.png)
Most recruiting language fails at least two of these four tests, which is why most recruiting language changes nothing. Here is how to run it.
SADR is a test for evaluating any claim a company makes to candidates. A claim earns the right to appear in a job post, an outreach message, or a career site only if it is all four of these things: Specific enough to picture, Attractive to the audience you actually want, Different from what competitors can say, and Real enough to survive a candidate's diligence. Fail one and the claim is weak. Fail two and the claim is wallpaper.
Why a test at all
Because the review meeting is not one. The standard process for approving employer brand language checks whether a claim is accurate, whether it is positive, and whether Legal can live with it. Those filters produce sentences like "we offer growth opportunities in a collaborative, fast-paced environment," which is accurate, positive, legally bulletproof, and completely inert. It has been approved by everyone and believed by no one.
Candidates do not evaluate your claims against your intentions. They evaluate them against every other company's claims, their own experience of being disappointed by identical language, and whatever they can verify in fifteen minutes of searching. SADR is the review meeting the market is already running on your language. Running it first just means you get the results before you pay for distribution.
The four tests
Specific: can the candidate picture it?"Growth opportunities" fails. "Most people in this role earn their charge certification inside eighteen months, and two of the last four moved into education roles" passes. The test is whether the claim produces an image or a shrug. Vague claims are not neutral; they actively signal that either you do not know your own offer or you are hiding something. Specificity is the tax you pay to be believed, and most companies refuse to pay it.
Attractive: does the right person actually want it?Attractive to the audience you need, not to the review committee. "Fast-paced environment" was written to sound energetic; to an experienced ICU nurse it reads as understaffed, and to a senior engineer it reads as no roadmap. Plenty of claims pass Specific and fail here: "you will present to the executive team monthly" is specific and, for some audiences, a threat. You cannot run this test without knowing who the message is for, which is the quiet reason most language stays vague: vagueness lets you skip the decision about your audience.
Different: could a competitor say it too?The brutal one. Take your claim, swap in your competitor's logo, and read it again. If it still works, it is not differentiation, it is category membership. Mission-driven, great culture, people-first, competitive benefits: every one of these is true of your competitors in exactly the proportion it is true of you, which means every one of them is a sentence spent saying nothing. Different does not require being unique in the universe. It requires being distinct within the choice set your candidate is actually comparing.
Real: does it survive contact with diligence?The claim has to hold up against Glassdoor, against the interview process, against a friend-of-a-friend who works there, and against the candidate's first ninety days. A specific, attractive, different claim that turns out to be aspirational does more damage than saying nothing, because it converts your employer brand from unknown to untrusted. If a claim is not yet real, it is not a messaging asset. It is a to-do list item for the business.
Running it: one claim, four verdicts
Take a real line: "We empower our people to do their best work."
Specific: fail. Empower how? Picture nothing. Attractive: unknowable, because it is aimed at nobody in particular. Different: fail; every competitor's careers page says it verbatim. Real: unfalsifiable, which is its own kind of fail, because a claim that cannot be checked cannot build trust.
Now the rebuilt version: "Engineers here ship without a release committee. You will push to production in your first week, and you will own what you push."
Specific: pass, you can picture Tuesday. Attractive: pass for the builder you want, and usefully repellent to the engineer who wants guardrails. Different: pass if your competitors run release committees, which is checkable. Real: pass only if it is true, and if it is not true, you have just learned that your problem is not messaging.
That last sentence is the real output of SADR. Sometimes the test improves the language. Sometimes it reveals that the language was never the problem, and the honest next step belongs to the business, not the brand.
Where to use it
Run SADR on every sentence of your highest-priority job post, your standard outreach template, and your career site's first screen. Anything that fails two or more tests gets rewritten or cut. The fastest version of this exercise has its own name and takes ten minutes: [the red pen test]. The most rigorous version puts the surviving claims in front of a synthetic panel of your actual audience: [The Panel].
Every claim a candidate reads is a small promise about what working for you is like. SADR just asks whether the promise is worth making.
