Our tech career data: sources and methodology
By the RoleMath Editorial Team · Last updated 2026-06-14. Every figure traces to a cited source; we sell none of the options discussed. Draft pending human review.
Our tech career data comes from a small set of official, public sources and nothing else: wages from BLS OEWS (May 2025), growth and openings from BLS Employment Projections (2024–2034), job characteristics and skills from O*NET, and certification facts only from official vendor pages — and we sell nothing and currently earn no referral on the paths we describe. Most career sites cite nothing, sell something, or both. We do neither — every figure on RoleMath traces to an official source, and we currently earn no referral or affiliate fee on the paths we describe. This page is the tech career data behind the site: where the numbers come from, how we read them, the caveats we keep visible, and — just as important — the claims we refuse to make.
Key takeaways
- Every data figure traces to an official public source — BLS OEWS (May 2025) for wages, BLS Employment Projections (2024–2034) for outlook, O*NET for job characteristics, and vendors' own pages for certification facts.
- We show the full wage percentile range and foreground the 10th–25th lower band as an early-career proxy — not the median that other sites quote as if it were a starting salary.
- Roles are mapped to BLS occupations, sometimes loosely; we flag approximate mappings and present wages as occupation-level context, not title-specific pay.
- We refuse to attach salary/ROI/pass-rate/placement/guarantee claims to certifications, name a single 'best' for everyone, or quote unsourced demand stats — and we currently earn no affiliate fee (any such link would be labeled inline and never affect a ranking).
- Our sources go beyond labor data: we also cite official government funding and eligibility (GI Bill, WIOA, Workforce Pell, IRS Section 127), published training and exam costs, and education-outcome data (NCES, College Board, CIRR) — the cost-and-funding reality most career sites leave out.
Where our numbers come from
We use a small set of official, public sources and nothing else for our data. Wages — including the full percentile distribution and the metro and state breakdowns — come from the U.S. Bureau of Labor Statistics Occupational Employment and Wage Statistics (BLS OEWS), May 2025 release. Growth outlook and annual openings come from the BLS Employment Projections, 2024–2034. Job characteristics, skills, and the job-zone preparation level come from O*NET, the U.S. Department of Labor's occupational database. Certification facts — exam codes, objectives, eligibility — come only from the official vendor or certifying body's own pages. We do not use self-reported salary aggregators, scraped job-board figures, or any source we cannot name and date. Why official data and not the salary aggregators most sites quote? Self-reported sites — the ones that ask users to submit their own pay — suffer from selection bias (who bothers to report skews the result) and carry no occupational discipline, so one inflated title can move the average. BLS surveys employers, not volunteers, and classifies by occupation. We are honest about its limits, too: BLS data lags (releases are roughly annual), it is occupation-level rather than title-level, and small samples can be suppressed — which is exactly why we show percentile ranges and a vintage instead of a single confident number.
The full source list, by category
Here is the complete picture of what we cite, grouped by what it answers. Every source is official, public, and named with a date — no aggregators, no scraped figures.
Labor and wages: the U.S. Bureau of Labor Statistics (OEWS for wages and percentiles, the Employment Projections for outlook and openings, the Occupational Outlook Handbook for context) and O*NET, the U.S. Department of Labor's occupational database, for skills, tasks, interests, and the job-zone preparation level.
Credential facts: only the certifying body's own pages — CompTIA, Cisco, AWS, Microsoft, Google, EC-Council, ISC2, ISACA, Red Hat, PMI, the Linux Foundation, GIAC, the Cloud Native Computing Foundation — for exam codes, objectives, prerequisites, and eligibility. We model a proctored certification and a course-completion certificate as different things.
Cost and pricing: published exam fees from each vendor and published training list prices from providers (plus federal GSA Schedule pricing where it exists), so we can show the total cost of a path — exam, training, and renewal — not just a headline.
Funding and eligibility: the official government programs most sites never mention — the VA (GI Bill certification-test reimbursement), the Department of Labor's CareerOneStop (WIOA training funds), Federal Student Aid (the new Workforce Pell rule for short-term programs), and the IRS (employer Section 127 educational assistance). Eligibility for several is decided locally and is never guaranteed, which we say plainly.
Education outcomes: the National Center for Education Statistics (earnings by educational attainment, graduation rates), College Board (published tuition), and CIRR (the only third-party-audited bootcamp outcomes) — with the honest note that most bootcamp placement claims are self-reported, not audited.
Definitions: NIST (its computer-security glossary and the canonical cloud-computing definition) and authoritative vendor explainers, so a term means what the standards body says it means.
How we read wage percentiles — and why the lower band
A single median is the statistic most career sites quote and the one most likely to mislead a career changer. The median is the midpoint across all workers in an occupation, experienced and new alike, so we show the full percentile range and foreground the 10th–25th percentile as a realistic early-career band. Important caveat we keep visible: these are cross-sectional percentiles across all current workers, not a literal entry-versus-experienced split — so we treat the lower band as a proxy for early-career pay, not a promise, and note that your own background can place you higher or lower.
The occupation-mapping caveat
Job titles and BLS occupations don't line up one-to-one. Each role we describe is mapped to a BLS Standard Occupational Classification (SOC) code, sometimes loosely — for example, "data analyst" maps to the BLS Data Scientists occupation, which skews toward more senior, higher-paid work, so those figures read high for an entry data job. Where a mapping is approximate we say so, we present wages as occupation-level context for the broader occupation rather than the specific title, and roles that share an occupation will show identical numbers because they draw on the same BLS row. Here is the full chain worked end to end for one role. We map data analyst to the BLS Data Scientists occupation (SOC 15-2051). From BLS OEWS (May 2025) we take that occupation's national wage percentiles — a 10th percentile of $67,240 and a median of $120,230 — and from the BLS Employment Projections (2024–2034) its typical entry education, a bachelor's degree. A data-analyst page therefore shows that early-career band and median as occupation-level context, flagged with the caveat that Data Scientists skews more senior than a typical entry data-analyst job, so the figures run high. Every number on the page traces back to those two cited BLS releases.
What we cite that competitors don't
Most career sites — including the big ones — share three blind spots, and our source list is built to fill them. First, cost of ownership: almost no competitor publishes the real exam fee, training cost, and renewal cost together, because the sellers among them would rather not. We cite each from the vendor's own page. Second, funding and eligibility: the GI Bill, WIOA, Workforce Pell, and employer Section 127 can cover much or all of a path's cost, yet you'll rarely see them named — we cite the official government program for each. Third, an honest cheapest-and-fastest read: because we sell nothing and cite outcomes data (including the audited bootcamp outcomes most providers won't submit to), we can rank paths by real cost and time instead of by what pays us. None of this requires a number we can't source; it just requires having no reason to hide it.
What we will not claim
Discipline is mostly about what we refuse to say. We do not attach a salary, return-on-investment, pass rate, placement rate, or job guarantee to any certification — credential facts and occupation wages are kept strictly separate. We do not declare a single "best" certification, role, or path for everyone, because the right answer depends on your situation. We do not invent demand or quote unsourced "X% of employers" statistics. We treat course-completion certificates and proctored certifications as different things. And every figure stays occupation-level context, never a personal earnings promise. If a page ever contained an affiliate or referral link, it would be labeled inline and would never influence a rating or ranking.
How current the data is
We label the vintage of every dataset so you can judge its freshness: wages are the BLS OEWS May 2025 release, and outlook and openings are the BLS Employment Projections covering 2024–2034. Official labor data is updated on a fixed government schedule — OEWS roughly annually — and we re-verify volatile facts (wage releases, exam prices, exam names, retirement status) against the latest official release as each one lands — OEWS annually each May, the Employment Projections on their annual cycle, and vendor exam facts whenever a vendor revises its page — and stamp each page with its vintage, rather than leaving undated numbers in place. Certification facts are checked against the vendor's current page. Where a figure can change, the page shows when it was last updated. Different sources refresh on different clocks, and we track each: BLS OEWS is annual (each May), the Employment Projections annual, O*NET updates quarterly with its main release in Q3, vendor credential and pricing pages are checked whenever the vendor revises them, and government funding rules are re-verified as they change — Workforce Pell, for instance, takes effect July 1, 2026. Each page stamps the vintage of the data it shows.
Frequently asked questions
What sources does RoleMath use for its data?
Wages and percentiles come from BLS OEWS (May 2025), including metro and state breakdowns; growth and openings from the BLS Employment Projections (2024–2034); job characteristics and skills from O*NET; and certification facts only from the official vendor or certifying body. We don't use self-reported salary aggregators or scraped job-board data.
Why do you show occupation-level data instead of exact role salaries?
Official wage data is published by BLS occupation, not by every job title, so a single title may map to a broader occupation. We present those figures as occupation-level context and flag where a mapping is approximate, rather than implying a precise per-title or per-certification salary, which the data can't support.
Do you sell certifications or earn affiliate commissions?
No. We sell no course, bootcamp, or certification, and we currently earn no referral or affiliate fee on the paths we describe; if a page ever contained such a link it would be labeled inline and never affect a ranking. That independence is why we can show the figures sellers omit — like the realistic lower-percentile starting band and the roles projected to decline.
How current is the data?
We label every dataset's vintage — currently BLS OEWS May 2025 for wages and BLS Employment Projections 2024–2034 for outlook — and re-verify volatile facts against the latest official release on a cadence, stamping each page rather than leaving undated numbers in place. Pages show when a figure was last updated.
Why don't you just tell me the best option?
Because there isn't one best for everyone, and any site that claims otherwise is usually selling it. The right role, certification, and path depend on your background, budget, location, and interests. We give you the cited numbers and honest trade-offs so you can decide, and we flag the traps to avoid.
Where do your cost and funding numbers come from?
Cost figures come from each vendor's published exam fee and each training provider's published list price (plus federal GSA Schedule pricing where it exists), so we can show the full cost of a path — exam, training, and renewal — not just a headline. Funding comes from the official government programs: the VA for GI Bill certification-test reimbursement, the Department of Labor's CareerOneStop for WIOA training funds, Federal Student Aid for the new Workforce Pell rule, and the IRS for employer Section 127 assistance. Eligibility for several of these is decided locally and is never guaranteed, which we say plainly.
How do you tell a certification from a course certificate?
We model them as different things, because conflating them is a common way sites oversell. A certification — CompTIA A+, Cisco CCNA, an AWS or ISC2 credential — is earned by passing a proctored exam against published objectives. A course-completion certificate — many Google, Coursera, or vendor 'academy' offerings — is awarded for finishing a course, with no proctored exam. Both can be worth doing; they are not the same signal to an employer, and we label which is which.
Related, with the cited detail
- How much do tech jobs actually pay?
- Tech salary by location
- Entry-level tech jobs compared
- Compare entry paths
- How to read a bootcamp outcomes report
- Start the RoleMath planner
Sources
Figures in this article trace to official sources — BLS OEWS (May 2025) and Employment Projections (2024–2034), O*NET, and OEM certification pages — named where they appear or on the cited page each links to. This page stays draft_noindex pending human citation review.
Citation Ledger
| ID | Supports | Evidence | Source |
|---|---|---|---|
| CIT-01 | Visible figures and claims | Official sources (BLS OEWS May 2025; BLS Employment Projections 2024–2034; O*NET; OEM certification pages) | Named inline and on each linked cited page |