The Interview Selling-Point Score: Headline, Support, or Hold

You have a bank of accomplishments and one target role, and you need to decide which one to build the interview around. This guide gives you a rubric for that single judgement call: score each proof point on four dimensions, name one headline differentiator, then rank everything else as support or hold. It is for anyone preparing a loop who is tired of leading with things every rival candidate also has.

Most published advice stops at "pick the relevant, quantified accomplishment." That is necessary and not sufficient. Relevant points are exactly the ones that feel safest and differentiate you least, because relevance is the bar every hired rival already cleared. What follows separates relevance from distinctiveness, then adds two dimensions the standard advice skips: whether the point survives probing, and how many prompts it can carry across the loop.

Why "relevant and quantified" is not enough

The trap is that table-stakes points feel the most relevant, which is exactly why people lead with them. A required-qualification bullet scores high on relevance and zero on distinctiveness, because everyone still in the loop met the same requirement. If you build your interview around it, you sound like a competent version of every other finalist.

Start from the posting's own structure. Required qualifications are what you need to do the job on day one. Preferred qualifications are wishlist items: they would love someone who has them, but they will hire someone who does not, and "nice to have" means genuinely optional unless two candidates are otherwise identical. So a proof point that only satisfies a required bullet is table-stakes. Distinctiveness lives in the preferred bullets and in things beyond the posting entirely.

There is a second reason relevance misleads. Many candidates are discouraged from applying by preferred qualifications they mistake for strict requirements. That means a preferred item you clearly hold may be rarer in the actual applicant pool than the posting makes it look, because the people who lacked it self-eliminated before applying. You cannot estimate that from the posting alone. You estimate it from how common the skill is among peers.

The four dimensions

Score every proof point on four dimensions before you decide what to do with it. The dimensions are relevance, distinctiveness, defensibility, and reach. Each one catches a different way a strong-sounding point fails.

• Relevance. Does the point map to something the posting asks for? A required bullet counts, but only as a floor. This is the one dimension the standard advice already covers.
• Distinctiveness. Does the point map to a preferred bullet or something beyond the posting, and is it rare in the pool the posting draws from? This is where you beat the field rather than join it.
• Defensibility. Does the point survive two probes, stated in the first person, with a number? Interviewers pick apart a story to reach the specific behaviour they want, a practice some call "digging a well." Employer-side guidance caps this at two follow-ups per question, aimed at the missing STAR component. A point that collapses under the first probe cannot carry your interview.
• Reach. How many prompts across the loop can this one point anchor? The same relevant, quantified accomplishment can serve "tell me about yourself," the greatest-accomplishment prompt, and "what sets you apart." A high-reach point is one you can deploy where it lands hardest.

Each dimension proves something specific, and each has a tell when it lies. The table below is the working version to keep open while you score.

Score each dimension on a simple three-point scale: 0, 1, or 2. Do not agonise over precision. The scores exist to force a comparison, not to produce a decimal ranking.

Scoring distinctiveness against the real pool

Distinctiveness is the dimension you cannot estimate by gut, so score it against how common the skill is among peers rather than how impressive it sounds. A skill tens of thousands of your peers list is table-stakes even if it took you years to learn. A skill only hundreds list is a candidate differentiator.

The gap is larger than most people assume. In Refolk's index of professional profiles, the commonness of "relevant" backend skills spreads across nearly two orders of magnitude.

Counts are US-based software engineers in Refolk's index; the multiples are derived by dividing each count by the Rust count. Python and Rust are both plausibly relevant to a backend role. One is listed by 55,038 people and the other by 614. If you lead with Python because it is "relevant," you have chosen the point 90 times more people can also claim.

Distinctiveness also moves with geography, not just skill. The line between table-stakes and differentiator depends on the market the posting draws from, so the same skill can be a differentiator in one place and background noise in another.

Counts are software engineers in Refolk's index; the ratio is US divided by UK. Kubernetes is roughly 6.6x scarcer in the UK, so a proof point that is merely competent in a US pool can headline in a UK one. Score distinctiveness against the specific pool, not a global impression.

Checking commonness by hand is the slow part. This is where Refolk removes friction: you can ask it how many peers in your exact market list a skill and get a count back, instead of guessing whether your edge is real. Score distinctiveness against that number rather than your feeling about how hard the skill was to earn.

The recency check runs last, and here is why

Run the recency check after scoring distinctiveness, not before, because recency degrades distinctiveness faster than it degrades relevance. A genuinely rare accomplishment from twelve years ago stays topically relevant while it quietly starts to read as decline, signalling that your edge is historical rather than current.

The published windows agree on the shape and differ on the number.

The dominant rule is a 10 to 15 year window of work history, shorter in fast-moving fields, because roles, programmes, and skills date faster there. The override is direct relevance: even an accomplishment 15 or more years old can stay if it ties straight to the target role and helps position you as a top candidate.

The sequencing matters. If you apply the recency filter first, you may cut a point that would have won on distinctiveness and never learn it. Score distinctiveness on the full bank, then let recency demote the ones that have aged out. A rare point inside the window is a headline candidate. A rare point outside it is a headline candidate only if it is directly role-critical, and otherwise a hold.

The scoring procedure

Work the bank in order. The whole pass takes about two hours the first time and far less once you have a scored bank to reuse.

For the defensibility score, apply a concrete test rather than a feeling. Rewrite the point in the first person and try to isolate what you specifically did. If the story only works as "we shipped," it fails, because interviewers probe precisely when a candidate hides inside the word "we." Then imagine two follow-up probes aimed at your weakest STAR component, usually the action or the number, and check whether you have an answer ready. A point that needs "we" or has no number scores 0 on defensibility no matter how impressive it sounds.

What to do with each point once it is scored

The scores map to three destinations: headline, support, or hold. The headline is the one point that scores high on all four dimensions. Supports are the ordered set you reach for as prompts arrive. Holds are relevant but table-stakes, or better saved for a probe. Use the matrix below to place any point you are unsure about.

Naming the headline is a commitment, not a shortlist. Highlight one or two experiences that differentiate you and stay focused on relevance. Do not chase novelty. You should not be intimidated by the word "unique"; employers want interesting and useful, not something they have never heard. Frame the point as employer benefit, answering "why hire me over others," not "what is quirky about me."

Reach decides where you spend the headline. "What sets you apart" is usually not part of the standard question set and is typically asked only of finalists, late in the loop. That is a sequencing constraint, not a footnote. If you spend your differentiator early on "tell me about yourself," you can arrive at the money question with leftovers. Map your headline's reach so you deploy it where it counts, and let high-reach supports carry the earlier, broader prompts.

One tactical note on order. Some employer-side guidance argues that reordering a STAR answer to lead with the quantified result, for example opening with "I saved 23 thousand dollars," is itself a differentiator, since it front-loads the number a probe would otherwise dig for. So order can precede selection: a merely-good point delivered result-first can outperform a better point buried behind setup.

How this goes wrong

The rubric fails in predictable ways, and each failure has a false positive that makes a weak point look strong. This section is the most valuable part of the standard, because these are the errors that survive a careless scoring pass.

• Leading with a required-bullet point. It feels relevant, so it scores high, but every hired rival cleared the same bar. Check: is the bullet in the required or preferred section? Required alone is table-stakes.
• Mistaking rare-on-paper for rare-in-pool. A preferred skill you hold may still be common among applicants, or the reverse. Check commonness against index counts, not gut feel, and against the specific market.
• Unquantified headline. Sounds impressive, collapses under "by how much?" Check: can you state a number in the first sentence? If not, demote it out of the headline slot.
• Team-credit story. "We shipped" reads fine until the probe, when you cannot isolate your action. Check: rewrite in "I"; if nothing survives, hold it.
• Stale differentiator. A genuinely rare accomplishment from twelve years ago reads as decline, not capability, especially in tech. Check it against the 10-year tech window unless it is directly role-critical.
• Over-reaching for "unique." Chasing novelty produces gimmicks; the prompt rewards employer benefit, not trivia. Check: does the point answer "why hire me over others"?
• Wasting the finalist-only prompt. Because "what sets you apart" often comes last, spending the headline earlier leaves nothing fresh. Check the reach map so the headline lands where it counts.

Once you have a scored bank, checking distinctiveness by hand across several skills and markets is the remaining slow step. Refolk turns your history into materials tailored to each posting and scores how well you fit, which is the same commonness-versus-relevance judgement applied at scale rather than one point at a time. Use it to confirm your headline is rare in the pool before you build an answer around it.

Keeping the scored bank current

A scored bank is not a one-time artefact; it drifts as your work advances and as markets shift. Re-score before each new loop rather than reusing last quarter's answer, because the two dimensions most likely to have changed are distinctiveness and recency.

Recency moves on its own: a point that sat comfortably inside the 10-year window slides toward the edge every year, and in tech it slides faster. Distinctiveness moves because the pool moves. A skill that was rare when you learned it becomes table-stakes as peers adopt it, which is why the commonness counts, not your effort, decide the score.

Re-run steps three through seven each time the target role changes, even for a near-identical posting, because the market and the required-versus-preferred split can differ. The bank of claims from step one is reusable; the scores are not. That is the difference between a standard and a script: the standard tells you how to judge, so you make the right call for the role actually in front of you.

CLAIM (I ... , resulting in [number]): ____________________
Posting tag: required / preferred / beyond-posting
Relevance (0-2): __   Distinctiveness (0-2): __   Defensibility (0-2): __   Reach (0-2): __
Pool check: skill listed by ______ peers in [market] -> table-stakes / differentiator
Recency: in-window / stale-but-role-critical / stale-drop
Destination: HEADLINE / SUPPORT (rank __) / HOLD