How do I find which gaps actually matter?

Read five current adverts for the role you want and write down what repeats. A skill named once is one company's preference; a skill named in four of five is what the market means by that role. This takes an hour and it replaces guessing with counting, which is the whole method.

How many gaps should I take on at once?

Three. A list of ten is a way of not starting, and progress on ten simultaneous fronts is invisible on all of them. Three is small enough to finish and large enough that finishing changes what you can apply for, which is the point.

Does ClearHire analyse my skills gap for me?

No. This is a guide, not an analyser — ClearHire holds no dataset of skills by role and does not rank gaps by frequency in live listings. The reading exercise above is the manual version of that, and it has the advantage that you see the adverts yourself rather than trusting a summary of them.

How do I know when a gap is closed?

When you have finished something with it that someone else can look at. A course completion certificate says you attended; a working thing says you can do it, and it is also what you will talk about in the interview. If you cannot point at an artefact, the gap is narrower than it was and not yet closed.

What if a gap is not realistic to close?

Decide that explicitly and adjust the target, rather than leaving it on the list. Some gaps need years, a different job, or a credential you are not going to pursue — and a plan carrying an item nobody intends to do quietly discredits the rest of it. Either the target changes or the timescale does, and saying which is the useful outcome.

Should I close gaps in public?

Where you can, yes, because it does two jobs at once. Working in the open — a repository, a write-up, a talk — creates the evidence you would otherwise have to assert, and it tends to raise the standard you hold yourself to. The exception is anything involving an employer's work, which needs permission first.

content:contentPage.section.career

Skills gap analysis

Work out which skills your target role expects that your current one does not, and which of them matter most.

Last reviewed: 19 August 2026
Length: About 5 minutes
Scope: Maintained, not published. Corrections welcome.

A skills gap is the distance between what you can do today and what the role you want expects. This page is the framework for closing one — how to find the gaps that matter, how many to take on at once, how to build evidence while you close them, and how to tell when one is actually closed. It is a guide and not an analyser: ClearHire does not scan job listings or rank your gaps for you, and a page that implied it did would be describing a product that does not exist. The ranking you need comes from reading five adverts for the role you want and writing down what appears in most of them, which takes about twenty minutes and is more honest than any score.

Who this is for

Where it applies.

Building a focused 90-day learning plan

Pick the top 3 gaps from the analyzer. Sequence them — usually 1 hard skill, 1 tooling skill, 1 process skill. 60–90 days of focused practice with one shipped artifact per gap closes more career distance than a year of scattered course-watching.

Deciding whether a gap is realistic to close

Some gaps are 3-month closes (a new framework). Some are 12+ month closes (a new programming paradigm, a new domain). Be honest about the time cost before committing. Underestimating closes you cannot afford to make is the most common career-change failure mode.

Knowing when a gap is actually closed

When you can ship a small but real project using the skill end-to-end without help, and explain the tradeoffs you made out loud. If you cannot teach it, the gap is not yet closed — you have just been exposed to it.

In order

How it works.

Read five adverts and write down what repeats
Not one advert — five, for the role you actually want, at companies you would actually join. What appears in three or more of them is the market speaking. What appears once is one hiring manager's preference.
Mark each one honestly: have it, half of it, none of it
Half is the interesting column. Those are the gaps that close in weeks rather than quarters, and they are the ones people skip because admitting to half feels worse than admitting to none.
Pick three, not ten
Usually one hard skill, one tool, one way of working. Above three, depth suffers and none of them reaches the point where you could be questioned about it.
Pair each gap with something you finish
A course gets you to working knowledge; a finished thing gets you to something you can be asked about. A certificate on its own answers a question nobody asked.
Close it in public, and check it before moving on
A repository, a write-up, a talk — something a stranger can look at. The test for closed is whether you could explain the trade-offs you made out loud, not whether you feel more confident.

Worked

Worked illustrations.

Constructed to show the shape of the decision, not reported cases. The figures are illustrative.

An engineer closing a systems-design gap

90 days: book + course + 5 paper designs + 1 production-grade side project. Public blog post explaining tradeoffs. Lands senior systems-design interviews at month 4; the visible artifact moved the conversation from "do you know systems design?" to "let's discuss your approach".

A career changer underestimating a domain shift

Initial plan: 90 days from finance to ML engineering. Realistic plan: 12 months. The honest re-estimation early prevents 6 months of disappointment when the original timeline cannot hold.

Asked often

Questions people bring here.

How do I find which gaps actually matter?: Read five current adverts for the role you want and write down what repeats. A skill named once is one company's preference; a skill named in four of five is what the market means by that role. This takes an hour and it replaces guessing with counting, which is the whole method.
How many gaps should I take on at once?: Three. A list of ten is a way of not starting, and progress on ten simultaneous fronts is invisible on all of them. Three is small enough to finish and large enough that finishing changes what you can apply for, which is the point.
Does ClearHire analyse my skills gap for me?: No. This is a guide, not an analyser — ClearHire holds no dataset of skills by role and does not rank gaps by frequency in live listings. The reading exercise above is the manual version of that, and it has the advantage that you see the adverts yourself rather than trusting a summary of them.
How do I know when a gap is closed?: When you have finished something with it that someone else can look at. A course completion certificate says you attended; a working thing says you can do it, and it is also what you will talk about in the interview. If you cannot point at an artefact, the gap is narrower than it was and not yet closed.
What if a gap is not realistic to close?: Decide that explicitly and adjust the target, rather than leaving it on the list. Some gaps need years, a different job, or a credential you are not going to pursue — and a plan carrying an item nobody intends to do quietly discredits the rest of it. Either the target changes or the timescale does, and saying which is the useful outcome.
Should I close gaps in public?: Where you can, yes, because it does two jobs at once. Working in the open — a repository, a write-up, a talk — creates the evidence you would otherwise have to assert, and it tends to raise the standard you hold yourself to. The exception is anything involving an employer's work, which needs permission first.

Worth knowing

Small things that help.

This is a framework, not an analyser. Nothing here scans listings or scores you.
Three gaps, not ten. Focus produces something you can be questioned about; breadth produces a list.
Pair every gap with something finished. Course completion on its own rarely moves a conversation.
Public evidence compounds; private learning is invisible to everyone but you.
The threshold is "could I explain the trade-offs out loud", not "do I feel productive".
Some gaps close in a month and some take a year. Being wrong about which is which is the expensive mistake, not being slow.

ClearHire Editorial

Written to be corrected. If something here is wrong, say so and it changes.