Tailor every CV, or not?

Tailoring helps for two reasons. Screening systems match on vocabulary, so using the posting’s words gets you past filters. And a human reader spends seconds deciding, so putting the relevant thing first changes what they see.

Both are real. The advice was sound; the problem was that doing it properly took an hour per application, so candidates either tailored a handful of applications or none at all.

The two reasons matter differently and are worth separating. Vocabulary matching is mechanical and easy to satisfy; ordering for a human reader is judgement, and it is the half that most automated tailoring does worst.

The cost. Tailoring now takes a minute, so the practical objection has gone and there is little reason not to do it for roles you care about.

What has not changed is the value of each tailored application — and now that everyone can do it, the relative advantage has shrunk. Tailoring has moved from an edge to a baseline: doing it no longer distinguishes you, and not doing it now stands out.

The asymmetry is the practical point. Tailoring costs a minute and buys little, but not tailoring costs nothing and can lose the screen outright — which makes it worth doing as insurance rather than as strategy.

If you are applying for one role type in one industry, a well-written general CV covers most of it. Tailoring then changes a few words and is not worth much, though it still costs nothing.

Where tailoring matters is when your background supports several stories. A developer who has done both data engineering and backend work needs the emphasis to shift depending on the posting, and a single document has to lead with one of them.

A middle option handles most of this without per-application work: keep two or three canonical versions, one per direction you are genuinely pursuing. That captures nearly all the benefit of tailoring, because the reordering is the part that matters and the vocabulary edits are marginal.

• One role type, one industry — one strong CV is mostly fine
• Several plausible directions — tailoring genuinely changes the read
• Career change — tailoring is essential, the framing is the whole case
• Highly technical niche — the exact vocabulary matters to filters

It cannot make you a fit for a role you are not suited to. Candidates use tailoring to stretch, and the tools encourage it, but the interview arrives and the gap is still there.

It also cannot substitute for the signals that survive automation — a referral, visible work, a specific reason for wanting this role. When everybody tailors, those are what separate applications, and time spent on them beats time spent on more variants.

And it cannot rescue a CV whose underlying content is weak. Rewriting the same vague bullet in the posting’s vocabulary produces a vague bullet in the posting’s vocabulary; if the achievements do not say what you actually did and what changed as a result, no amount of reframing supplies it.

Tools left to themselves converge on inserting the posting’s phrases into your existing bullets. It passes a keyword filter and reads badly to a person, because the vocabulary no longer matches the sentence around it.

The worse version invents. Asked to align a CV with a posting requiring Kubernetes, a model that has nothing to work with will produce a plausible sentence about Kubernetes, and a candidate skimming a fluent document will not notice a line that was not there before.

Reading every version before it goes catches both, and it takes ninety seconds. Diffing against your canonical CV is faster still: anything added that you did not do is the whole thing you are looking for.

• Keyword stuffing that reads as machine-written to a human
• Invented experience the model supplied to close a gap
• Seniority inflation — "led" where the truth was "contributed to"
• The previous employer’s name surviving into the new version

Keep one strong canonical CV that reflects you accurately, and tailor from it for roles you actually want. Read every tailored version before it goes, because your name is on it.

Skip tailoring for speculative applications where you are testing the market — those are low-value by design, and spending effort on them is the trap that makes a job search feel busy without moving.

Improve the canonical document rather than the variants. An hour spent making the underlying achievements specific and quantified improves every application you will ever send, which is a better return than an hour spread across forty rewrites.