Refolk
FrameworkReading the market

The Metro Target Score, Six Dimensions to Move, Apply From Away, or Drop

Score any metro for your own occupation on six dimensions and reach a move-first, apply-from-away, remote-only, or drop verdict that a second scorer would match.

23 min readLast reviewed November 16, 2026Read as Markdown

Key takeaways

  • Score each city for your own six-digit occupation code, not on all-occupation rankings, because a city's overall job growth says nothing about how many employers hire your role there.
  • Price adjustment can flip a pay ranking: nominal pay 30% higher in San Francisco than in Tulsa buys the same at national prices, because 115.6 divided by 89.2 is 1.30.
  • An employer-concentration index (HHI) above 2,500 means your role's market there is no deeper than four equal-sized employers, and the average occupation market sits at 4,378, about 2.3 employers.
  • The distance penalty comes early: applicants are 35% less likely to apply to a job 10 miles away, but only 9% less likely at 110 miles than at 100, so the real choice is local versus not local.
  • In Refolk's index, only 14 of 64,856 US Data Analyst profiles say they are open to relocation in the headline, so a clear relocation line is rare enough to be noticed.
  • No public source measures how hard applicant tracking systems filter out non-local applicants, so local preference should be scored only from explicit wording in your own sample of job postings.

This guide is for people who have three to eight cities they could live in and one occupation to sell. It rates each metro for that occupation on six dimensions, using public government tables and a 30-day pull of job postings. Each city then gets one of four verdicts: move first, apply from away, remote-only from home, or drop. The cut lines are written down so that a second person with the same inputs reaches the same verdict.

What does the Metro Target Score decide?

For one occupation, the score decides whether a metro is worth moving to before you have an offer, worth applying to from where you live, worth only remote applications from home, or not worth any more applications. It grades your occupation in each city, not the city's economy as a whole.

The pages that rank the best cities for jobs measure something else. Indeed's Best Cities for Work, for example, counts posting volume averaged over a 30-day window and scaled per 1,000 residents, across all occupations. Indeed's tightness index ranks 800+ metropolitan and micropolitan areas, also across all occupations. A city can top those lists and still have two employers hiring your role. The missing measure is a 30-day posting pull for your own occupation, and this score is built around it.

Distance matters in a specific way:

  • Applicants are 35% less likely to apply to a job 10 miles from their home ZIP code.
  • Beyond 100 miles, distance adds little. Applicants are only 9% less likely to apply at 110 miles than at 100.
  • In CareerBuilder data, 81% of applications stay inside the applicant's commuting zone, so about 19% cross one. A commuting zone is a cluster of counties that share one labor market.

Most of the distance penalty comes in the first few miles. Once you are applying outside your metro, the extra miles matter less than the plain fact that you are not local. That is why the score focuses on local preference instead of distance.

Three terms come up throughout:

  • SOC-6 code: the six-digit Standard Occupational Classification code that federal occupation tables use.
  • MSA: a metropolitan statistical area, the unit the government tables report.
  • OEWS: the Occupational Employment and Wage Statistics program at BLS. It publishes employment and wages by occupation for about 530 metropolitan and nonmetropolitan areas.

Which six dimensions does the score use?

Each dimension scores 0, 1 or 2, so the total runs from 0 to 12. Depth, concentration and pay come from public tables. Employer thickness, live openings and local preference come mostly from your own posting pull.

DimensionScores 2Scores 1Scores 0
DepthTop third of your list by OEWS employmentMiddle thirdBottom third, or "Estimate not released"
ConcentrationLQ 1.0 or higherLQ 0.7 to 0.99LQ below 0.7
Employer thicknessHHI below 1,500HHI 1,500 to 2,500HHI above 2,500, or no postings
Live openingsProjected openings and 30-day count both in top half of listOne of the two in top halfNeither
Price-adjusted payAt or above home on both all-items and housing RPPAt or above home on oneBelow home on both
Local preferenceUnder 25% of postings say local-only or no relocation25% to 50%Over 50%

The 0.7 location quotient line comes from a BLS example, covered below. The HHI bands are the federal antitrust bands that labor economists use. The 25% and 50% lines for local preference are my own convention, because no published share of local-only postings exists. When you split a list into thirds or halves and it does not divide evenly, put the extra metro in the lower band.

Depth: OEWS employment count

What it proves: how many people in the metro already do your job, which is the pool of jobs people leave and get replaced in.

How it lies: OEWS employment does not include the self-employed, so it undercounts freelance-heavy occupations. Estimates are rounded to the nearest ten. If the cell reads "Estimate not released", Depth scores 0. OEWS says it cannot give the specific reason an estimate was withheld, so a suppressed cell tells you nothing about the true size.

Concentration: location quotient

A location quotient (LQ) is the ratio of an occupation's share of employment in an area to its share of US employment. At 1.0, the metro matches the national share.

What it proves: whether the metro specializes in your role. Specialization tends to bring local training pipelines, peer networks and employers who already know how to hire for it.

How it lies: in a small metro, a high LQ can sit on very few jobs. BLS records an area that employed over 500 purchasing agents but had an LQ of 0.7. That is a real market that the LQ undersells. My rule is that when Depth is 0, Concentration cannot score above 1.

Employer thickness: HHI from your posting pull

What it proves: how many separate employers are competing to hire your role right now, which is the number of doors you can knock on.

How it lies: staffing agencies, reposts and multi-location chains each look like separate employers. Count them uncleaned and a thin market looks thick.

Live openings: projections plus a posting count

What it proves: whether jobs are open, both in the long run (projected annual openings) and right now (the 30-day count).

How it lies: projections can be a year or more out of date, or suppressed for small occupations. A 30-day count can spike because of one employer's hiring wave, which is why Thickness is scored separately.

Price-adjusted pay at your percentile

What it proves: what the metro's pay for someone at your level buys compared with staying home. The next section has the arithmetic.

How it lies: the all-items price index hides housing, the biggest cost for most renters.

Local-candidate preference

What it proves: how openly the metro's employers screen for local applicants. This decides between moving first and applying from away.

How it lies: it records only what employers write in postings. Any screening that is not written down is invisible to you, and no public source measures it.

How do you price-adjust pay at your own percentile?

Divide the metro wage at your percentile by that metro's Regional Price Parity divided by 100. The result is a national-equivalent wage that you can compare across cities. Regional Price Parities (RPPs) are BEA's measure of price-level differences across states and metro areas for a given year, as a percentage of the national level.

The formula is: national-equivalent wage = metro wage / (RPP / 100). BEA's latest RPPs cover 387 metros. The highest is San Francisco-Oakland-Fremont at 115.6 and the lowest is Monroe, LA at 83.6.

MetroRPP$100,000 at national prices
San Francisco115.6$86,505
New York112.6$88,810
Tulsa89.2$112,108
Monroe, LA83.6$119,617

Among metros with at least one million residents, Tulsa's 89.2 is the lowest price level. The ratio 115.6 / 89.2 = 1.30 means a San Francisco salary 30% above a Tulsa salary buys the same goods at national prices. So the metro with the highest nominal pay can score lower on this dimension than a cheaper metro that pays less.

Use the wage at your own percentile, not the median. If you are senior, the median understates what you would be paid. If you are early in your career, it overstates it. Pick the OEWS percentile column closest to your level and use it for every metro, including home.

Then run the formula a second time with BEA's housing RPP. The all-items RPP spans 32 points, from 115.6 down to 83.6. BEA's rent index spans about 170 points, from 211.9 in San Jose to 42.3 in Johnstown, PA. For a renter, that housing gap can swamp everything else, which is why Pay scores 2 only when the metro beats home on both measures.

How many employers make a market thick enough?

A market is thick enough when your 30-day posting pull, after cleaning, has an HHI below 1,500, which works out to roughly seven or more equally sized employers. Above 2,500, the market is no deeper than four equal employers, and one hiring freeze can close a large share of it.

The Herfindahl-Hirschman Index (HHI) measures how concentrated hiring is. For each employer, take its share of the metro's postings as a percentage, square it, and add up the squares. If one employer posts everything, HHI is 10,000. If four employers each post a quarter of the jobs, HHI is 2,500. Divide 10,000 by the HHI to get the equivalent number of equally sized employers.

HHIBandEqual-sized employers
1,500Moderately concentrated floor6.7
2,500Highly concentrated floor4.0
4,378US average market2.3
5,000Very high2.0

The bands come from the federal antitrust guidelines. Azar, Marinescu, Steinbaum and Taska apply them to labor markets, defined as one SOC-6 occupation in one commuting zone over one quarter. They find the average market has an HHI of 4,378, the equivalent of 2.3 recruiting employers, and 60% of markets are highly concentrated. Concentration is lower in large commuting zones, but the average is still thin. This is the dimension most likely to send a small metro to drop.

Thin evidence warning: those bands were built on quarterly data, and no 30-day employer-count threshold has been published. A 30-day window catches fewer employers than a quarter does, so your HHI will tend to come out higher than a quarterly figure would. I use the quarterly bands as the closest available anchor and apply them to every metro in the same way. That way the comparison between your cities holds, even if the absolute level is a little harsh. Indeed's own index mixes concentration with engagement data, so do not swap its figure in for yours.

How do you score a metro, step by step?

Run these seven steps for every candidate metro, with one occupation code, one OEWS vintage and one 30-day window across the whole list. Expect about half a day per metro, most of it spent on the posting pull.

Scoring one metro

  1. Fix the SOC-6 code
    Pick the one six-digit SOC code that matches the work you would be hired to do, and use it in OEWS, the state projections and the posting pull. Done when all three sources resolve to the same code, which takes about 20 minutes.
  2. Pull the OEWS metro row
    For each metro, record employment, employment per 1,000 jobs, location quotient and the wage at your percentile, all from one OEWS vintage. Done when every field is filled, or a state wage has been substituted and flagged where the cell reads Estimate not released.
  3. Price-adjust pay at your percentile
    Divide each metro's percentile wage by RPP/100, once with the all-items RPP and once with the housing RPP, and do the same for your home metro. Done when every metro, home included, has two national-equivalent figures.
  4. Pull projected annual openings
    From the state labor market information site, take MSA or substate projected annual openings for the code and note the projection vintage. If the figure is missing or suppressed, scale the state figure from Projections Central by the metro's share of state employment, and flag it.
  5. Run the 30-day posting pull
    Collect every posting for the code in each metro over one 30-day window, remove reposts, collapse employers to legal entity and drop staffing agencies, then compute HHI. Done when you have a deduplicated employer list, a posting count and an HHI per metro.
  6. Code local-preference language
    Tag each posting in the same sample for explicit local-only text, no-relocation text, onsite requirements and remote availability. Done when each metro has a local-preference percentage and a count of remote postings.
  7. Score, assign a verdict, get a second scorer
    Score all six dimensions against the cut lines, run the decision order, then give the raw inputs to a second person to score without seeing your results. Done when both of you reach the same verdict, or when the disagreement is settled by a written coding rule and both of you rescore.

Notes on the slower steps:

  • Step 2. The state fallback for a withheld wage cell is my own rule, because OEWS publishes no sample-size cutoff. Substitute the state wage, flag it, and do not substitute for employment. A withheld employment cell scores Depth 0. OEWS also no longer publishes metropolitan divisions, so the largest metros only come as whole MSAs. If you are aiming at one side of a big metro, the table cannot isolate it.
  • Step 4. Projections Central is the federal site for state-level occupational projections. It reports average annual openings per occupation and offers REST and direct download. Metro figures are only on each state's own site, and coverage varies. Ohio, for example, covers the state, its eight largest MSAs and six JobsOhio regions. Allow 30 to 90 minutes depending on the state.
  • Step 5. Budget 2 to 4 hours per metro. Dedupe on the legal entity, not the brand name on the posting, and keep the agency exclusions in a separate list so a second scorer can check them.

How do the scores turn into a verdict?

Run the six scores through a fixed decision order and stop at the first rule that fits. The order matters: a weak market is dropped before its pay is considered, and pay is checked before local preference.

  1. Drop if Depth is 0 and Employer thickness is 0.
  2. Drop if the total is 5 or less.
  3. Remote-only from home if Pay is 0 and the sample contains remote postings. Drop if it contains none.
  4. Move first if the total is 9 or more and Local preference scores 0 or 1.
  5. Apply from away if Local preference scores 1 or 2.
  6. Remote-only from home for everything left over (total 6 to 8, Local preference 0) if the sample contains remote postings. Drop if it contains none.

The reasoning behind the rules: moving first only pays off where the market is strong and its employers write a preference for locals into their postings, because that is the screen that being there gets you past. A strong market that does not screen for locals does not need you to move. A market that pays less in buying power than home is only worth its remote roles, paid at your home rate.

Market strength against local preference

Strong market (sum high)Weak market (Depth to Pay sum low)
Apply from away
Send tailored applications with a dated relocation line
Move first
Being local is the screen; move before the offer if savings allow
Remote-only from home
Apply only to remote roles, priced at your home rate
Drop
Stop applying and move the effort to a better metro
Low local preferenceHigh local preference
Strong markets that screen for locals justify moving first; strong markets that do not can be worked from away, and the decision order settles the borderline cases.

The score is not a full move decision. It leaves out your savings, your lease and your household, and moving without an offer depends on all three. If a metro gets move first and you cannot afford a move without an offer, treat it as apply from away until you can.

Once a metro gets apply from away, the slow part is volume: every posting needs a resume that matches its requirements and a cover letter that carries your relocation line. Refolk writes the resume from your own history, tailors it to each posting and drafts the cover letter, so a list of four apply-from-away metros does not become four lots of hand editing.

Metro score sheet
OCCUPATION: [title] | SOC-6: [code]
OEWS vintage: [month year] | RPP year: [year] | Posting window: [start date] to [end date]
Home metro: [metro] | Wage percentile used: [P__]
Home national-equivalent: all-items $[ ] | housing $[ ]

METRO: [name]
1 Depth           OEWS employment [ ]   rank [ ] of [ ]            score [ ]
2 Concentration   LQ [ ]   (capped at 1 if Depth = 0)              score [ ]
3 Thickness       postings [ ]  employers [ ]  HHI [ ]             score [ ]
4 Live openings   projected annual [ ] (vintage [ ], scaled? y/n)
                  30-day count [ ]                                 score [ ]
5 Pay             nat-equiv all-items $[ ]  housing $[ ]           score [ ]
6 Local pref      local-only or no-relo [ ] of [ ] = [ ]%
                  onsite [ ]  remote [ ]                           score [ ]
TOTAL [ ] / 12

Decision order: rule fired [1-6] -> VERDICT: [move first / apply from away / remote-only / drop]
Flags: [state wage substituted / projection scaled / agencies excluded: list]
Second scorer: [name]  verdict [ ]  match [y/n]  rule added if not: [ ]

One block per metro; copy the header once per list and give the filled sheet plus raw files to your second scorer.

How does the Metro Target Score go wrong?

It usually goes wrong in one of seven ways. Each one makes a metro look better or worse than its inputs support, and each has a check you can run in a few minutes. This section matters more than the rubric: a score built on these errors gives you a confident wrong answer.

  1. LQ read without employment. A high LQ in a small metro looks like a deep market and is not. The BLS case of over 500 purchasing agents at an LQ of 0.7 shows the opposite error. Check: score Depth beside Concentration every time, and apply the cap.
  2. Comparing two OEWS vintages. BLS warns that OEWS is not designed as a time series. Each estimate pools six semiannual panels collected over three years, so consecutive vintages overlap. A metro that looks like it is growing between two vintages may only be showing that overlap. Check: confirm every figure on the sheet comes from one vintage.
  3. All-items RPP hiding housing. With a 170-point rent spread against a 32-point all-items spread, a renter who uses only all-items can rank an expensive city too high. Check: compute both national-equivalent figures and score Pay on both.
  4. Inflated employer counts. Staffing agencies, reposts and multi-location chains each look like a separate employer in a raw pull. Four brands owned by one company are one employer. Check: dedupe on legal entity, exclude agencies, recompute HHI, and keep the exclusion list.
  5. Treating local preference as measured. Every source claiming that applicant tracking systems filter by ZIP code is advice content. A registered resume audit that randomizes applicant ZIP codes across local short commutes, local long commutes and non-local addresses exists, but its results are unverified. The share of postings that state local-only, onsite or no-relocation requirements is not established publicly. Check: score only explicit text in your own sample.
  6. Stale or missing projections. MSA projections can be a year or more behind. South Dakota, for example, publishes its MSA projections well after the base period. South Dakota also suppresses projections for occupations with base-year employment under 30. Check: write the vintage on the sheet. Where the figure is missing, scale state openings by the metro's share of state employment and flag it.
  7. Coarse profile locations. If you check Depth against professional profiles, remember that many list only a country. In a 25-profile sample from Refolk's index, 28% of Canadian Data Analyst profiles listed only "Canada", against 4% of US profiles that listed only "United States". Check: treat any metro headcount taken from profiles as a lower bound, never as a replacement for OEWS.
A tidy score on a dirty employer list is the most common way a thin market gets a move-first verdict.

One more mistake to watch for: if your second scorer reaches a different verdict, do not average the two. The difference almost always comes from one coding call, usually whether a brand counts as a separate employer or whether "must be based in the area" counts as local-only. Write the rule down, and both of you rescore.

What does applying from away look like in practice?

Applying from away means making your relocation obvious, dated and easy to check, because in your own field almost nobody else does. Recruiters scanning for out-of-town candidates who are serious about moving have very few to choose from.

Refolk's index shows how rarely the signal is used. Data Analyst is the example occupation here, but the pattern is what matters for your role:

SliceProfilesExact-title share of sampleCountry-only location
US Data Analyst64,85614/25 (56%)1/25 (4%)
Canada Data Analyst8,28121/25 (84%)7/25 (28%)
US, "open to relocation"1412/142/14
14
US Data Analyst profiles, out of 64,856, that say open to relocation in the headline
In Refolk's index that is about 0.02% of the base, so anyone searching on the signal finds almost nobody.

Who signals relocation

  1. US Data Analyst profiles
    64,856

    Full base in Refolk's index

  2. Headline says open to relocation
    14

    About 0.02% of the base

  3. Exact Data Analyst title among them
    12

    2 of the 14 list a country only

Out of 64,856 US Data Analyst profiles in Refolk's index, 14 state openness to relocation and 12 of those carry the exact title.

The limits of this evidence: it shows the signal is rare. It does not show that recruiters reward it, and I know of no public study that tests that. An older Indeed figure found 27.8% of American job seekers searching in another state, so many people are looking across state lines without saying so. Being one of the few who says it plainly costs nothing to try. Whether it gets you more interviews is something to track in your own sent log.

How to work an apply-from-away metro:

  • Put the target metro and a month in your resume header and on your profile headline. Leave your current city off the header if the relocation line already explains it.
  • In every cover letter, state the move date and when you are available for onsite interviews.
  • Prioritize postings without local-only text. Your coded sample already tells you which employers those are.
  • If you reach a final round and the move-first conditions now apply, rescore that metro instead of deciding on momentum.
Relocation line for resume header and cover letter
Resume header:
[Your Name] | Relocating to [Metro], [State] by [Month] | [email] | [phone]

Profile headline addition:
[Role title] | Relocating to [Metro] by [Month]

Cover letter sentence:
I am relocating to [Metro] by [Month] and can attend onsite interviews there with [number] days' notice; my start date is not dependent on relocation assistance.

Replace the brackets and delete any clause that is not true for you; never promise a date you cannot keep.

Because Refolk tailors each resume and cover letter to the posting, the relocation line goes into every version without you pasting it by hand. Refolk's fit score also shows which postings in an apply-from-away metro actually match your history before you spend time on them.

How do you keep the scores current?

Rescore the whole list when any input changes period: a new OEWS release, a new RPP year, a new projection vintage, or a 30-day pull that is more than a window old when you act on it. Do not update one metro's inputs while the rest of the list stays on old ones.

What to check, and when:

  • OEWS updates once a year, normally in March. When it does, replace every metro's row together, and never read the change between vintages as growth.
  • RPPs come out once a year. Check the release date on BEA's Regional Price Parities page instead of assuming one, and replace both the all-items and housing figures for every metro.
  • Projections follow each state's own schedule. Note the vintage on the sheet. If a state publishes a new MSA vintage, rescore Live openings for its metros, and keep the other states' figures as they are, flagged with their own vintage.
  • The posting pull has the shortest shelf life. Rerun it before you act on any move-first verdict, and before giving up on a metro that dropped on Thickness alone. One new employer can change a thin market's HHI.

Before you act on a verdict

  • Every metro uses the same SOC-6 code, OEWS vintage, RPP year and 30-day window.
  • Every substituted state wage and scaled projection is flagged on the sheet.
  • The employer list is deduplicated on legal entity, with staffing agencies listed separately and excluded.
  • Pay was scored on both the all-items and the housing RPP against your home metro.
  • The local-preference percentage counts only explicit local-only or no-relocation text from your own sample.
  • Any remote-only verdict rests on a sample that contains at least one remote posting.
  • A second scorer reached the same verdict from the raw inputs, or a written coding rule settled the difference.
  • Any move-first metro has a posting pull dated within one window of today, and you have checked your savings separately.

Once the verdicts are settled, put your effort where they point. Drop means stop applying to that metro. Remote-only means filtering that metro's postings to remote roles and pricing your asks at your home-metro rate. Apply from away means a dated relocation line on every application. Move first means rerunning the pull, checking your runway, and setting a date.

Questions job seekers ask

Should I move to a new city before finding a job?

Only when the city scores 9 or more out of 12 for your own occupation and its postings show real local preference. A strong market that screens out non-local applicants is the one case where being there changes your odds. If the market is strong and postings do not ask for locals, apply from where you are. The score does not cover your savings or lease terms, so check those separately before you move without an offer.

Do employers prefer local candidates?

Some do, but no public source measures how many. Claims that applicant tracking systems filter applicants by ZIP code come from career-advice blogs, not vendor documentation, and a registered resume audit testing local versus non-local addresses has no verified results. So the only evidence you can use is explicit wording in your own posting sample, such as local candidates only or no relocation. Score that wording and do not guess.

What is a location quotient and how should a job seeker use it?

A location quotient is an occupation's share of a metro's jobs divided by its share of all US jobs, so 1.0 means the metro matches the national share. Use it to see whether a metro specializes in your role, but never on its own. A small metro can have a high quotient and few jobs, and a big metro can have a low quotient and hundreds of jobs. Always score the employment count beside it.

How do I adjust a salary for cost of living by city?

Divide the metro wage by that metro's Regional Price Parity divided by 100. With an RPP of 115.6 for San Francisco, $100,000 is worth about $86,505 at national prices. With Tulsa's 89.2, it is worth about $112,108. Do this at your own wage percentile. If you rent, run it a second time with the housing RPP, because rents vary far more between metros than overall prices do.

Should I put a friend's local address on my resume?

I would not. Sources disagree on this: Ask a Manager has suggested using a friend's address, and Monster advises against it. The address does not hold up once interviews start, and nobody has shown that ZIP filtering is common enough to justify the risk. Instead, put a plain relocation line with the target city and a month in your resume header and cover letter.

How often should I rescore my candidate cities?

Rerun the 30-day posting pull before you act on any move-first verdict, and rescore every dimension when a new OEWS vintage comes out. OEWS normally updates in March. Never compare two OEWS vintages to claim a metro is growing, because BLS says the series is not designed as a time series. Check the Regional Price Parity and projection release dates on their own pages.

Put this to work

Paste your career in once. Every application after that is written for you.

Drop a resume or a LinkedIn URL. I rank the live openings against it, rewrite the resume and write a cover letter for the best of them, and fill in the employer's form when you press the button. You read, you decide what goes out.

  1. 01Drop your resume

    A PDF or a LinkedIn URL. About a minute, once.

  2. 02I rank the openings

    Every weekday morning, the live catalog scored against your history. Up to 20 worth your time, not two hundred links.

  3. 03Each one is written up

    Resume rewritten for the posting, a cover letter, a fit score. Press send, or let me fill in the form.

  • New matches ranked and written before you are up.
  • Every bullet stays inside what your history supports. Nothing invented.
  • Queued, submitted, interviewing, offer: one screen, not a spreadsheet.

Free, no card. Up to 25 applications sent a week, and nothing is sent until you say so.

Read next