The Free-Text Angle, Scored Per Posting to Lead, Support, or Cut

You are staring at a character-limited "why you / why this role" box, and you have five selling points but room for one, maybe two. This guide is for job seekers filling in application free-text and supplemental fields, and it gives you a scoring model to decide, for the posting in front of you, which selling point leads, which supports, and which you cut. It applies to the prompt and the posting on your screen right now, not to a generic template.

This is a different call from the cover letter, where you have a page, and from the spoken self-pitch, where you can read the room and adjust. Here the field is screened, length-capped, and read in seconds. The scarcity is real, so the decision has to be deliberate.

Why the lead angle decides the field

The angle you put in sentence one is the angle that gets read. In Refolk's index there are about 110,700 US recruiter profiles, and the average opening now pulls in over 300 candidates, roughly triple the 2021 baseline. At 6 to 7 seconds of scanning per application, only sentence one is reliably read, which means a mis-ordered angle is effectively a cut angle.

That is the whole reason this decision matters. If reviewers read everything, order would be cosmetic. They do not, so order is the decision. The free-text box is not an essay you fill; it is a single shot at the front, followed by whatever survives the cap.

Employers are explicit about what these fields do. Public HR sources describe supplemental questions as a way to assess specialized experience and writing skill beyond the minimum qualifications, and as a ranking gate: responses are used to determine whether an applicant qualifies. A staffing authority frames them as separating the field. So the box is doing two jobs at once: proving you clear the bar, and sorting you against everyone else who also clears it. Your lead has to serve both.

What the fields actually assess

Application free-text fields test relevance, evidence, and differentiation, in that order of weight. Relevance gets you counted, evidence makes the claim survive scrutiny, and differentiation breaks the tie when many applicants post the same keyword match.

Three signals separate a lead from a support from a cut. First, requirement overlap: the point must map to what the posting asks for, and the posting tells you what matters most by repetition. Terms that appear two or more times in a job description are the highest-priority keywords, and the order of listed responsibilities shows the employer's priorities. Second, evidence strength: you might have experience with fifteen programs, but if you can only tell compelling results stories for three, those three are your only real leads. Third, differentiation: when many candidates clear the same bar, the documented multiplier is relationship-based, with referred candidates 11 times more likely to be hired than inbound applicants.

Above all sits the must-have gate. For skills stated as "must have," "required," or "minimum qualifications," if you lack them, tailoring will not help. A lead that hits a nice-to-have while your answer never addresses a must-have is a wasted lead.

Length changes the call before content does

Field length decides how many angles you can run before you even look at content. Under about 150 words or a single line, you have room for one angle only; above that, you can carry a lead plus support plus a detail. Get this wrong and the strongest content still fails, because a longer opening buries the qualification that should grab attention, and an over-stuffed short field truncates silently at the cap.

The mechanism is a space-versus-evidence tradeoff. At very short lengths, listing several points dilutes all of them, so you commit to the single angle with the strongest proof. With more room, you lead first, then support, then anchor to one specific verifiable detail about the employer.

Match the field type to the strategy before you write a word.

These ranges come from platform practice and adjacent domains, not a single published standard. A government HR office says a sufficient supplemental answer is typically 3 to 4 well-developed sentences; practitioner guidance for "why us" boxes targets 70 to 140 words. Concrete caps observed run to 1500 characters for a single short-answer field. When your portal states a cap, that number overrides the table. When it does not, treat the field's visible size as the cap and write to the low end.

Score each angle: the four dimensions

Score every candidate selling point on four dimensions, then let the composite decide its label. This turns a gut call into a repeatable one, so the same posting and the same points produce the same answer twice.

The four dimensions, each scored 0 to 3:

• Requirement overlap. Does the point map to a posting requirement, and how high does that requirement rank? A term repeated two or more times, or listed near the top of the responsibilities, scores 3. A single mention scores 2. An adjacent-but-implied match scores 1. No match scores 0.
• Evidence strength. Can you attach a verifiable result? A quantified outcome with scope scores 3. A concrete but unquantified result scores 2. A claim without proof scores 1. Nothing behind it scores 0.
• Differentiation. Could 250 other applicants write this same line? A referral, unique access, or a distinctive result scores 3. Something uncommon scores 2. Generic-but-true scores 1. A cliche scores 0.
• Must-have status. This is a gate, not a slider. If the point satisfies a stated must-have, mark it; if the posting has a must-have that none of your points cover, that is a red flag on the whole application, not just this point.

Add the first three for a composite of 0 to 9. Optionally validate the overall picture with a match-rate checker, but treat the number as a signal, not a verdict. Thresholds disagree across published tools.

A "pass" on one tool is a "weak" on another. Use the tool to confirm you have not missed a whole cluster of the posting's language, then trust your dimension scores over any single percentage. The raw list of terms that appear two or more times is the most reliable input you have, because it comes straight from the posting rather than from a model of it.

The procedure

Run these seven steps against the box in front of you. The whole pass takes under an hour once you have a base list of selling points; tailoring from a base typically takes 10 to 20 minutes, and in one test across 2,840 resumes a score swung an average of 36 points after a 12-minute tailoring pass. The gain is real and cheap.

Some sources say to preview these questions first and answer the largest field first, then repurpose downward. Others insist each field be tailored per posting. Both are right in part: draft once from your highest-limit field, but retailor the lead for every posting. Never paste an unedited answer across applications.

How this goes wrong

Most bad free-text answers fail in one of eight predictable ways. The scoring model catches most of them, but only if you check for them explicitly before you submit. This is the part worth slowing down for.

Two of these deserve extra weight because they masquerade as strengths.

Keyword-stuffing produces a high match-score and a confident feeling, and it is a false positive. Both applicant tracking systems and recruiters flag unnatural repetition. A keyword only counts when it is load-bearing inside a result. "Led SQL reporting" is stuffing; "cut close time 40% by rebuilding the SQL reporting pipeline" is evidence. Over 97% of Fortune 500 companies and 75% of mid-size employers use an applicant tracking system, so the keyword has to survive both the machine pass and the human read, and stuffing fails the second.

Reusing an essay across postings is the most seductive trap because the repurpose shortcut is genuinely efficient. The false positive is a fluent, polished answer that never names this employer's priorities. The check is one question: does the text cite a requirement unique to this posting? If you could paste it into another application unchanged, it is not tailored, it is recycled.

There is a geographic wrinkle worth knowing. Screener supply is lopsided: Refolk's index shows about 110,700 US recruiter profiles versus about 7,202 in the UK, a roughly 15.4x gap. A thinner reviewer pool per posting raises the odds that an automated screen, not a human, makes the first cut, which reinforces the case for a keyword-anchored lead in markets with fewer recruiters per opening.

Assembling the lead, support, and cut

Once each point has a label, the writing is mechanical. Lead with the highest-composite point that clears a must-have, phrased as a result. Add support only if the field allows it. Cut everything else for this posting, and keep the cuts in a bank for postings where they will score higher.

Here is a scoring sheet you can copy for any box.

Posting: __________  Field cap: ____ words / ____ chars  Type: single-line / paragraph

Selling point | Overlap (0-3) | Evidence (0-3) | Differentiation (0-3) | Hits must-have? | Composite | Label
1. ________________ | __ | __ | __ | Y/N | __ | lead/support/cut
2. ________________ | __ | __ | __ | Y/N | __ | lead/support/cut
3. ________________ | __ | __ | __ | Y/N | __ | lead/support/cut
4. ________________ | __ | __ | __ | Y/N | __ | lead/support/cut

Unaddressed must-haves in this posting: ________________

And a skeleton for the two field sizes that cover most applications.

[Single-line / short, one angle]
[Result that matches the posting's top requirement], which is why [this role] fits: [one specific detail about the employer].

[Paragraph, lead + support + detail]
Sentence 1 (lead): [Your strongest matching result, quantified].
Sentence 2 (support): [A second matching result or scope that reinforces the lead].
Sentence 3 (employer detail): [One specific, verifiable thing about this company or team that maps to your points].
Sentence 4 (close): [How your matching achievements meet this role's stated responsibilities].

Refolk can carry the friction that makes this slow at scale: it writes your resume from your own history, tailors your materials to each posting, drafts the cover letter, and scores how well you actually fit. When you are filling a dozen boxes a week, that fit score is the fastest way to catch a lead that scores strong for you but weak against the posting. You can start from your own history at Refolk.

Knowing who screens your answers also sharpens the lead. If a role is high-volume and machine-screened first, weight requirement overlap; if a human recruiter reads early, weight differentiation. Refolk lets you find and read the people who screen a given pipeline.

Keeping the model current before each batch

Re-run the ranking step for every posting, because the requirement frequencies change even between near-identical roles, and that is the input that sets your lead. The scores are cheap to recompute and expensive to skip.

Two things drift and are worth re-checking rather than trusting old values. First, competition volume: greenhouse-style data showed 244 applications per job, up from 116 a few years earlier, and Ashby data puts the average opening over 300, so the pressure on your lead only rises. Re-check whether the role you want is entry-level, where postings often see 400 or more, and weight your keyword lead harder if so. Second, match-tool thresholds move, and they already disagree at 60 versus 80 percent, so re-verify what "strong" means on whatever tool you use before you trust its verdict.

Run this checklist before you call any box done.

The skill this model builds is cutting. Most applicants think they have more usable leads than they do, because evidence strength caps how many angles they can honestly run. When you focus on the three points you can actually prove and lead with the one the posting asks for most, the box stops being a blank you dread and becomes a decision you can make in ten minutes.