The third pillar of our standard is a documented record of commercial results. In practice, most of the resumes we see have the results and not the documentation — and the gap between those two is where good candidates lose to weaker ones who present better.
This is the standard of evidence a client applies, and how to meet it.
The four-part test
A claim gets believed when it has all four of these. Missing one weakens it; missing two means it is read as a responsibility rather than an achievement.
A number. Not “significantly improved” or “drove efficiency.” A figure, a percentage, or a count.
A baseline. Improved from what? A number without a starting point is unreadable — a 15% reduction could be recovering ground you lost or it could be best in the industry.
A timeframe. Over what period, and was it sustained? An improvement that reversed the following year is a different claim.
An attribution. What did you do that caused it? This is the one most resumes skip entirely, and it is the one a hiring manager cares about most.
Compare:
Improved plant efficiency and reduced costs.
Cut scrap from 4.2% to 2.6% over eleven months by rebuilding the changeover procedure and retraining three shifts. Held at 2.7% through the following year.
Same achievement. The second is checkable, and it also tells the reader how you think — which the first does not.
Attribution versus participation
This is where this population loses the most ground, and it is cultural rather than a skills gap.
A career where credit belongs to the team produces people who write “we” and describe their contribution at the level of the unit. A hiring manager reading that cannot tell whether you led the effort, contributed to it, or were in the room. And the default assumption is the weakest of the three, because that is the assumption that makes them least wrong.
You are not being asked to overclaim. You are being asked to be specific, which is the opposite. “I owned the changeover procedure rebuild; the retraining was run by the shift supervisors” is more credible than either “we improved scrap” or “I reduced scrap,” because it shows you know exactly where your contribution ended.
A useful sentence pattern: “I was responsible for X. The result was Y. The part I would not claim is Z.” That last clause is unusual on a resume and it does more for your credibility than anything else you can write, because almost nobody does it.
Three rules for the number itself
Round honestly, in the direction that costs you. If it was 14.6%, say 14%. A suspiciously precise number invites arithmetic; a conservatively rounded one invites belief. If somebody later finds the real figure was better, that is a good discovery for them to make.
Give the denominator. “Reduced overtime by $400,000” means nothing without knowing whether the base was $600,000 or $40 million. Include it.
Never present a forecast as a result. “Delivered a program projected to save $2M annually” is a plan. If the savings have not been realized, say so — “on track to save $2M annually, $600K realized in year one” is both honest and stronger than the vague version, because it shows you know the difference.
The confidentiality problem
A great many of the best results are inside numbers you are not permitted to disclose. This is real, it applies to classified work and to ordinary commercial confidentiality, and it is solvable.
Use ratios and percentages instead of absolutes. “Reduced cost per unit 19%” discloses nothing about volume or revenue.
Use bands. “A P&L between $100M and $150M.” Nobody needs the exact figure to size your scope.
Describe the scale in units nobody owns. “Eleven sites, four states, 68 technicians” is often enough, and headcount is rarely confidential.
Say when something is constrained. “I cannot give the revenue figure, but the business was in the top quartile of our segment and grew for six consecutive quarters” is a credible sentence. An interviewer reads a stated constraint as discipline. What they read badly is a vague claim with no explanation, because they cannot tell whether you are being careful or being vague.
The same approach works for classified work. Scale, duration, decision authority, what went wrong, what you would do differently — none of those require any detail about the work itself.
What a reference has to be able to confirm
Most people choose references for seniority. Choose them for proximity to the claim instead.
A director who worked alongside you daily and can say “yes, he owned that, here is what it was actually like” is worth several VPs who can confirm you were well regarded. The second kind of reference produces warm generalities. Warm generalities do not move a hiring decision at this level.
Before you list somebody, ask yourself: can this person confirm the specific claims on my resume? Then — and this is the part people skip — tell them what those claims are. Not to coach them. So they are not surprised, and so they have the figures to hand rather than saying “I think it was around thirty percent?”
If nobody can confirm a result, reconsider leading with it. A number that collapses at the reference stage does more damage than never claiming it, because it recasts everything else you said.
Build the file before you need it
The single most common avoidable problem we see: somebody has genuinely delivered a result and cannot evidence it, because they left the company, the people who could confirm it have moved on, and the reporting they would have pointed to is behind a login they no longer have.
Keep a running file, updated quarterly, with:
- The result, in the four-part form above
- The source of the number and the date you read it
- Who could confirm it, and where they are now
- What you would say if you could not use the raw figure
Ten minutes a quarter. It is worth more than any resume rewrite, and it is the thing you cannot reconstruct later.
What good looks like
A strong record at this level is usually three to five results of this quality across two roles. Not twelve. Twelve bullet points of similar weight reads as a job description, and the reader cannot tell which ones you are proud of.
Pick the ones where your attribution is clearest and the numbers are hardest, lead with those, and let everything else be context.
If you are not sure whether what you have clears the bar, send it to us. A person reads it. Frequently the answer is that the substance is there and the presentation is hiding it — which is a good problem, and a fixable one.