# The Employer-Logo Credit Read: Full, Partial, or No Credit

*Score one employment line at a famous company on six public dimensions and give a Full, Partial, or No Credit verdict that a second recruiter would match.*

- Canonical URL: https://www.refolk.ai/guides/employer-logo-credit-read
- Pillar: Recruiting and sourcing
- Format: Framework
- Published: 2026-11-14
- Last reviewed: 2026-11-14
- Reading time: 25 min
- Keywords: how to evaluate a candidate's previous company, does ex-FAANG experience mean a strong candidate, company pedigree bias in resume screening, contractor vs full-time at big tech on a resume, was the hiring bar lower during hypergrowth, how much weight to give a well-known employer

## Key takeaways

- In Refolk's index, only 220 of 180,086 US profiles with Google in the headline say "contractor" (0.12%), while Google's reported contingent share in 2019 was about 54%, so a missing vendor tag is not evidence of full-time status.
- Employer prestige has no published validity figure as a predictor of job performance, while structured interviews sit at .42 in the Sackett et al. 2022 revision.
- In interviewing.io's recruiter study, candidates with a top company on their resume were 35% more likely to be picked, yet 46% of people who got offers at those companies had neither a top school nor a top company.
- Count a join year as hypergrowth when the company's headcount grew 20% or more, about twice big tech's 11% annual baseline from 2011 through 2019, and let it lower credit but never veto a candidate.
- Acquired employees turn over at a higher rate than directly hired ones, with the widest gap at senior ranks, so an acquisition entry earns no entry-route credit for the logo.
- Score six dimensions from 0 to 2 each, cap the verdict at Partial for vendor entry or scale mismatch, and read 9 or more as Full Credit, 6 to 8 as Partial, and 5 or below as No Credit.

This is the read I run when a candidate's profile shows a famous employer and I have to decide, before screening or ranking them, how much that company name should count toward the role I am filling. It is for in-house recruiters, sourcers, talent leaders, and founders who rank their own pipelines. It breaks one employment line into six public dimensions, scores each from evidence you can cite, and returns a verdict of Full Credit, Partial Credit, or No Credit for the logo.

## Why a famous employer name is weak evidence on its own

A famous employer name bundles several claims about a person: how they were hired, when, where they sat, and what they did. None of those claims has a published validity figure as a predictor of job performance. Recruiters reward the name anyway, so check it before you let it move a rank.

"Logo credit" here means the weight you allow the employer name itself, apart from the evidence of the work. Most public advice on how to evaluate a candidate's previous company stops at one of two slogans: "trajectory matters more than the logo" or "pedigree is bias". Both are partly right, and neither tells you what to do with the profile in front of you.

The evidence on how recruiters behave is clear enough:

- In interviewing.io's recruiter study, candidates with experience at a top company were 35% more likely to be picked.
- In a separate interviewing.io analysis, 46% of candidates who got offers at these companies did not have top schools or top companies on their resumes.
- In Lauren Rivera's *Pedigree*, which draws on 120 interviews with employers and 32 with candidates, firms rarely tracked whether school prestige actually predicted later job performance.
- interviewing.io's summary is the fairest one I know: where people have worked does carry some signal, but it is not nearly as important as other traits.

Compare that with the selection methods that do have published validity. "Validity" is the correlation between a selection method and later job performance.

| Method | 1998 estimate | 2022 revision |
|---|---|---|
| Structured interview | .51 | .42 |
| Work sample | .54 | .33 |
| Cognitive ability | .51 | .31 |
| Unstructured interview | .38 | .19 |
| Employer prestige | not published | not published |

Two things in that table matter for this read. First, the 2022 revision by Sackett, Zhang, Berry and Lievens cut most estimates by .10 to .20, and work samples dropped the most. Even the strong methods are weaker than older articles claim. Second, the last row is empty. Employer prestige has no published validity at all, so it should never anchor a ranking that structured evidence could anchor instead.

So does ex-FAANG experience mean a strong candidate? Not by itself. It means a set of claims you can check, and that is what the rest of this guide does.

## The six dimensions of an employer line

The read scores six things a company name implies without saying them: entry route, worker classification, join-year hiring era, org or team, in-company level trajectory, and scale relevance to your open role. You tag each one from public evidence and score it 0, 1, or 2.

#### What one employer name bundles

1. **Brand** - The company name in the headline, which is what most rankings stop at
2. **Entry and payroll** - Whether they came through the main bar, an acquisition, or a vendor, and whose payroll they were on
3. **Hiring era** - How fast the company was adding headcount in the year they joined
4. **Org and level** - Which team they sat in and how they moved up the company's own ladder
5. **Scale of the work** - What the work actually was, measured against the role you are filling

*The brand is the outer layer a recruiter sees, and each layer underneath is a separate claim that can be checked from public evidence.*

| Dimension | Tags and points | What it proves | How it lies |
|---|---|---|---|
| Entry route | Direct 2, Unknown 1, Acquisition 0, Vendor 0 | Passed the company's own hiring bar | An acquisition looks direct once the startup line is deleted |
| Classification | FTE 2, Unknown 1, TVC 0 | On the company's payroll, hired under its bar | Vendor name dropped from the profile |
| Join-year era | Normal 2, Hypergrowth 1 | Hired when volume pressure was lower | A firm-level number applied to one person |
| Org or team | Core 2, Adjacent 1, Support/Unknown 1 | Worked on what the brand is known for | No validated public method exists |
| Level trajectory | Fast 2, Typical 1, Stalled 0 | Performed against the company's own ladder | Titles hide levels, and terminal levels look flat |
| Scale relevance | Match 2, Partial 1, Mismatch 0 | The work transfers to this role | Big-company scale read as startup-relevant |

The points and caps are Refolk's weighting, not a published standard. No source documents a consensus on how to weight these dimensions. The rubric is useful because it is written down: two recruiters who apply the same weighting to the same evidence should reach the same verdict, and that is the test this read has to pass.

Org can add credit but never takes any away. Support and Unknown score the same as Adjacent because there is no validated way to place someone in an org from public evidence, and I will not penalise a candidate on a guess.

## Entry route and worker classification: who hired them, and on whose payroll

Entry route asks how the person got in: through the company's main hiring bar, through an acquisition, or through a vendor. Classification asks whose payroll they were on. You usually have to infer both, because profiles almost never state them.

### Why classification has to be inferred

As of March 2019, Google worked with roughly 121,000 temps and contractors worldwide against 102,000 full-time employees, according to an internal document obtained by The New York Times. That makes about 54% of the workforce contingent. Inside Google these workers are called TVCs: temps and contractors who are typically employed through a temp service or outsourcing company. A contracting job site estimated that contingent labour makes up 40% to 50% of the workers at most technology firms. That is a job site's estimate, not a regulator's.

The two groups were on different terms. Per Glassdoor, median pay was $90,000 a year for contractors against $128,000 for full-time employees. Ten US senators asked Google to convert contractors to full-time after six months. A contractor who shows the brand on their profile passed a vendor's screen, not necessarily the company's own hiring bar.

Here is what profiles disclose:

**0.12% - Share of US profiles with "Google" in the headline that also say "contractor"**

In Refolk's index, 220 of 180,086 US headlines match, against a reported contingent share of about 54% at Google in 2019.

| Measure | Value |
|---|---|
| US headlines matching "Google" | 180,086 (Refolk's index) |
| US headlines matching "Google contractor" | 220 (Refolk's index) |
| India headlines matching "Google contractor" | 39 (Refolk's index) |
| Contractor share of US Google headlines | 0.12% |
| Reported Google contingent share, 2019 | about 54% |

The two shares cover different populations, so the gap is not a precise measure of undercounting. The direction is not in doubt, though. Vendor work is a different legal status, and people pay a social cost for disclosing it, so they drop the vendor name. A profile with no vendor tag tells you very little about payroll.

Even the 220 matches need reading. In a 25-profile US sample of "Google contractor" headlines from Refolk's index, only 6 list Google as the current employer. Several are Google Ads consultants who serve building contractors, which makes them false matches.

**6 of 25 - "Google contractor" headlines in a US sample that list Google as the current employer**

Several of the rest are Google Ads consultants serving building contractors, so a keyword match is a lead, not a tag.

In Refolk's index, people who do disclose tend to word it like this:

- "Wipro Contractor for Google"
- "(Client) | via Virtusa"
- A vendor named as the current employer, with the client named only in the title or description

Nobody has published a validated list of wording that reliably flags vendor hires. Keep your own list and add to it as you find new patterns.

### Tagging classification

- **TVC:** the current-employer field names a staffing or outsourcing firm, or the line carries "via", "(Client)", or "contractor" wording that refers to the company itself.
- **Unknown:** no vendor wording, but the tenure is short and ends at a round 6, 12, or 24 months, or there is nothing else to go on.
- **FTE:** no vendor markers, plus at least one positive sign, such as a title change inside the company or a tenure that is neither short nor round.

Requiring a positive sign before tagging FTE is my rule. It follows from the numbers above: a missing marker is not evidence of full-time status.

### Spotting an acquisition entry

The pattern is a startup role that ends in the same month the acquirer role starts, where that startup was bought by this company. Recognising it is practitioner inference. No source validates it as a detection method. Check by hand that the earlier employer was acquired by the same company, and that the acquisition month matches the join month.

Why does it score 0 for entry? An acquired employee arrived through a deal and did not pass the main hiring bar to get the logo. The retention evidence also points one way:

- Gehrke and Maug studied 1,043 German deals from 1997 to 2014 and found that net employment at target firms halves within two years of an acquisition. That is a net figure for the target firm, not a measure of how many individuals stayed. No one-year figure has been published from that study.
- A Strategic Management Journal study using over 30 million resumes found that acquired employees turn over at a much higher rate than matched employees who were hired directly. The gap is larger for acquired employees in higher job ranks and with advanced degrees.

Those senior acquired staff are exactly the people a logo flatters most. An acquisition entry is not a mark against the person. It means the logo's credit has to be earned later in the read, through trajectory inside the acquirer.

> **Watch out:** A missing vendor tag is not evidence of full-time status
>
> Fewer than 1 in 800 US Google headlines in Refolk's index mention contracting, against a reported contingent share above half. Tag FTE only when you can point to a positive sign, and otherwise tag Unknown.

Checking acquisition timing by hand means knowing every deal the company has done and lining up month-level dates across two employers. It is the slowest part of the read. In [Refolk](/) you can ask for the pattern in plain English and start from profiles that already show it:

I ran this search: `Engineers whose Google or Meta role started the same month their startup was acquired by that company` - [see the full result list](https://www.refolk.ai/s/cz3wcn9s1w).

*Returns profiles where a startup role ends as the Google or Meta role begins, so you can confirm the deal and tag the entry route as Acquisition.*

## Join-year hiring era: was the company hiring unusually fast?

The era dimension asks whether the person joined in a year when the company was adding headcount unusually fast. I count a join year as Hypergrowth when headcount grew 20% or more that year. That is roughly twice the 11% a year that big tech companies grew from 2011 through 2019.

The 20% line is Refolk's suggested rule, not a published threshold. Here is how it plays out on year-over-year headcount change:

| Year | Meta | Microsoft (fiscal year) | Alphabet |
|---|---|---|---|
| 2019 | +26.3% | +9.9% | +20.4% |
| 2020 | +30.4% | +13.2% | +13.8% |
| 2021 | +22.8% | +11.0% | +15.7% |
| 2022 | +20.2% | +22.1% | +21.6% |
| 2023 | -22.2% | 0% | -4.1% |

Notes on reading it:

- The Alphabet figures for 2020 to 2022 are derived from year-end headcounts of 118,899, 135,301, 156,500, and 190,234.
- Microsoft reports by fiscal year, so map the join month to the right fiscal year before you tag it. Its FY2022 growth took it from about 181,000 to 221,000.
- Meta went from 44,942 staff in 2019 to 86,482 in 2022, then fell to 67,317 in 2023.
- Amazon went from 798,000 staff in 2019 to 1,608,000 in 2021. Its counts include warehouse staff, so they tell you little about corporate or engineering hiring. Do not use them for an engineering line.
- The cycle turned: Google cut 12,000 employees in January 2023, and Meta shrank 22.2% that year.

Under the 20% rule, the Hypergrowth years in this table are Meta 2019 to 2022, Microsoft FY2022, and Alphabet 2019 and 2022.

### Why era is only a modifier

Era is a firm-level number applied to one person. Meta's +30.4% in 2020 tells you about hiring volume, not about any individual's interview. The claim that the bar dropped during the boom comes from anecdotes on Blind and has not been established. So was the hiring bar lower during hypergrowth? Possibly for some teams, but you cannot show it for any one person. That is why Hypergrowth scores 1, not 0, and can never produce No Credit by itself.

> **Rule:** Era is a modifier, never a veto
>
> A Hypergrowth join year can move a borderline line from Full to Partial. It must never be the reason a candidate is screened out, because that discards people who came through the main bar in a busy year.

For a company that is not in this table, use the same method: pull year-end headcount from the company's annual filings or a public headcount tracker, compute the year-over-year change, and compare it with the 20% line. Write the figure next to the tag so the second reviewer can check it.

## Org, level trajectory, and scale relevance

These three dimensions ask what the person did inside the company: which org they were in, how they moved up, and whether the scale of that work matches your role. Trajectory has conventions to lean on, org has no validated method, and scale has to be settled with the hiring manager.

### Org or team

There is no publicly established method for placing someone in a core or peripheral org. No source validates engineering blogs, talks, or code contributions as evidence for it. Treat this as a practitioner check only:

- **Core (2):** a named product the company is known for, confirmed by a public talk, post, or project name on the line.
- **Adjacent (1):** a title or project that names a product area next to the core, without outside confirmation.
- **Support/Unknown (1):** an internal-tools or support function, or nothing you can confirm.

A title claiming core infrastructure is weak until something public backs it up.

### Level trajectory

No primary levelling dataset publishes time in level. Levels.fyi publishes pay, not promotion intervals. The figures below come from secondary sources about Google's engineering ladder, so treat them as convention, not data:

| Move | Conventional interval |
|---|---|
| L3 to L4 | about 1.5 to 2 years |
| L4 to L5 | about 2 to 3 years |
| L5 to L6 | 3 to 5 years or more, and many never make it |

At Google, L4 is described as a terminal level with no organisational pressure to advance. Treat a long stay at L5 the same way.

- **Fast (2):** at least one promotion faster than the low end of the interval for that move.
- **Typical (1):** promotions within the intervals, or a long stay at a terminal level.
- **Stalled (0):** below a terminal level, well past the interval, with no title change and no other sign of a promotion.

Titles hide levels. "Software Engineer" covers L3 to L4, so a flat title can contain a promotion. When the profile shows no levels, default to Typical and record that the level was inferred.

For any company other than Google, find its ladder and its terminal level before you score. The intervals above do not carry over automatically.

### Scale relevance

Scale relevance is the dimension most often skipped and most often wrong. Big-company scale gets read as relevant to a startup role by default. A Blind commenter puts the counterpoint simply: smaller companies often look for people with specific expertise.

Before you score any line, sit down with the hiring manager and write down four things about the open role:

1. The company stage and team size the hire will work in.
2. The core problem the role owns.
3. What the hire will build or run alone, without platform teams behind them.
4. What a strong first six months looks like.

Then tag the line:

- **Match (2):** the work on the line was the same kind of problem at a comparable scope.
- **Partial (1):** a related problem, or the same problem with much more support around it than this role will have.
- **Mismatch (0):** a different problem, or a scope that will not transfer.

Write the scale brief once per req and reuse it for every candidate on that req.

## Turning six tags into a verdict

Add the six scores for a total from 0 to 12, then apply two caps. A total of 9 or more is Full Credit, 6 to 8 is Partial Credit, and 5 or below is No Credit.

The two caps:

- **Payroll cap:** a Vendor entry or a TVC classification caps the verdict at Partial.
- **Scale cap:** a scale Mismatch caps the verdict at Partial.
- **Both caps together** make the verdict No Credit, whatever the total.

| Verdict | Rule | What you do with it |
|---|---|---|
| Full Credit | 9 to 12, no cap applied | Count the employer line as evidence of a passed bar and relevant work |
| Partial Credit | 6 to 8, or a capped line | Credit the work, not the brand, and probe the low dimensions |
| No Credit | 5 or below, or both caps | Rank the candidate as if the company name were blank |

Some worked cases:

- A direct hire with positive FTE signs, a Normal-era join, an unconfirmed org, typical promotions, and a scale Match scores 2 + 2 + 2 + 1 + 1 + 2 = 10, which is **Full Credit**.
- The same line with a Hypergrowth join scores 9, which is still **Full Credit**. Era moves the score without deciding it.
- An acquisition entry with typical trajectory and everything else equal scores 8, which is **Partial**. The same person promoted fast inside the acquirer scores 9, which is **Full**, because they earned the logo after arriving.
- A vendor line with a scale Mismatch is **No Credit**, whatever the total.

#### Bar evidence against scale relevance

Horizontal axis runs from Work does not transfer to this role to Work transfers to this role. Vertical axis runs from Weak bar evidence (vendor, TVC, acquisition) to Strong bar evidence (direct, FTE).

| Quadrant | What it means |
| --- | --- |
| Bar without fit | Partial at most; ask what they would do without the platform behind them |
| Full Credit candidate | Score the remaining dimensions and let the logo count |
| Logo only | No Credit; rank on the rest of the profile |
| Relevant work, unproven bar | Partial; credit the work, then test the bar in a structured interview |

*The logo earns Full Credit only when the person both came through the company's own bar and did work that transfers to the open role.*

> **Rule:** No Credit is a verdict on the logo, not the person
>
> A No Credit line is ranked as if the company name were blank, using the rest of the profile and structured evidence. It is never a reason on its own to reject a candidate.

Record every read in the same format so the second reviewer can check it line by line:

**Employer-logo credit record (one employment line)**

```
Candidate:
Req and scale brief reference:
Employment line: Company | start month and year to end month and year | titles in order

1. Entry route: Direct / Acquisition / Vendor / Unknown   Score:
   Evidence:
2. Classification: FTE / TVC / Unknown   Score:
   Evidence (positive sign required for FTE):
3. Join-year era: Normal / Hypergrowth   Score:
   Join-year headcount growth and source:
4. Org or team: Core / Adjacent / Support-Unknown   Score:
   Evidence (named product or public talk required for Core):
5. Level trajectory: Fast / Typical / Stalled   Score:
   Ladder used, terminal level, inferred or stated:
6. Scale relevance: Match / Partial / Mismatch   Score:
   Brief line it matches or misses:

Total (0 to 12):
Payroll cap applied: yes / no
Scale cap applied: yes / no
Verdict: Full Credit / Partial Credit / No Credit
First reviewer:            Second reviewer (blind):
Disagreements and resolution:
```

*One record per employment line; every tag needs the profile text or public source it rests on.*

> A logo is a bundle of claims. Unbundle it, and most of the claims can be checked.

## The read on one profile, step by step

Run eight steps on one employment line. It takes roughly 15 to 20 minutes of sourcer time, plus the hiring-manager conversation on scale, which you have once per req. Then a second reviewer scores the same line blind.

#### The employer-logo credit read

1. **Isolate the employment line** - Write down the company, the start and end dates, and every title listed under it. Done when you have one dated line with every sub-role, in about a minute.
2. **Tag the entry route** - Look for vendor names, "via", "(Client)" or "contractor", and check whether a startup role ends in the same month the big-company role starts after that startup was acquired. Done when the line is tagged Direct, Acquisition, Vendor, or Unknown, in 2 to 3 minutes.
3. **Tag worker classification** - Check whether the current-employer field names a staffing firm, and whether tenure is short and ends at a round 6, 12, or 24 months. Done when the line is tagged FTE, TVC, or Unknown, with a positive sign behind any FTE tag.
4. **Tag the join-year era** - Look up the company's headcount growth in the join year and compare it with the 20% line. Done when the line is tagged Normal or Hypergrowth, with the growth figure written beside it.
5. **Tag the org or team** - Check titles, project names, and public talks or posts, in 3 to 5 minutes. Done when the line is tagged Core, Adjacent, or Support/Unknown, with Core only allowed when a named product or public talk confirms it.
6. **Tag the level trajectory** - Count title changes per year against the company's interval conventions, reading the ladder for terminal levels first. Done when the line is tagged Fast, Typical, or Stalled.
7. **Match scale to the open role** - With the hiring manager, write down the stage, team size, and core problem of the role, then compare it with the work on the line. Done when the line is tagged Match, Partial, or Mismatch.
8. **Score, cap, and get a second read** - Add the six scores, apply the two caps, and record the verdict with the evidence for every dimension. Done when a second reviewer has scored the same line blind and any disagreement is settled against the cited evidence.

A note on order: sources disagree on whether classification should come before entry route. Practitioner threads on Blind treat "contractor" as the first thing to check. I run entry route first because the acquisition check needs the previous employer line, which you are already reading. If your pipeline is heavy with vendor-placed profiles, check classification first. The verdict comes out the same either way.

If you are building a comparison set before you start scoring, such as direct, pre-boom joiners with at least one promotion, Refolk takes that as a plain sentence. "Software engineers who joined Google directly before 2019 and were promoted at least once within three years, currently in the US" is a complete search. Check each result against the read rather than trusting the match.

## How the read goes wrong

Most errors come from treating a missing signal as evidence: no vendor tag read as full-time, a flat title read as stalled, a direct-looking start read as a direct hire. These are the failure modes I see most, with the check for each.

1. **Reading silence as full-time.** Only 0.12% of US Google headlines in Refolk's index say "contractor", against a reported contingent share of 54%. *Check:* the current-employer field, any "via" or "(Client)" wording, and tenures that end at a round number. Tag Unknown when nothing positive shows up.
2. **Keyword false positives.** "Google contractor" also matches Google Ads consultants who serve building contractors. In one 25-profile US sample, only 6 listed Google as the current employer. *Check:* read the title and current employer before you tag anyone as TVC.
3. **Hidden acquisition entries.** An acquisition entry looks exactly like a direct hire once the startup line is deleted or merged into the acquirer's line. *Check:* whether the previous employer was acquired by this company in the join month, and whether the first title at the acquirer reads like a continuation of the startup role.
4. **Era used as a veto.** Penalising every 2020 to 2022 joiner throws out people who came through the main bar in a busy year, and the claim that the bar dropped is anecdote. *Check:* that Hypergrowth scores 1 and that no verdict anywhere in your pipeline rests on era alone.
5. **Terminal levels read as stalls.** Google's L4 is a terminal level, and a long stay at L5 is normal. A long stay at either is Typical, not Stalled. *Check:* read the company's ladder and find its terminal level before scoring trajectory.
6. **Flat titles read as flat careers.** "Software Engineer" covers L3 to L4, so a promotion can sit inside an unchanged title. *Check:* default to Typical and record the level as inferred.
7. **Core claims taken on trust.** No validated public method exists for placing someone in an org. *Check:* allow Core only when a named product or public talk confirms it, and never let org pull a score down.
8. **Big-company scale read as startup fit.** Scale is the easiest dimension to assume and the one hiring managers argue about most. *Check:* write the scale brief before scoring, and tag against the brief rather than against how impressive the company is.

There is one more failure mode, and it sits above all the others: letting the logo back in after the read. The interviewing.io finding that a top company made candidates 35% more likely to be picked describes a pull that does not go away because you scored a sheet. If a No Credit line keeps ending up at the top of the ranking, the brand is doing the ranking again.

## Keeping the read consistent between recruiters

The read is only worth running if a second recruiter scoring the same profile reaches the same verdict. Calibrate on shared profiles, track where you disagree, and re-check the inputs that change over time.

How to keep it honest:

- **Score blind.** The second reviewer scores the line without seeing the first record, then the two of you compare dimension by dimension.
- **Log disagreements by dimension.** If most disagreements land on org or scale, tighten the evidence rule or the scale brief for that dimension.
- **Re-pull headcount once a year.** Era tags rest on company growth figures that change as each year closes. Recompute year-over-year change from annual filings and update your era table.
- **Keep a ladder note per company.** Write down each company's levels, conventional intervals, and terminal level the first time you score a line there, and mark which parts are convention.
- **Grow your vendor-wording list.** Every new disclosure pattern you find, such as a vendor named as employer or a "(Client)" suffix, goes on a shared list.

When a hiring manager asks why an ex-big-tech candidate ranks below someone from a smaller company, send the record instead of an argument:

**Rank note to the hiring manager**

```
Subject: Why [candidate] is ranked where they are

I scored the [company] line on six dimensions before ranking.
Verdict: [Full / Partial / No] Credit, total [score] of 12.

What moved it:
- [Dimension]: [tag], because [one line of cited evidence].
- [Dimension]: [tag], because [one line of cited evidence].

What this means: the [company] name counts as [evidence of a passed bar and relevant work / evidence of the work only / nothing beyond the rest of the profile].
The rank rests on [the other evidence you used]. The full record is attached, and I am glad to walk through any tag.
```

*Send with the credit record attached; keep it to the dimensions that moved the verdict.*

#### Before you call the read done

- [ ] The employment line is written down with its dates and every title under it.
- [ ] Any FTE tag cites a positive sign, not just the absence of vendor wording.
- [ ] Any Acquisition tag is confirmed by a deal whose month matches the join month.
- [ ] The era tag shows the join-year headcount growth figure and its source.
- [ ] The trajectory tag names the ladder used and whether the level was stated or inferred.
- [ ] Any Core org tag cites a named product or public talk.
- [ ] The scale tag points to a specific line in the written scale brief for this req.
- [ ] Both caps have been checked and recorded as applied or not applied.
- [ ] A second reviewer has scored the line blind and every disagreement is resolved in the record.
- [ ] The final rank does not rest on the company name for any line scored No Credit.

## Frequently asked questions

### Does ex-FAANG experience mean a strong candidate?

No, not on its own. A big-tech employer name carries some signal, but no meta-analysis publishes a validity figure for employer prestige, and interviewing.io found that 46% of people who got offers at top companies had neither a top school nor a top company on their resume. Treat the name as a bundle of claims about entry route, payroll, era, org, trajectory, and scale. Check each one, and give credit only for the claims the profile backs up.

### How can I tell if someone was a contractor at Google rather than full-time?

Usually you have to infer it, because almost nobody says so. In Refolk's index, 0.12% of US profiles with Google in the headline say "contractor", while Google's reported contingent share in 2019 was about 54%. Look for a staffing firm in the current-employer field, wording like "via" or "(Client)", and short tenures that end at round 6, 12, or 24 months. If none of those show up, tag the line Unknown, not FTE.

### Was the hiring bar lower during hypergrowth?

Nobody has established it. Headcount data shows several large tech companies grew 20% or more a year in some years, against an 11% annual baseline from 2011 through 2019. That tells you about hiring volume, not about any one person's interview. The claim that the bar dropped comes from anecdotes on forums. Use the join year as a one-point modifier in the score and never reject anyone for it.

### How much weight should I give a well-known employer when ranking candidates?

Give it only as much weight as the scored read supports. Full Credit lets the employer line count as evidence that the person passed a selective bar and did relevant work. Partial Credit counts the work but not the brand. No Credit means you rank the candidate as if the company name were blank. Even at Full Credit, the logo should never outrank structured interview evidence, which has published validity when the logo has none.

### Is screening on company pedigree a form of bias?

It often works as one, because recruiters reward the logo more than outcomes justify. In interviewing.io's study, a top company made a candidate 35% more likely to be picked. But throwing the logo out entirely discards real information. The fix is to break the name into checkable parts and give credit only where the evidence holds, so you can defend every rank to a hiring manager.

### What do I do when the profile shows no levels, only titles?

Score trajectory from title changes, and be conservative. Titles hide levels: "Software Engineer" at Google covers L3 to L4, so a flat title can contain a promotion. Without any title change, tag the trajectory Typical, not Stalled, unless the tenure runs well past the conventional interval for a level below the terminal one. Write down that the level was inferred so the second reviewer knows how firm the tag is.

---

*From the Refolk guide library. I revise these guides rather than replacing them, so the current version is always at https://www.refolk.ai/guides/employer-logo-credit-read*
