You've written a clean draft, the client loves the tone, and then someone asks: "Where did this number come from?" If you can't answer in ten seconds, you have a problem. Fact-checking isn't a final polish — it's a core part of the editing process that separates professionals from hobbyists. A single unverified figure can turn a routine copywriting job into a reputation crisis, and it's the most common reason freelance writing projects go sideways.
The demand for writers who can verify their own work is rising. Clients increasingly expect SEO writing that's not just keyword-rich but accurate, and editors want copy that won't expose them to legal or credibility risk. This article walks through a practical fact-checking workflow — the kind you'd build during a serious Writing and Editing programme — and shows where it breaks, what it costs, and how to learn it without wasting months on trial and error.
Why fact-checking is a career skill, not a chore
Most writers think fact-checking is about catching typos. It's not. It's about source hierarchy: knowing that a primary source (the original study, the official document, the person who said the quote) outranks a secondary source (a news article summarizing the study, a blog quoting the quote). When you cite a secondary source without checking the primary, you inherit every error the intermediary made — and you own them.
In freelance writing, your reputation is your only real asset. A client who has to correct a factual error once will review your next draft line by line. That slows the editing process, increases revisions, and quietly erodes your rates. Fact-checking is how you protect your pricing power.
A walkthrough: verifying a single claim in a client article
Imagine you're writing a piece for a fintech client. Your draft says: "The average household saves 15% of its income." Here's how a practitioner verifies it step by step.
Step 1: Identify the claim type
This is a statistical claim — a number attributed to a population. Statistical claims need a named source, a date, and a methodology. Without all three, the claim is unusable.
Step 2: Trace to the primary source
You find the number in a news article that cites "a recent survey." The news article is secondary. You follow its link to the survey — if there is one. If the link is dead or the survey isn't named, the claim fails. You cannot verify it, so you don't use it.
Suppose the survey exists. Now check: who conducted it, when, and on what sample? A survey of 500 people conducted three years ago is not the same as a national statistics office report. If the survey is outdated, you either update the number or qualify it: "According to a [year] survey of [population]..."
Step 3: Check the math and the units
Even real numbers get misreported. Does "15% of income" mean gross or net? Before or after tax? A figure that shifts meaning with a definition is a trap. You write the definition into the sentence or you drop the number.
Step 4: Verify quotes and dates
If your draft quotes a person, the quote must come from a transcript, a recording, or a published interview — not from a tweet summarizing the interview. Dates are just as fragile: "last year" in a source written in 2023 means 2022, not 2025. Convert every relative date to an absolute one.
Step 5: Run the pre-publish checklist
Before sending to the client, you check: every number has a source; every source is linked or named; every quote is verbatim; every date is absolute; every name is spelled correctly. This takes ten minutes and saves hours of revisions.
What breaks if you skip a step? If you skip Step 2, you might publish a number that a competitor can debunk in one search. If you skip Step 3, you might report a real figure with the wrong definition. If you skip Step 4, you might misquote someone — the fastest route to a public correction request.
Where the fact-checking mechanism itself breaks
No workflow is foolproof. Here are four scenarios where the standard process fails, and how to handle them.
1. The primary source disappears
You verified a claim six months ago. The source page is now offline. What goes wrong: you can't prove the claim if challenged. How to spot it early: archive every source the moment you verify it — save the page or take a screenshot with the date. If it's too late: replace the claim with a verifiable one or add a note that the original source is no longer available.
2. The client insists on an unverifiable claim
A client wants a statistic that sounds good but has no source. What goes wrong: you either comply and risk your reputation, or refuse and risk the contract. How to spot it early: ask for the source during the brief, not after the draft. If it's too late: offer a qualified alternative — "industry estimates suggest" — and explain the risk in writing.
3. The deadline makes full verification impossible
You have two hours to file. What goes wrong: you skip checks and hope. How to spot it early: triage claims by risk — numbers and quotes first, descriptive details last. If it's too late: flag the unverified claims to the editor explicitly. Editors respect honesty more than silent risk.
4. The tool you rely on stops working
You depend on a link checker or a plagiarism tool that goes down. What goes wrong: your process stalls. How to spot it early: keep a manual backup — a simple spreadsheet of sources and dates. If it's too late: do the checks manually; it's slower but reliable.
Learning paths: self-study, courses, and practice
There are three realistic ways to learn this workflow, and they differ in time, cost, and outcome.
| Path | Time | Cost | Outcome |
|---|---|---|---|
| Self-study (blogs, YouTube) | 3–6 months of scattered reading | Free | You learn concepts but lack feedback; mistakes go unnoticed until a client catches them |
| Structured course | 6–10 weeks | Moderate | Guided practice with feedback; you build a repeatable system |
| On-the-job practice | Ongoing | Free but risky | You learn by making real mistakes — costly for reputation |
Self-study is cheapest but slowest to produce competence. You can read about source hierarchy in an afternoon, but applying it under deadline pressure requires repetition and correction. A structured course compresses this by giving you exercises and feedback — the course's AI chat can review your written answers, which is the closest thing to an editor looking over your shoulder.
On-the-job practice is how many freelancers learn, but it's the most expensive path: a single factual error can cost you a client. The trade-off is real — you gain speed but pay in trust.
Who this path suits — and who it doesn't
Fact-checking and editing suit people who are detail-oriented, comfortable with research, and willing to slow down before publishing. If you enjoy the hunt for a primary source and the satisfaction of a clean, verified draft, this work fits you.
It does not suit writers who see verification as bureaucracy. If you want to write fast and loose, you'll resent the checklist — and clients will notice. It also doesn't suit those who expect instant results: building a reliable workflow takes weeks of practice, not days.
For those who want a structured way in, the Writing and Editing programme covers the editing process, fact-checking, and client work with practical exercises. It's one path among several — self-study and on-the-job learning are valid alternatives — but it's the one that gives you feedback on your decisions before a client does.
Fact-checking is the difference between a writer who is trusted and one who is replaced. Start with one claim today: pick a number in your last draft and trace it to its source. If you can't, you've found your first lesson. The Writing and Editing course walks through this workflow step by step and gives you feedback on your decisions — details and enrolment are on the course page.
You can start here: Writing and Editing.