How to Choose Certifications Based on Your Current Skills

The most expensive assumption in certification planning is "I should start from the beginning." Plenty of professionals with years of hands-on work buy foundational courses that re-teach what they do every day — while others, with thinner experience than they realise, book associate-level exams that assume habits they have not built yet. Both mistakes come from the same omission: nobody audited what they actually know before choosing what to certify.

This article is that audit. You will inventory your current skills, compare the inventory against real exam objectives, and use the gap — small, medium or large — to decide which certification to take and at what level. It is the inverse of choosing by demand: our sibling guide on choosing a certification based on your target job starts from job postings and works backwards, and the general certification decision framework balances all five selection factors at once. Here, the starting point is you.

Why a skills inventory comes before any shortlist

Exams are written for a candidate profile. Providers say so openly: CompTIA recommends 12 months of hands-on IT support experience before A+, and Network+ plus two years in a security or systems administrator role before Security+; AWS pitches its Cloud Practitioner exam at people with up to about six months of AWS exposure — including non-IT backgrounds — but recommends a year of hands-on solution-design work before the Solutions Architect – Associate. At the far end, the ISC2 CISSP requires five years of cumulative paid experience across its domains before you can hold the full credential.

Read those profiles as calibration marks on a ruler. Your inventory tells you where you sit on that ruler; the right certification is the one written for roughly that position — or one deliberate notch above it. Skip the inventory and you are choosing a point on a ruler you have never looked at.

Build the inventory: three lists in ninety minutes

You do not need a formal competency framework. Three lists, written honestly, are enough.

List 1 — skills you use weekly. Go through your actual work: tickets closed, systems administered, scripts written, spreadsheets built, incidents handled. Name tools and tasks specifically ("reset AD passwords and troubleshoot group policy", not "Windows administration"). Weekly use is the bar because exams at every level lean on fluency, not familiarity.

List 2 — skills you have used, but not recently or deeply. The technologies from a previous role, the platform you touched during one project. These are cheaper to revive than to learn, which matters when you compare study timelines later.

List 3 — skills you are regularly near but have never owned. The firewall a colleague manages, the cloud console you have viewed but not configured. Adjacency is where "extending" certifications pay off best, because context accelerates new learning.

Then pressure-test the lists. Self-ratings drift optimistic, so anchor each claimed skill to evidence: could you demonstrate it in an interview tomorrow? Have you done it unsupervised? If a skill on List 1 fails that test, move it down a list.

Turn the inventory into a gap analysis

Now bring in the exams. For each certification you are curious about, download the provider's official exam objectives or exam guide — every serious provider publishes one — and score each objective domain against your lists: have it (List 1), rusty (List 2), new (List 3 or absent). Ten minutes per exam is plenty at this stage.

The pattern of scores tells you which of three situations you are in, and each situation has a different right move.

Mostly "have it": certify to validate

If two-thirds or more of an exam's objectives sit on List 1, that exam is a validation play: you are buying legible proof of existing competence, and your preparation is mostly about the exam itself — question formats, timing, the provider's vocabulary — rather than new learning. Validation candidates often over-prepare at the wrong altitude, re-studying content they know instead of drilling exam mechanics. A timed practice-test simulation is the efficient tool here: it confirms the knowledge is retrievable under time pressure and surfaces the handful of genuinely unfamiliar objectives, which become your entire study plan. If you are an established professional certifying mainly to strengthen positioning, the adjacent question of which credentials round out a track record is covered in choosing certifications that complement your experience.

A meaningful minority "new": certify to extend

If roughly a third to a half of objectives are new — but the rest are solid — you have found the sweet spot for growth. The certification's syllabus becomes a structured curriculum for exactly the skills adjacent to yours, and your existing base keeps the timeline sane. This is where List 3 earns its keep: an exam whose new material lies in your adjacent zone (the security exam, when you already administer the systems being secured) extends you far more efficiently than one whose new material is alien.

Mostly "new": step down a level, deliberately

If most objectives score "new", the exam was written for someone you are not yet — and the honest move is to certify one tier lower, or to build hands-on exposure first. This is not a demotion; it is sequencing. Foundational credentials exist precisely for this position on the ruler (AWS's Cloud Practitioner, to take one documented example, assumes no prerequisites at all), and some providers make the ladder financially explicit — AWS gives anyone who passes any of its exams a 50% discount voucher toward the next, which rewards starting where you actually are. Beware, too, of hard gates: experience-gated credentials such as CISSP or ISACA's CISA cannot be shortcut by study alone, so a "mostly new" score on one of those means it belongs on a multi-year plan, not this year's. For sequencing several certifications across that longer arc, see building a career roadmap with certifications.