.com data on coding jobs and points the same way: employers were willing to pay more for workers with tailored applications before tools like ChatGPT arrived, "but not after". It is also an unreviewed working paper about freelance bids.
What both suggest is narrow but plausible. A tailored letter used to be evidence that someone had made an effort. When software can tailor a letter in seconds, it stops being evidence of much. Whether that has happened in UK hiring, nobody has measured.
What we looked for and didn't find
We looked for a UK source with nothing to sell: the CIPD, which regularly surveys employers on recruitment; the REC, the recruitment industry's trade body; the Institute of Student Employers; and the large job boards. We found nothing from any of them on whether cover letters get read.
The closest was Salt, a recruitment agency, which ran polls on LinkedIn in 2025 and reported that 30% of hiring managers said they rarely read cover letters. It didn't say how many people answered.
So there is no independent UK figure. If you see one quoted with confidence, ask who ran the survey and what they sell.
What the evidence supports
Nobody can tell you what share of UK employers read cover letters. Anyone who gives you a precise figure is quoting a survey by a company that sells CV or cover letter tools, often an American one.
What the evidence supports is smaller. In one experiment from before ChatGPT, run by a CV writing company, a letter written for the job did better than no letter. Cover letters seem to work partly as proof of effort, and early research from one freelance platform suggests that proof counts for less once AI can tailor a letter for anyone. Letters probably count for more in some sectors and roles than others, but nobody has measured which, so treat anyone who tells you exactly where they matter with the same caution.
Our cover letter page sets out when we think a letter is worth sending and when it is likely to be ignored. That is advice, not data, and we sell the letters, so weigh it the same way.