The Resume-Line Probe Reference, One Row per Follow-Up an Interviewer Asks

An interviewer has stopped on one line of your resume and asked a follow-up: what was the baseline, was that you or the team, how long did it take. This reference is for candidates heading into a deep-dive round, or already in one. Each row names the probe, says what it verifies, shows which resume elements trigger it, and lists the facts a passing answer must contain. Find your row, check you have those facts, and move on.

What is a follow-up probe in an interview?

A probe is a follow-up question meant to complete an inadequate or incomplete answer, or to get additional or clarifying information. In structured interviews, the probes are often written before you walk in and asked of every candidate.

A structured interview fixes the questions, and often the follow-ups and the scoring, in advance so every candidate is judged the same way. Campion, Palmer and Campion (1997) list fifteen components of interview structure, and one of them is limiting ad hoc follow-up. Their four levels of probing control were restated in a later review in Personnel Psychology:

• No probes. Probing is prohibited entirely.
• Planned probes only. The interviewer gets a list of follow-ups written in advance and must not stray from it.
• Unlimited probing. The interviewer is encouraged to probe freely.
• No guidance. The interviewer decides.

US federal guidance from OPM tells interviewers: "Prior to the interview, establish the desired range of probing (for example, no probes, a limited number of probes, unlimited probes)." Its example probes are worth memorising because versions of them turn up everywhere: "What factors led up to the situation?", "How did you respond?", "What was your role?", and on a one-page example question, "Who was involved? What specific actions did you take? What was the outcome?"

What is not established: how often employers use planned probes compared with unlimited probing. None of the public sources I checked gives a percentage. One research experiment capped probes at two per question, but that was a study design choice, not an industry norm. Don't assume a probe limit. Assume each high-weight claim can be pressed until the interviewer has what they need.

Which resume lines trigger which probes?

Numbers trigger the metric probes, claims involving others trigger the role and decision-authority probes, dates trigger the timeline walk, and any claim the job hinges on can draw a manager-view or proof request. Interviewers go through the resume bullet by bullet, decide how to test each one, and keep the claims that matter most to the role.

The trigger map:

• A before-and-after metric. A bullet such as "improved ROAS from 2.1 to 4.3" (ROAS is return on ad spend) leads to questions about the baseline, measurement window, budget authority, attribution model and concurrent changes.
• A claim involving a decision or a budget. This leads to the decision-authority probe: what could you approve on your own?
• Work done with others. This leads to the role probe: who was involved and what did you specifically do?
• Date ranges and project phases. These lead to the timeline walk.
• Any claim central to the role. This can lead to the manager-view probe or a proof request.

Published interviewer guides don't all list the same probes. The table below shows which named probe appears in which source. A "no" means the source doesn't publish it, not that its interviewers never ask it.

Read the table as a coverage map. The role probe is the most widely published, so expect it in nearly every deep-dive. The timeline appears in four of the five sources. Baseline probes show up in the guides focused on testing resume achievements, so any resume with numbers on it should expect them.

You can't prepare every line equally, so triage. Weigh each claim by how much the role depends on it and how complete your facts are. If you're tailoring for a specific posting, Refolk scores how well you fit it, which is a quick way to see which lines that role leans on most.

The probe reference: one row per follow-up

Each row below gives the probe's usual wording, what it verifies, what triggers it, the minimum facts a passing answer contains, and how it misleads. The metric probes come first because a single number sets off all five of them.

Metric probes: the five checks behind any number

Any number on your resume triggers a fixed set of about five checks: baseline, time period, measurement method, concurrent changes and repeatability. They usually come in sequence, so prepare all five as a unit.

What the set verifies: that the number describes a real change you can explain, not a lucky window or someone else's result.

How it misleads: a candidate with a precise baseline can still fail if the window is cherry-picked or a concurrent change explains the result. Precision on one check doesn't carry over to the others.

Decision-authority probe

• Sounds like: "Which of these decisions could you make without approval?" or "Did you own the budget?"
• Verifies: the size of your scope, meaning what you could actually decide, not what you were near.
• Triggered by: budget figures, "owned" or "led" claims, any result that needed spending or sign-off.
• Passing answer contains: the decisions you made alone, the ones that needed sign-off and from whom, and your budget limit if there was one.
• How it misleads: needing approval isn't a weakness. Hiding the approval chain is, because your manager will describe it differently.

Role probe ("was that you or the team?")

• Sounds like: "What was your role?", "What specific actions did you take?", "Who was involved?"
• Verifies: your individual contribution inside a team result.
• Triggered by: any project, launch or outcome that clearly took more than one person.
• Passing answer contains: who was involved, by role; two or three specific actions you took, as verbs; and where your part ended and someone else's began.
• How it misleads: interviewers sometimes count "I" against "we." That distinction often clarifies contribution but doesn't settle it on its own, and repeated "I" on a team-based initiative is itself a published red flag.

Situation probe

• Sounds like: "What factors led up to the situation?"
• Verifies: that you understood why the work existed, not just what you did.
• Triggered by: any bullet that starts with a result and skips the problem.
• Passing answer contains: the trigger, the constraint you worked under, and why it mattered to the business.
• How it misleads: a long backstory feels thorough but uses up the time the interviewer set aside for probing your actions.

Outcome probe

• Sounds like: "What was the outcome?"
• Verifies: that the story ends in a result that can be checked.
• Triggered by: bullets that describe activity ("built," "ran," "supported") with no result.
• Passing answer contains: the result, how it was measured, and what you would do differently.
• How it misleads: an outcome with a number in it immediately triggers the five metric probes. Don't add a number here that isn't already on your resume.

Timeline walk

• Sounds like: "Walk me through the timeline."
• Verifies: that the story holds together in time. Dates and phases should come easily if you lived them.
• Triggered by: date ranges, multi-phase projects, short tenures. Topgrading builds a whole chronological interview around it, described as a "4-hour tandem Topgrading interview."
• Passing answer contains: a start date, the phases in order with rough durations, and an end date that matches the resume.
• How it misleads: hesitating over dates you never wrote down reads like invention. That's a preparation gap, not a credibility gap, so close it before the round.

Manager-view probe

• Sounds like: "What would your manager say you contributed?" In Topgrading's Threat of Reference Check (TORC), the interviewer collects each boss's name and asks "What will so-and-so say about you?", sometimes adding what score from 1 to 10 each manager will give "when" (not "if") they call.
• Verifies: that your account matches what a third party would say. In TORC, it turns your answer into a claim the reference call can test.
• Triggered by: vague answers to the role probe, or any high-weight claim.
• Passing answer contains: the manager's name, what they would credit you with, a realistic weakness they would mention, and evidence they could check afterwards.
• How it misleads: this sounds like a self-assessment, but it's a verification hook. An inflated score is a prediction that can be proved wrong.

Proof request

• Sounds like: "Is there anything that documents this?" Crosschq names this probe the "Proof Request": documentation, performance reviews, emails from leadership.
• Verifies: that the claim left a trace outside your memory.
• Triggered by: large or unusual claims, and claims the role depends on.
• Passing answer contains: which artifact exists, what you can share without breaching confidentiality, and who can confirm the rest.
• How it misleads: having no shareable document isn't failing. Most work artifacts are confidential. Naming a person who can confirm the claim usually does the same job.

Probes the published guides don't name

Three follow-ups candidates report all the time have no named row in the guides I checked: choice rationale ("why that tool?"), implementation depth ("how did that work under the hood?") and setback ("what went wrong?"). What follows is my reading, not a published standard.

• Choice rationale most likely checks whether you made the decision or inherited it, which makes it a variant of the decision-authority probe. Pass with the options you considered, the deciding constraint, and who signed off.
• Implementation depth checks the false-precision failure covered below: technical detail you can't explain. Pass with one level deeper than the resume line, and a plain "that part was handled by another engineer" where that's true.
• Setback usually tests the repeatability and concurrent-change checks from the other direction. Pass with what broke, what you changed, and what the result was afterwards.

CLAIM (exact resume wording):
VERSION SUBMITTED (file or posting):

METRIC CHECKS (fill if the line has a number)
  Baseline value + period:
  Result window (start to end):
  Measured in (report / tool / attribution model):
  Concurrent changes in the same window:
  Held, repeated, or one-time spike:

OWNERSHIP
  Who was involved (by role):
  My specific actions (verbs):
  Decided alone:
  Needed sign-off from:

TIMELINE
  Start / phases / end:

THIRD PARTY
  Manager name:
  What they would credit me with:
  Weakness they would name:
  Score I expect (1 to 10):
  Artifact that exists / who can confirm:

STATUS: supported / unclear / contradicted

How to prepare a resume line for probing, step by step

Prepare each high-weight claim the way the interviewer will test it: list the claims, rank them, fill the probe slots, check them against your public record, and sort them before the debrief does it for you. Each step below matches a stage of the interviewer's process.

Step four is where tailored resumes cause trouble. If you've sent five versions of the same history, the 4.3 in one can easily become a 4.5 in another. Refolk tailors each version from the same underlying history, so the figure a probe lands on is the same figure in every version an interviewer might have.

For the live moment when you don't have a fact, keep honest wording ready so you don't improvise a number.

Baseline unknown:
"I don't have the exact starting figure in front of me. It was in the [report name] for [period], and the change was from roughly [range]. I can confirm the exact number afterwards."

Team result:
"That was a team of [number] - [roles]. My part was [action 1] and [action 2]. The [specific decision] was mine; [other decision] went to [role] for sign-off."

Concurrent change:
"[Other change] happened in the same window, so I wouldn't credit the whole result to my work. What I can isolate is [the part you can isolate and how]."

Detail outside your part:
"That layer was built by [role]. What I owned was [your layer], and I can go as deep as you like there."

How probing goes wrong: failure modes and false positives

The most expensive mistakes are reading a hard probe as doubt, and inflating the exact point being pressed. Both come from treating the probe as a judgment instead of a pre-written check. Each failure mode below comes with the check that catches it.

The research behind the biggest one deserves a careful statement. Researchers expected follow-up questions to reduce deceptive impression management (presenting yourself as better than you are through false or inflated claims). They found the opposite: follow-ups increased it, because probing was "perceived not as response verification but rather as a cue signaling what types of answers were important and critical." That finding comes from the research program behind the 2007 interview faking behavior scale, which was built and validated across 6 studies (N = 1,346). Reviews disagree on which study produced the probing result, one citing Study 5 and another Study 6, so cite the finding rather than a study number. A related review found that 93% to 99% of undergraduate applicants had used at least one faking tactic, and that situational questions drew more faking than past-behavior questions.

1. Reading hard probing as doubt. Probes are often written before the interview and asked of everyone, so how hard one lands isn't a verdict. Check: would this question make sense, worded the same way, for any candidate with a line like yours? If so, it's a planned probe.
2. Embellishing where the probe lands. Probing works as a cue, so candidates inflate exactly the point being pressed. Check: does the number you're about to say match your resume and your LinkedIn profile?
3. Counting "I" to judge ownership. The "I" versus "we" distinction isn't conclusive on its own. The false positive is a candidate who says "I" all the way through a team project, which Forbes flags as a warning sign for clearly team-based initiatives. Check: name who was involved before you name what you did.
4. Treating a baseline as a single number. A baseline with no measurement window, or with concurrent changes ignored, fails the confounder probe. Check: can you say the period the baseline covers and one thing that changed alongside your work?
5. Quoting false precision. Adding an exact percentage, timeline, title or technical detail you can't explain. Check: for every precise figure, can you name the report or system it came from?
6. Treating TORC as a bluff. In Topgrading, the reference check is real and is used to double-check the career story. Check: would your manager give the score you predicted?
7. Citing a single study version. If you mention the probing research in an interview or a write-up, reviews disagree on which study it came from. Check: state the finding, not the study number.

Who is doing the probing, and does anyone check afterwards?

Most deep-dive probes come from trained recruiters and hiring managers, and they far outnumber the coaches who prepare candidates for them. Whether claims get checked after the interview depends on the employer's method. No public survey isolates post-interview checking.

In Refolk's index of professional profiles, recruiter supply is large and concentrated in the US:

The UK count may be low because of title variants such as "Talent Partner." That's unverified, so treat the multiple as a rough guide. Ten of the 25 UK profiles sampled are in London.

Set that against the people who prepare candidates:

The practical point: the person probing you has usually done it many times, and you're probably preparing alone. The prep sheet above is meant to close that gap.

What's known about verification is partial and sometimes contradictory:

• A Checkr survey of 3,000 US managers found 60% had caught candidates lying about qualifications and another 13% suspected it.
• CareerBuilder reported that 75% of HR managers spot resume inaccuracies.
• Greenhouse found 39% of hiring managers are doing more in-person interviews to confirm identity.
• In a survey of 200+ HR professionals, 86.1% check candidates' online presence at least occasionally and 27.2% always do.
• Robert Walters found 64% had viewed a candidate's professional networking profile, while Fast Company reported only 29% of hiring managers check LinkedIn. The figures conflict, so don't lean on either.

Topgrading is the clear exception, where verification is announced and certain. It says to tell candidates at every step: the careers page, the application form and the introduction to every interview. A Topgrading reference call runs about 30 minutes. If you see that language anywhere in a process, treat every manager-view answer as one that will be checked.

Before the next round: checklist and keeping the sheet current

The night before a deep-dive, run the checklist below against your top three to five claims. After the round, update the sheet with the probes you actually got, because the next interviewer at the same company often works from the same guide.

To keep the sheet current:

• Log the probes you got. Write the exact wording of each follow-up and which row it maps to. Planned probes repeat across candidates and rounds.
• Re-sort after every round. Any claim that came out unclear goes back to step three before the next interview.
• Re-check after every resume edit. Tailoring a line for a new posting resets its prep sheet. When Refolk drafts a tailored version, check the changed lines against the sheet before you send it.
• Re-check the verification picture yourself. Survey figures on how often employers check claims change and conflict. If it matters to a decision, look for the original survey and its sample, not a secondhand summary.