📊 Numbers, Data & Metrics
17 exercise sets. Read SLA percentages, performance benchmarks, version numbers, CI pipeline output, sprint data, and business metrics — then use them in sentences.
SLA Uptime Reference — Quick Guide
These figures appear in SLAs and reliability engineering. Know how to read them and say them aloud.
| Uptime % | Allowed Downtime (per year) | How to say it |
|---|---|---|
| 99% | 3.65 days / year | "ninety-nine percent" |
| 99.9% | 8.77 hours / year | "three nines" |
| 99.95% | 4.38 hours / year | "three-and-a-half nines" |
| 99.99% | 52.6 minutes / year | "four nines" |
| 99.999% | 5.26 minutes / year | "five nines" |
- Intermediate
Performance Metrics
Read and talk about p99 latency, throughput, error rates, and response times. "Our p95 is 120ms, but p99 spikes to 800ms under load."
- Intermediate
SLA & Uptime Percentages
"Five nines" (99.999%), calculating downtime budgets, expressing error budget burn rates. Convert percentages to allowed downtime.
- Beginner
Version Numbers & Semver
Reading version numbers like v3.14.2, major/minor/patch meaning, pre-release tags (alpha, beta, rc), and how to say them aloud.
- Beginner
Data Sizes & Units
Bytes, kilobytes, megabytes, gigabytes, terabytes, petabytes — and the difference between MB and MiB. Reading data specs correctly.
- Intermediate
Benchmark Data & Comparisons
"40% faster", "3x throughput improvement", "reduced memory usage by 512MB" — comparative figures in technical English.
- Beginner
Time, Dates & Timestamps
Unix timestamps, UTC, ISO 8601 dates, duration expressions ("within 30 minutes", "over the past 7 days"), timezone references.
- Intermediate
Reading Service Status Updates
Read real-style incident reports, status pages, and maintenance notices. Extract duration, impact, root cause, and next steps.
- Advanced
Writing SLA Language
Choose and evaluate SLA clauses — uptime tiers, incident response times, service credits, and professional support tier descriptions.
- Intermediate
Core Web Vitals
LCP, CLS, INP, FCP, TBT, and Lighthouse scores — read performance reports and communicate results to clients and stakeholders.
- Intermediate
CI Pipeline Summaries
Read build output, discuss failure rates, coverage thresholds (exit codes, test counts, flaky tests) — essential for daily CI/CD work.
- Intermediate
Sprint Burndowns
Read sprint charts and velocity data, discuss scope creep, completion rates, and rollover with the vocabulary used in standups and retros.
- Intermediate
Business Metrics
DAU/MAU ratios, conversion funnels, churn rate, MRR, p99 latency — understand and present product and API metrics to stakeholders.
- Intermediate
Code Coverage Numbers
Read coverage reports, discuss threshold breaches, understand line vs branch coverage, and talk about test debt in professional English.
- Intermediate
CloudWatch & Grafana Dashboards
Read monitoring dashboards and describe what you see: latency spikes, throughput drops, throttle events, concurrency limits — the language of observability.
- Intermediate
Data Pipeline Numbers
Describe pipeline capacity in English: events per second, GB per hour, consumer lag, SLA compliance, and batch job performance metrics.
- Intermediate
Salary & Compensation
Practise discussing compensation professionally: understanding total comp, RSU vesting, comparing offers, stating salary expectations, and negotiating a raise.
- Advanced
Performance Benchmarks
Read and discuss performance benchmark reports: latency percentiles, throughput comparisons, database benchmarks, frontend performance, and load-testing vocabulary.
Frequently Asked Questions
Why do IT professionals need to study numbers and data language in English?
Numbers appear constantly in technical communication: performance metrics (p99 latency 450ms), SLA percentages (99.9% uptime), data sizes (200GB database), version numbers (v2.4.1), code coverage percentages (87% coverage), sprint velocity (42 story points). Misreading or misphrasing these in reports, incidents, or client presentations leads to real misunderstandings.
How do I say percentages and uptime figures in English?
Percentage phrasing: "99.9% availability" = "three nines" (industry shorthand), "99.99%" = "four nines", "99.999%" = "five nines". In sentences: "The service maintained 99.95% availability last quarter" or "We achieved four-nines uptime in December." SLA violations: "We fell below our 99.9% SLA by 47 minutes this month."
How do I read and say version numbers correctly?
Version numbers follow semantic versioning: MAJOR.MINOR.PATCH — "version 3.12.4" is read as "three dot twelve dot four" or "version three, twelve, four." Major version increments signal breaking changes: "We're upgrading from v2 to v3." Pre-release versions: "v4.0.0-beta.1" = "version four zero zero beta one." In conversations: "We're on version twelve of the API."
How do I describe performance metrics in English?
Performance metric vocabulary: "p50 latency is 120ms" (50th percentile), "p99 latency spikes to 850ms under load", "we're seeing 3,400 requests per second at peak", "throughput is 1.2 GB/s", "CPU utilisation is at 78%". In reports: "Response time degraded from 200ms to 450ms following the deployment, with p99 reaching 1.2 seconds."
What is the difference between KB and KiB in data size?
KB (kilobyte) = 1,000 bytes (SI standard). KiB (kibibyte) = 1,024 bytes (binary standard). This matters: "1 GB hard drive" ≠ "1 GB when formatted" because hard drive manufacturers use SI (1,000) while operating systems often use binary (1,024). In precise technical documentation, use KiB/MiB/GiB for binary values. In informal speech, "gigabyte" usually means whichever fits the context.
How do I describe code coverage and quality metrics in English?
Coverage language: "We have 87% test coverage", "Line coverage is at 92%, but branch coverage is only 74%", "The coverage report shows 3 uncovered code paths in the authentication module." Quality metrics: "Cyclomatic complexity of 15 — should refactor", "Technical debt is estimated at 40 developer-days", "DORA metrics show a deployment frequency of 3.2 per week."
How do I discuss sprint velocity and burndown in English?
Agile metrics language: "Our velocity over the last 5 sprints averages 38 story points", "The burndown shows we're on track to complete 42 of 48 points by end of sprint", "We're carrying 12 points of scope creep into next sprint", "Our cycle time has improved from 4.2 to 2.8 days per ticket." In retrospectives: "Our velocity dropped due to two unplanned incidents."
How do I describe timestamps and timezones in incident reports?
Incident time language: "The incident began at 14:32 UTC on 15 July 2024", "Detection to resolution took 47 minutes", "The deployment window closed at 22:00 BST (21:00 UTC)", "All timestamps in this report are in UTC." In incident bridges: "What time is it at your end?" "It's 15:45 CET here, which is 14:45 UTC."
How do I present benchmark results professionally?
Benchmark language: "Benchmark results show a 23% improvement in query time, from 2.3s to 1.8s", "Results are normalised against the reference implementation", "The optimised version outperforms the baseline by 3.4× on average across 10 runs", "These results were obtained on a t3.medium instance — production hardware may differ." Always state hardware, test conditions, and caveats.
What data transfer speed vocabulary should I know?
Speed vocabulary: Mbps = megabits per second (network bandwidth), MB/s = megabytes per second (file transfer), Gbps = gigabits per second (high-speed networking), GB/s = gigabytes per second (storage). "Our API response size averages 1.4MB and returns in 380ms — that's about 3.7 MB/s effective throughput." Distinguish bits (Mbps) from bytes (MB/s) — factor of 8 difference.