A Resume Review That Doesn't Ask What Job You Want Is Just Proofreading

Almost every resume review returns the wrong class of feedback. You get formatting notes, verb swaps, keyword density, a score out of 100. What you needed to hear was that you are applying one level too high, that your bullets describe the job rather than you, and that the top third of page one never states the claim the rest is supposed to prove.

Here is the part most people get backwards. They read shallow feedback as a lazy reviewer. It usually isn't. Surface notes dominate because surface is cheap to evaluate and easy to hand over as a deliverable. Structure takes someone who understands the role you are targeting and will say something uncomfortable about your positioning. That costs more to produce, so almost nobody sells it.

Why does every resume review come back with the same five notes?

Because those five notes can be generated without knowing anything about you. Font consistency, bullet length, passive verbs, missing keywords, a parsability score. Each can be judged from the document alone, by a person or a script that has never seen the job you want. That is what makes them sellable.

The scan those notes claim to optimise for is genuinely fast. Ladders' eye-tracking work clocked recruiters at an average 7.4 seconds on a first-pass resume screen, and every review report quotes it as the reason to tighten your formatting. But run the logic properly. In seven seconds nobody grades your verb choice. They are checking whether you look like the shape of the hire they are making. Formatting matters only because it decides how fast that question gets answered. It is not the question, and where those seconds land is covered in the seven-second scan.

The second reason surface feedback stopped paying: the surface got fixed for everyone at once. When every candidate runs the same draft through the same models, the polish converges and stops sorting anyone, which is why AI resume tools made everyone identical rather than better. Improving your formatting relative to 2015 does nothing for you relative to the pile you are in.

What feedback actually decides whether you get called?

The kind that requires someone to reason about a specific role. Recruiters do not weigh your document evenly. In SHRM's study of over 1,030 recruiters, 84% named years of experience and 83% named type of experience as important at screening, against 37% for credentials and 35% for education level. The weight sits on whether your experience fits this job at this level. The study's own framing: recruiters want "enough experience to succeed in the job, but not so much that a more senior role would be a better fit."

That is a structural judgment, and no scanner makes it. Here are the five failures that cost interviews, none of which a font check finds.

Structural failure	What it looks like	Why surface reviews miss it
Wrong level	Director applications with the scope of a senior IC	Needs the posting's level, not your document
Duties, not results	Every bullet restates the job description, no outcome	The bullets read as clean, correct English
No claim in the top third	Page one opens with adjectives or education	Scores reward a tidy header, not a load-bearing one
Title-to-evidence gap	Title says Head of; evidence shows one project, no headcount or budget	The reviewer would have to doubt you out loud
Wrong market	The role is contracting where you are applying	Requires live hiring data, not your file

The second is fastest to fix: a bullet naming what changed and by how much outruns a perfect duty list every time, and the rewrites are in resume bullets that describe you, not the job. The first is the most expensive and the least often said out loud. Compare two notes a reviewer could give the same person.

Weak review note: "Consider replacing 'managed' with a stronger action verb and adding two or three keywords from the job description. Formatting is clean, but bullet lengths are inconsistent."

Strong review note: "You are applying to Senior Engineering Manager roles with the evidence of a strong tech lead. One team of four, no hiring, no budget, no cross-team programme. Either aim at EM roles where a team of four is the standard span, or spend six months collecting the evidence the senior title needs and apply then."

The second note is unwelcome, specific, and worth a hundred verb swaps. It is also impossible to write without knowing which job the reader is chasing.

Can a resume be reviewed without a job posting attached?

Barely. Without a target role a review can only check hygiene, and hygiene is table stakes. Harvard's career services guidance says it plainly: "tailor your resume to the type of position you're seeking." If the reviewer does not know the position, the useful half of the review is unavailable to them.

The employer side agrees. In the Harvard Business School and Accenture study Hidden Workers: Untapped Talent, which surveyed 2,275 executives and 8,720 workers across the US, UK and Germany, 88% of employers agreed that qualified high-skills candidates get vetted out because they do not match the exact criteria established by the job description, rising to 94% for middle-skills roles. The screen runs against the posting, so the review has to run against the posting too. NACE's Job Outlook 2026 Spring Update, a survey of 185 employer respondents, lands in the same place: it is not enough to list your skills, employers want examples, connected to the jobs you are applying for.

Then there is the failure no document review can reach: the market. In August 2026, US leisure and hospitality added 62,000 jobs while information and financial activities lost a combined 34,000. A resume can be excellent and still point at a category shedding roles where you live. Before paying anyone to sharpen the document, count the live postings for the exact title and level you are chasing, which is the first thing Praxy Jobs is for.

Is a "free resume review" feedback, or a lead magnet?

Usually both, and the order matters. The mechanic is simple and entirely legal. You upload a resume, a scan or a reviewer returns a report naming several problems, and the report sits beside a paid service that fixes exactly those problems. TopResume, as a public example of the model, advertises personalised feedback in under 48 hours including an ATS analysis and a read on strengths and weaknesses, alongside a paid catalogue of resume writing and coaching. Nothing hidden. It is a funnel, disclosed on a product page.

What matters is what the incentive does to the content. A free review has to find enough wrong to make the paid fix feel necessary, in minutes, at zero marginal cost, for a stranger whose target job it does not know. Those constraints select hard for surface findings: formatting gaps, missing keywords and a low score are abundant, fast to detect and easy to render as a list. "Your evidence does not support the level you are applying at" is slow, singular and hard to productise.

Treat it as what it is: a hygiene pass worth twenty minutes, plus a sales conversation you can decline. Do not mistake the findings for a diagnosis.

When is a professional resume writer genuinely worth the money?

Often enough that dismissing the category would be dishonest. The real question is whether you are buying judgment or typing.

On price, use published figures rather than folklore. TopResume's pricing guidance puts entry level at around $200, professional level at $200 to $400 and executive level at $350 to $700, and says to expect $200 to $1,000, sometimes more. That is one large vendor's published range, not an industry average. Independent writers and boutique executive firms sit outside it in both directions.

Four cases where paying is the right call:

• Senior and executive positioning. Above a certain level the resume is a narrative about scope, P&L and organisational change, not a record of tasks. You are buying an outside read on how large your career looks to a stranger.
• A hard career change. When fifteen years in one function has to read as preparation for another, that is a structural rewrite. The skill being sold is framing, not prose, which makes it the most defensible thing here to outsource.
• A language you do not write natively. Fluency in speech and fluency in the compressed register of a resume are different skills.
• Rigid formats. USAJOBS expects each role to carry employer name and address, job title, dates, duties and hours per week for qualification purposes, and a missing field can mark an application incomplete. Academic CVs have their own conventions. Where the rule is rigid and the penalty binary, expertise in the rule is worth buying.

Where it is usually not worth it: a normal mid-career record, targets you already know, and a document that reads flat. That is a content problem, and the content is yours.

The 20-minute self-review, in order of what moves callbacks

Run these in order and stop when time runs out. The order is the point: most reviews run it backwards.

#	Check	Time	Why it ranks here
1	Name the target. One real posting beside the resume. Not a role family.	2 min	Everything below is meaningless without it
2	Level test. Does your evidence show the scope that posting implies: team size, budget, blast radius, decisions owned?	5 min	Type and years of experience dominate screening at 84% and 83%
3	Top-third claim. Cover everything below the top third. Can a stranger state what you are and the proof behind it?	3 min	The only part reliably read on the first pass
4	Duty audit. Mark every bullet true of anyone with your title. Rewrite the top three into outcomes.	5 min	Employers want examples, not lists of skills
5	Criteria pass. Line the posting's requirements against your page. Anything you meet but never said, say it.	3 min	88% get vetted out for missing the posting's exact criteria
6	Market check. Live postings for this exact title and level in your location?	1 min	A perfect resume aimed at a shrinking category still fails
7	Hygiene. Single column, real text, standard headings, clean parse.	1 min	Worth doing once, not obsessing over

Note where the score sits: nowhere. A score out of 100 measures hygiene, not fit, and the "an algorithm rejected you for a keyword" story it arrives wrapped in does not survive contact with how these systems work. The automatic screen-outs that exist are human-configured knockouts. Hidden Workers found almost half the companies surveyed screen out any resume showing an employment gap over six months, on that basis alone. Somebody typed that rule in, and it says nothing about your verbs. The full argument is in the ATS myth, and nothing here contradicts it.

This is also why Praxy gives resume feedback against a specific posting and will not write the resume for you. A resume somebody else wrote is a document you then have to defend in an interview, line by line, to a person who does this all day. If you cannot explain how you measured the number in bullet two, that number is a liability. Feedback improves the document and leaves you able to stand behind it. Ghostwriting does the opposite.

What to do now

1. Pick one posting and put it next to your resume. Not three, not a role family. The rest only works against a specific job at a specific level.
2. Run the level test first. Write down the scope the posting implies, then find that evidence on your page. If it is missing, you have a targeting decision, not a formatting one.
3. Rewrite your top three bullets into outcomes. What changed, by how much, over what period, and defensible under questioning.
4. Do the hygiene pass once and close the tab. Single column, text not images, standard headings.
5. Check the market before you pay anyone. If live postings for that exact title and level are thin, the problem is your target, and no writer at any price fixes that.

Want a resume review that starts by asking which job you are actually going for? Message Praxy on WhatsApp, send the posting and the resume, and get a read on whether your evidence supports the level, what your top third is claiming, and which bullets describe the job instead of you. Praxy will not write it for you, on purpose, because the version you can defend is the one that works.