Read enough "how to hire a copywriter" guides and you'll notice they all converge on the same advice: check the portfolio, ask for case studies, look for results. Good advice. Also incomplete, because none of them tell you what to actually do once the case study is sitting in front of you.
Here's the problem nobody names out loud: a screenshot of a graph going up and to the right proves almost nothing on its own. It could be real. It could be someone else's result with the serial numbers filed off. It could be a metric that has nothing to do with the copy. You have no way to tell — unless you know what to ask next.
The Advice Everyone Gives, and Where It Stops
"Ask for a portfolio." Fine — but a portfolio is a list of finished pages, and finished pages tell you someone can write. They don't tell you whether the writing did anything. "Ask for case studies showing results." Better — but a case study is a story someone chose to tell about themselves, edited for the best possible version of events. Neither one, on its own, separates a strategist from someone who got lucky once and built a slide deck about it.
Five Questions That Actually Verify a Result
These take two minutes to ask and are uncomfortable for anyone who's exaggerating. That discomfort is the point.
1. Can I see the actual page, not just a screenshot of a chart?
A real result lives somewhere real — a live URL, an archived version, an actual email that went out. A number floating in a slide with no page attached to it is a claim, not evidence.
2. Is there a before and an after, or just an after?
"Conversion rate: 4.2%" means nothing without knowing what it replaced. A page that went from 0.5% to 4.2% is a real story. A page that's just at 4.2%, with no baseline mentioned, could've been performing that well before the copywriter ever touched it.
3. Does the number have a timeframe attached?
"Increased sales" over what window — a week, a quarter, a year with three other changes happening alongside it? Vague timeframes are how unrelated growth gets credited to a rewrite.
4. Can they explain what changed and why, in their own words, without notes?
Someone who actually did the strategic thinking can talk through their own case study conversationally — what the research found, what they changed, why they expected it to work. Someone reciting a client's summary tends to sound like they're reading it, because they are.
5. Is there a name attached, or is it "a SaaS company" forever?
Anonymized case studies are sometimes legitimate — NDAs are real. But when every case study is anonymous, with no verifiable detail anywhere, that pattern itself is worth noticing.
This gap — "I don't know what I'm actually paying for until it's already failed" — shows up consistently in research on why buyers hesitate to hire a copywriter. It's not paranoia. There's currently no standard way to verify a claimed result before you've already paid for one yourself.
What a Legitimate Case Study Actually Looks Like
Specific, not impressive-sounding. "We rewrote the pricing page headline based on a repeated theme in cancellation surveys — free-trial users said they didn't understand what happened after the trial ended. Rewrote the headline to answer that directly. Trial-to-paid conversion moved from 11% to 15% over the following six weeks, no other changes to the page." That's boring to read and hard to fake, which is exactly why it's convincing. Compare that to "Wrote copy that increased conversions by 340%" with nothing else attached — impressive-sounding and unfalsifiable, which is exactly the problem.
When There's No Portfolio to Verify
Sometimes the honest answer is that a copywriter is newer and doesn't have a results-based portfolio yet — everyone starts somewhere, and a thin track record isn't automatically a red flag. What matters at that point is whether they're upfront about it instead of padding the gap with vague claims. A newer copywriter who says plainly "I don't have client results to show you yet, here's my process instead, and here's a free sample of how I'd approach your actual page" is giving you something you can evaluate. One who invents a results screenshot to compensate is giving you the exact problem this article is about.
Either way, the fix is the same: don't evaluate claims. Evaluate what you can check.