The Field-Switch Level Call, Scored to Apply Lateral, Step Down, or Bridge

This guide is for anyone changing fields, returning after a break, or taking a lower-level role after a layoff who has a posting open and needs to decide which level to apply at. It turns your work history into the countable years a screener will actually use, scores five dimensions, and gives one verdict: apply lateral, step down one level, go entry, or take a bridge role first. It also sets out when a second application at a different level at the same employer is safe.

What the field-switch level call decides

The level call picks the level you apply at for one specific posting: your current seniority, one level down, entry level, or none of them yet. You make it posting by posting because the same history counts differently against each posting's wording, and the wording is what the screener enforces.

Terms used throughout:

• Lateral: applying at the same seniority as your last role.
• Step down: applying at the level directly below your last role.
• Entry: the first professional level in the new field.
• Bridge role: a role that changes only your function or only your industry, so your old years keep counting while you build years in the other.
• Countable years: the part of your history that meets a requirement as worded, after pro-rating for hours and duties.
• Knockout screener: a required application question that rejects you automatically if your answer misses the threshold.

The split that matters most is between two kinds of switch. An industry change keeps your function: a product manager in retail becomes a product manager in fintech. A function change changes what you do all day, whether or not the industry also changes: an accountant becomes a data analyst. The two behave differently on both pay and years.

On pay, the UK Office for National Statistics (ONS) found that job changers who moved occupation averaged 12.4% earnings growth, against 8% for those who stayed in the same occupation. Changing industry added a premium of 2.1 percentage points. On years, the function change is the expensive one, because most of your old years fail a function-specific screener. So changing function pays more and costs more countable time. A US study covering three decades adds a caution: moving to an occupation with 10% higher average wages raised a worker's own wage by only 3.1%. You do not inherit the new field's average pay.

This is not a niche problem. Pew found that in an average month across its study window, 49.5% of workers who changed employers also changed occupations. What is missing is outcome data. No published source compares time-to-hire or level-at-hire for people who changed function with people who changed only industry. So the framework below works from how screeners and recruiters count, not from outcome benchmarks, and it flags where that matters.

How do you count relevant years of experience in a career change?

For each period of your history, multiply the months by hours per week divided by 40, then by the share of time you spent on duties the posting asks for, and count overlapping periods once. This is the US federal method. It is the only detailed counting rule anyone has published. No private employer has published an HR policy on pro-rating or on credit for years in adjacent functions.

That last point is the honest limit. A private recruiter is not bound by federal rules. But a private-sector screener asks one question, usually "How many years of experience do you have with [skill]?", and accepts whatever number you type. The federal method gives you a number you can defend line by line when a recruiter asks how you got it.

The crediting rules from the Office of Personnel Management (OPM) and the US Geological Survey (USGS):

• Credit is based on the normal work week and work year. Time below that norm is credited in proportion.
• "A year at 20 hours per week is credited as 6 months of experience."
• Jobs with mixed duties are credited by the share of time spent on qualifying duties: 2 years at 75% budget work equals 18 months.
• Experience with little or no pay earns the same credit as comparable paid experience, so volunteer work counts if the duties match.
• Concurrent jobs do not stack: two clerical jobs of over 6 months each inside one 12-month period earn only one year.
• Work-study credit needs 12 months, with at least 320 hours at the next lower grade.

A worked example using only the published cases: 2 years at 75% on qualifying duties gives 18 months, and 3 years at 20 hours a week gives another 18 months. If those periods did not overlap, the countable figure is 36 months, or 3 years. If they overlapped, you count the shared calendar time once, and the figure drops. To check your arithmetic, divide total qualifying hours by 2,080, the hours in a full-time year.

REQUIREMENTS FROM THE POSTING
R1: "__ years in __"   type: function / industry   status: required / preferred
R2: "__ years in __"   type: function / industry   status: required / preferred

MY HISTORY
Period | Months | Hours per week | Share on R1 duties | Countable months for R1
__     | __     | __             | __                 | months x (hours / 40) x share
__     | __     | __             | __                 | months x (hours / 40) x share

Overlap check: concurrent periods counted once (yes / no)
Hours check: total qualifying hours / 2,080 = __ years
Countable figure for R1: __ years
Countable figure for R2: __ years
Screener answer I will type: __ (must match the figure above)

The figure only helps if your resume backs it up. A resume that lists job titles and dates but not hours or duty share makes your pro-rated answer look like a typo. Refolk writes the resume from your own history and tailors it to each posting, so the details behind your number stay visible on the page the recruiter checks.

A screener cannot pro-rate. You can, and the number you type has to survive the resume behind it.

The five dimensions and how to score each one

Score the posting at your current level on five dimensions, 0 to 2 points each: change type, countable function years, countable industry years, level equivalence, and knockout exposure. The total runs from 0 to 10, and knockout exposure also works as a hard gate.

The thresholds are my calibration, not a published standard. Below, I explain what each score proves and how it can mislead, so you can overrule it when the posting gives you reason to.

1. Change type

Proves: how much of your history can count at all. An industry-only switch keeps your function years intact. A function-only switch keeps your industry years. Changing both leaves you few countable years in either. Misleads when you call a switch industry-only because the titles match but the duties do not. "Analyst" can mean different work in two industries. Classify by duties, not titles.

2. Countable function years

Proves: you clear the "X years in [function]" line as worded. If the employer also posts the level below, use that posting's figure for the 1-point threshold. If it does not, treat "within one year of the requirement" as a working stand-in. No public data breaks down years asked by level for associate or mid-level postings, so this stand-in is a judgement call. Misleads when you entered calendar years instead of the pro-rated figure.

3. Countable industry years

Proves: you clear a separate industry test. LinkedIn treats Work Experience and Industry Experience as separate screener types, so an industry requirement is a test of its own, not part of the function count. Misleads when you read "5 years in fintech" as met by 5 years of product work in retail. It is not.

4. Level equivalence

Proves: you have already done the work one level below your target. OPM credits specialized experience only if it is equivalent to at least the next lower grade, and outside applicants need 1 year at that standard. A year counts only if its scope, independence and consequence matched the level. Misleads when you held the title but not the decisions. To score above 0 you must be able to name one decision you owned.

5. Knockout exposure

Proves: your application reaches a human. A failed required screener is an automatic rejection, and LinkedIn job posters can automatically archive candidates who fail screening questions and send a rejection email. Indeed is different: custom questions there cannot be marked as required, so a posting that shows no required screener scores 1, not 2. Misleads when you pass only by typing calendar years. That pass is a false positive your resume will contradict.

What the score means: lateral, step down, entry, or bridge

Apply the knockout gate first, then read the total. A score of 8 to 10 with full level equivalence means apply lateral. A score of 5 to 7 means step down one level, 3 to 4 means go entry, and 0 to 2 means take a bridge role first. A knockout fail at any level rules that level out whatever the total, and you re-score one level down.

Score from the top level down. Under federal rules, meeting a grade's requirements also means meeting them for the lower grades in the same occupation. The same logic works here: if you score lateral, a step down is always open as a fallback, but the reverse is not true.

Entry level is not a guaranteed fallback. Postings labelled entry-level often ask for years you may not have as a changer:

The TalentWorks figure is the older of the two, so read it alongside the LinkedIn figure rather than on its own. Either way, a third or more of entry postings would knock out someone with under three countable years in the function. Score the entry posting on the same five dimensions before you treat it as your safety net.

That sample suggests the associate level in product management may be hard to find as a visible title, which makes "step down to associate" a less obvious target than the ladder implies. Check by searching the employer's own postings at that level before you plan around it.

If Refolk's fit score for a posting disagrees sharply with your total, go back to the labels from step 1 before trusting either number. The usual cause is a function requirement that was counted as an industry one, or the reverse.

What does stepping down one level actually cost?

Usually less than the headline gap between levels, because pay bands for adjacent levels overlap. On the US federal GS ladder, the first step of each grade sits roughly 16 to 18% above the first step of the grade below. But the top step of the lower grade pays more than the first step of the upper grade.

The figures come from one year's published GS base pay table; the gaps are derived from it. The lower rows jump two grades because most professional entry-to-mid occupations advance in two-grade intervals. The table is republished each year, so re-check the current one before quoting any figure.

The limit: no one publishes level-to-level pay gaps for private-sector role families, and this table covers the federal ladder only. To check locally, find the posted pay ranges for both levels at the same employer. If the top of the lower range reaches the bottom of the upper range, the step down becomes a negotiation about where you land in the band, not a fixed cut. The ONS and wage-study figures above point the same way: a step down into a better-paid occupation can still pay off, but not by the full difference between the two occupations' averages.

The number of roles at each level also matters, and it differs by market. In Refolk's index, Data Analyst-family titles show this pattern:

In both samples, 13 of 25 held the plain "Data Analyst" title.

Can you apply to two levels at the same company?

Yes, if the portal's limits leave room and no single recruiter covers both postings. Whether a second application hurts depends on system rules and permissions, not on etiquette.

Greenhouse's Candidates page also lists a person once per application, so two levels means two entries. Whether that registers depends on who looks.

The second-application rule:

1. Apply only at adjacent levels. Pair lateral with one level down, or one level down with entry. Applying at both lateral and entry on the same team tells the recruiter you have not worked out your level.
2. State the same countable figure in both. Two applications with different year counts are the clearest sign of inflation a recruiter can see.
3. Check the cap first. On a Google-style limit, the second level uses one of three slots in 30 days, and a rejection sets a 90-day wait before you can reapply to that job.
4. Decide both levels up front where reapply blocks are possible. If the employer blocks rejected candidates, a lateral rejection can stop the lower-level application from going in at all.
5. Pick one level when one recruiting team covers both. Two postings in the same job family under one recruiting team is the unsafe case.

Practitioner advice puts tailoring ahead of spacing: each application should be written for its own level. The lower-level version should lead with the decision you owned, not with your old title. Refolk tailors a separate resume to each posting, which keeps the two versions consistent on dates and years while their emphasis differs.

Step-by-step: scoring one posting to a verdict

Run these seven steps once per posting. The timed steps take about 65 to 75 minutes the first time, and much less once your worksheet exists.

Sources disagree on the order. Practitioners tailor first and space applications second. OPM treats level as a precondition and duration second: a year that fails the level test does not count at all. I put counting before the level test because counting is mechanical and quick. But if step 4 shows a period was below the level just under your target, go back to step 2 and remove those months from the function count.

Employer and posting:
Posting level: lateral / one down / entry

D1 Change type (industry only 2, function only 1, both 0): __
D2 Countable function years vs required (__ / __): __
D3 Countable industry years vs required (__ / __): __
D4 Level equivalence, decision I owned one level down: "__": __
D5 Knockout exposure (any required screener failing?): __

Total: __ / 10
Gate: does D5 score 0 at this level? yes / no
Verdict: lateral / step down / entry / bridge

Second level at this employer?
Portal cap, waiting period and reapply block checked: __
Same recruiting team covers both postings: yes / no
Countable figure stated in both applications: __
Decision: submit second / hold / one level only

How the level call goes wrong

Most bad level calls come from counting mistakes a recruiter can see, not from aiming too high. Each failure below makes the score look better than it is, and each has a check that takes minutes.

Over-counting part-time years

Twenty hours a week for 3 years is 1.5 years under the OPM rule. The false positive happens when you answer the screener with "3" and your resume then shows part-time hours. Recruiters check the answer against the documents, and the standard advice to employers is that a candidate's resume and LinkedIn profile should reflect what they claimed. Check: total qualifying hours divided by 2,080.

Function years mistaken for industry years

"5 years in fintech" is not met by 5 years of product work in retail, however strong the product work was. The mistake inflates dimension 3 and hides an industry knockout. Check: label each requirement as function or industry before you count anything.

Duration without level

A year counts only if its scope, independence and consequence matched the level just below the one you target. Three years of support work in a function do not make you ready to apply at that function's middle level. Check: write down one decision you owned at the level below. If you cannot, dimension 4 scores 0.

Assuming every hiring manager sees everything, or nothing

In Greenhouse, who sees your other applications depends on permissions, not on a blanket rule. Assuming everyone sees everything makes you avoid safe second applications. Assuming no one does makes you send unsafe ones. The case that is actually unsafe is a recruiter who covers both postings. Check: are both postings in the same job family under one recruiting team?

Treating the cap as a style rule

Google's 3-in-30-days limit is a hard limit, and each application uses up a slot. Some Ashby employers block rejected candidates from reapplying, so a lateral rejection can close the lower-level door entirely. Check: read the portal's FAQ for limits and waiting periods before the first submission, not the second.

Trusting raw title counts

Title matching over-counts. Refolk's US Data Analyst sample included an Electrical Assembler and founders. A raw total can make a level look busier than it is. Check: rely on the split between the top titles, and confirm the level exists at your target employer by searching its own postings.

Stale benchmarks

The 61% entry-level figure comes from the older of the two studies. Quoting it alone overstates how closed entry level is now, and quoting the 35% alone may understate it. Check: use the dated LinkedIn figure alongside the older one, and note the date of any benchmark you rely on.

Before you submit, and keeping the call current

A level call is done when every number in it would survive a recruiter asking where it came from, and when any second application has been checked against the portal's rules.

Keeping the call current means re-running it when the inputs change:

• Your countable figure grows every month. In a bridge role, re-score your original target each time the figure crosses a threshold a target posting uses. Do not wait for a fixed date.
• Portal rules change. Caps, waiting periods and reapply blocks are employer settings. Re-read the portal's FAQ for each new employer, and again before any reapplication.
• Pay tables republish. The GS base table is reissued, so recompute the band overlap from the current table before you quote it.
• Benchmarks age. When you cite a share of entry postings asking for experience, attach its date, and look for a newer figure with a clear method before relying on the old one.
• Rejections start clocks. On a portal with a 90-day reapply wait for the same job, note the date of the rejection so you know when that posting, or its next opening, is available to you again.