Resume tower · floor

Are cover letters necessary, and who actually reads them?

Whether are cover letters necessary has an answer depends on a question you can settle in ten seconds before any advice becomes relevant: what does this particular application ask for? Every posting falls into one of three cases, and only one of them is a decision.

That trichotomy removes most of the anxiety, and it is the part almost no page about this subject leads with. Check the posting first. If it says required, the discussion is over.

Which brings us to the optional case, where people actually get stuck — and to the evidence, which exists, is better than the folklore, and comes with a caveat that has to be read alongside it.

What those numbers mean if you accept them at face value. The gap that matters is not between sending a letter and not sending one. It is between a tailored letter and a generic one. Eighty-one per cent of the surveyed professionals distinguished sharply between the two, which fits the USAJOBS description of purpose: a letter earns its place by carrying something the résumé does not. A generic letter carries nothing new by definition, so it costs you time and returns very little.

That reframes the practical question completely. It is not “should I write a cover letter”. It is how many applications can I afford to tailor — and the honest answer for most people is fewer than they are currently sending. Ten applications with real letters will beat forty with a template, and it is less work.

What to put in one, given all of the above. The purpose statement does the work here. Anything that is on your résumé is the wrong material. What belongs is the part a list of roles cannot carry: why this employer rather than a similar one, how a piece of your experience maps onto something specific in the posting, and an explanation for anything on the résumé that raises an obvious question — a change of field, a gap, a move to a different country. That last use is the one people underuse most, and it is where a letter is genuinely irreplaceable.

Length is not the interesting variable, but since people ask: it should be shorter than the résumé it accompanies, and it should be readable in one screen. If it takes longer to read than the document it is introducing, something has gone wrong with the idea.

Should AI write it? That question belongs to a different floor, but the short version follows directly from everything above. The evidence says a generic letter is worth little, and the fastest way to produce a generic letter is to ask a model for one and send what comes back. Used to draft from material you supply about a specific posting, the tools are useful. Used to generate a letter from the job title alone, they are an efficient way to produce exactly the thing the research says does not work.

Who is telling you that you need one. Two of the five results currently above this page on Google are companies that sell cover letter builders. That does not make them wrong — the evidence above broadly supports writing one — but it is worth knowing that the pages most confident about the answer are the ones with a product attached to it. The same is true in reverse of the advice to skip letters entirely, which tends to come from people applying at volume.

The position this floor takes is narrower and, we think, more defensible. Check the posting, because that settles most cases. Where it is optional, write one only if you have something to say that the résumé does not already carry — and if you do not, that is a real answer rather than laziness. And treat the fifty-three per cent figure as what it is: the best measurement anybody has published, six years old, funded by a firm that sells the service.

What this tower will not do

It will not tell you cover letters are dead. The only large experiment on the question says the opposite, and the pages saying otherwise are usually describing high-volume applying rather than evidence.

It will not present the fifty-three per cent figure as a fact either. It is quoted above with its date, its sample and the disclosed commercial interest of the firm that produced it.

And it will not sell you a cover letter builder, which is what two of the five pages currently above this one on Google are ultimately for. What holds instead is simple: the purpose statement here is quoted from a United States government application system, and the experiment is quoted with its dates, its sample size and the disclosed interest of the company that ran it.