How to Quantify Your Achievements: A Practical Guide to Adding Metrics to Your CV & Performance Review

It is Sunday night. Your performance review is next week and your impact column is a ghost town. Same problem when you open the CV: plenty of things you did, not one number to prove any of them mattered.

I see this constantly. Capable engineers and PMs who genuinely moved the needle, then describe a whole year of it in language so soft a recruiter skims straight past. The work was real. The proof was missing.

Here is the good news. You can almost always recover the numbers, even when nobody handed you a dashboard.

Why metrics matter

Recruiters and engineering managers skim for digits first. A number is the fastest way to tell, at a glance, whether you had impact or just held a title.

No numbers, and one of two things happens. They assume there was no real impact, or your CV simply lands softer than the person next to you who wrote "cut cloud costs 30 percent." Same work, weaker signal. The signal is what gets read.

The fix is not to invent numbers. It is to surface or estimate the ones already sitting inside your work.

The 5-step Metric-Magic playbook

1. Mine the exhaust. The numbers already exist in the tools you use every day. GitHub deploy counts, CloudWatch uptime, Jira story points, Stripe dashboards. Go and look before you decide there is nothing there.
2. Derive defensible estimates. When there is no clean figure, calculate one and own the maths. Time saved = (old hours - new hours) / old hours x 100. Round it, label it, be ready to explain it.
3. Scale beats silence. If you cannot show a lift, show the size of what you touched. "Served 15M API calls at 99.95 percent uptime" tells me the scope you operate at.
4. No numbers? Instrument now. Ping product for the figure, or stand up a log counter today, so next quarter you are not guessing.
5. Quality-only backup. When a number genuinely does not exist, use a strong impact verb, but sparingly. This is the fallback, not the default.

You do not need all five on every bullet. One real number, time saved, percentage lifted, or dollars cut, already puts you ahead of most of the CVs I read.

Two worked examples

Example A: Product Manager, prop-tech SaaS

You shipped a "1-Tap Rent Pay" feature for a property-management portal. No fancy analytics in place.

| Step | What you did | Quick maths | |---|---|---| | Exhaust | Stripe dashboard shows weekly on-time rent payments | Before: 5,100/week. After: 6,900/week | | Estimate | Conversion lift | (6,900 - 5,100) / 5,100 ≈ +35% | | Scale | Size of audience | 18k tenants use the portal monthly | | Cost impact | Admin time saved chasing late rent | 1,800 fewer follow-ups x 8 min x A$35/hr ≈ A$8.4k saved/month | | Fallback | If Finance disputes the labour rate | Stick to "+35% on-time payments" |

Finished bullet: "Rolled out 1-Tap Rent Pay to 18k tenants, lifted on-time payments 35 percent, cutting admin costs roughly A$8k a month."

Example B: Software Engineer, Sydney fintech

You optimised a serverless API by adding DynamoDB caching. No previous logs.

| Step | What you did | Quick maths | |---|---|---| | Exhaust | CloudWatch logs | p95 latency 650ms to 140ms | | Estimate | Latency drop | (650 - 140) / 650 x 100 ≈ -78% | | Scale | Traffic volume | 2.4M API calls/month | | Cost avoided | Shorter Lambda run time | 120ms shaved x 2.4M calls ≈ A$1.2k saved/month | | Fallback | If the cost calc gets messy | Keep the latency percentage and the traffic scale |

Finished bullet: "Optimised serverless caching, cutting p95 latency 78 percent (650ms to 140ms) across 2.4M calls a month, trimming the AWS bill roughly A$1.2k and dropping alerts from 4 to 1 a week."

The one question to run before you submit

Before your next review or CV edit, run a single test over every bullet: if I had to prove this impact tomorrow, what one number, time-span, or scale point could I pull from logs, reports, or memory to back it up?

That question surfaces more numbers than you think you have. Hard numbers are not optional in this market. They are the difference between a recruiter believing you had impact and a recruiter guessing, and the guess rarely goes your way.

If you want bullets that already do this, the Careersy tech resume templates are built around metric-first impact lines for the ANZ market.