A consultant’s skills dossier is a sales document

A skills dossier is written the way a proposal is written: for one client, for one assignment, and under the firm’s signature rather than the consultant’s. You are allowed to select from a career history, to put it in order and to rephrase it in the client’s words. You are never allowed to add to it, and it is precisely because the line is that simple that so many dossiers cross it without noticing.

The confusion comes from the format. A dossier looks like a CV, it borrows its headings, and it is often produced by taking the consultant’s CV, stripping the name and pouring it into the firm’s template. But the reader, the author and the stakes have all changed, and a document whose reader, author and stakes have changed is no longer the same document, even when its layout has not moved.

How is a skills dossier different from a CV?

A CV is written by a person describing themselves to every possible employer; a skills dossier is written by a company presenting someone to one specific client. Three things follow from that, and they decide everything else.

The first is authorship. When a candidate polishes their CV, it is their word against reality, and they answer for it alone. When a consulting firm sends a dossier, the firm is the one asserting that this consultant can do what is written, and the client signs a purchase order on the strength of that assertion. Sometimes the consultant has not even read the version that went out.

The second is the reader. A buyer or a project lead receives several dossiers for the same need and reads them looking for an answer to their question, not a portrait. A dossier that tells a career in chronological order forces the client to do the selection themselves, and the client does not bother.

The third is scope. A CV is good for a whole job search; a dossier is good for one assignment. The one that won a cloud migration at an insurer has no business being resent unchanged for a data assignment at a manufacturer, even when the consultant is the same and so are their skills.

What makes a dossier win the assignment?

A dossier wins when its first half page answers the client’s need better than the others do, and almost everything else is secondary. The client reads the title, the summary and the first assignment, and decides at that point whether to read on.

The title uses the wording of the need when it is accurate, not the consultant’s internal job title. The summary runs to four or five lines that say why this profile for this assignment: the closest experience, the requested technology with its real duration, the comparable context (sector, team size, delivery method). Assignments are ordered by relevance rather than by date, and each one says what the consultant achieved instead of listing their tasks.

Presentation matters too, and it is well covered elsewhere. Cobalt has catalogued the presentation mistakes that get a dossier dropped before the interview: a generic title, assignments written like job descriptions, certifications buried at the bottom, stale availability. We will not go over them again, because the question this article asks lies elsewhere: once all of that is fixed, how far can you go to make the dossier fit the need?

Where is the line between tailoring a dossier and lying?

The line runs between four things you can do to a career history: select, order, rephrase and add. The first three are the job; the fourth is a lie, even when it is called making the most of someone.

Selecting means keeping only the assignments and skills that serve the need. Nobody blames a dossier for leaving out a 2014 internship that has nothing to do with the role. Ordering means putting the closest experience first, even when it is not the most recent. Rephrasing means describing the same reality in the client’s vocabulary: if the consultant “owned the product catalogue data” and the client is looking for someone with master data management experience, it is the same thing, and it is legitimate to write it that way.

Adding is a different kind of act. Two years of Kubernetes when the consultant spent six months on it in one project, a lead role when they were the most senior person on a team without being in charge of it, a certification “in progress” they never started: none of these is a rephrasing, because the reality being described does not exist. The most common additions fall on exactly this ground. In Resume Builder’s January 2025 survey, 24% of American candidates admitted to lying on their CV, and of those, 38% had inflated their years of experience and 34% their skills.

The practical test fits in one question, asked line by line: could the consultant defend this sentence in front of the client, at the first technical question? If the answer is no, the sentence comes out. The test has the advantage of needing no casuistry. It does not measure the writer’s intent, only what will happen in the interview, and that is the only measure the client cares about.

Why does lying cost a consulting firm more than a candidate?

Because a candidate who lies is almost never caught out, while a firm that lies almost always is, and by the only party that matters to it.

The survey quoted above is striking on this point: among candidates who lied, only 12% say they faced any consequence at all. Employers do check more than people think: according to HireRight’s 2025 report, more than three quarters of the employers surveyed found discrepancies in candidates over the previous twelve months, most often about education and employment history. But a rejected candidate moves on to the next opening, and for them the cost ends there.

A consulting firm goes back to the same client. The consultant placed on an inflated dossier is found out in the first week, when they are handed the task that headed their summary, and the client’s judgement falls on the firm that presented them rather than on the consultant. We have shown that 79% of firms earn at least half their revenue from clients who come back: that is exactly the revenue a false dossier puts at stake, to win a single assignment.

There is a second, quieter cost. A consultant credited with skills they do not have lands in an assignment where they are struggling from day one, and they know it. The inflated dossier therefore also manufactures mid-assignment departures and early terminations, which will later be blamed on the market, the client or the consultant, and rarely on the document that made them likely. The fraud everyone talks about, entirely fabricated profiles, is a separate subject; here nobody lied about who they are, and the damage to the client relationship is of the same order.

What can an agent prepare, and what must a person sign?

An agent can prepare the whole dossier; it cannot release it. The dossier leaves the company and commits its word to a client, which puts it on the wrong side of the one line we treat as non-negotiable: anything that goes out passes through an approval.

The preparation is nonetheless substantial, and it is exactly what goes missing when an account manager builds three dossiers in an afternoon. Reading the need and pulling out what is actually required. Finding the consultant’s assignments that answer it, including those that appear in an assignment report rather than in their CV. Ordering, rephrasing in the client’s words, checking that dates and durations agree from one section to the next, and removing the names of previous clients that confidentiality forbids citing, which is the most frequent mistake and the most expensive one when the dossier goes through a tool nobody approved.

One thing belongs in that preparation and almost never appears in it: the list of what the profile does not have. An agent comparing the need with the history can say “requested: three years of Kafka; found: one eight-month project in 2023; not verifiable: the volume handled”. That line stays internal: it is there for whoever will sign, so they know exactly where the dossier is thin before deciding how to present it.

Signing then takes two people. The consultant reads first, because they will have to defend every sentence and they are the only one who knows whether it is true; this is the interview test applied directly. The account manager releases second, because it is their client relationship on the line. A dossier generated without anyone having spoken to the consultant is the textbook case of what we describe in work that only looks finished: plausible, quick to produce, and expensive for whoever receives it.

Should the dossier say what the profile is missing?

Most of the time, yes, in one sentence and with how the gap will be closed, and it is the recommendation that meets the most resistance.

The resistance is well founded, and it is worth admitting. A dossier that says “Kafka experience limited to one project, ramp-up planned over the first two weeks” will sometimes lose to a competing dossier that claims three years. A busy client reads quickly, and a written caveat stands out more than a hollow claim. On any single assignment, honesty can cost you the assignment.

But the gap will be found anyway, in the interview if the client asks the right question and in the first week if not, and a discovered gap always costs more than an announced one. The client who read the caveat and chose the profile anyway bought with their eyes open; the one who finds the gap alone reads every later dossier from the same firm with the same suspicion. Over a relationship of several years, the written caveat is the most profitable sales argument in the dossier, because it makes everything else in it credible.

At Balt, this is the part of the work we are happy to give the agent: the preparation, the comparison with the need and the list of what does not fit, laid out in the conversation where the account manager decides. The dossier goes out once a person has read it, and not before.

One question comes up well before the dossier: how long had the consultant been without an assignment before anyone built one for them? The most accurate dossier in the world arrives too late if the end of the assignment was known three weeks after everyone else knew it.