An achievement does not require a percentage. Describe the problem, your action and the output you can verify. Use counts, coverage or an observed change when reliable records exist, and say when a benefit was intended rather than measured. Leaving out an unsupported metric strengthens the accuracy of the claim.
Decide what the number was supposed to explain
Percentages often stand in for a missing explanation. Before looking for a substitute, ask what you want the reader to understand: how much work you handled, how quality changed, how quickly a task moved or how widely a resource was used. These are different questions. A single attractive percentage cannot answer them all, and a vague estimate can hide more than it clarifies.
Write the relevant claim in ordinary words first. For example, a revised onboarding pack may have brought scattered instructions into one reviewed document. That describes a useful contribution without asserting that onboarding became faster. If speed was never measured, an improvement in organization should not be presented as an established improvement in elapsed time. The sentence can explain purpose while keeping its observed result modest.
Use evidence that fits the claim
Look for records you are allowed to consult: a dated version history, an approved project summary, a published resource or your own contemporaneous notes. A count needs a clear unit and boundary. Five training sessions is meaningful only if the reader can tell whether you designed them, delivered them or scheduled them. Attendance and learning outcomes are separate claims that need their own evidence.
Qualitative evidence also needs care. A manager accepting a document shows acceptance, not universal satisfaction. A customer saying one explanation was helpful does not establish a customer-satisfaction trend. Describe what the record actually supports and avoid converting one comment into a broad conclusion. You can sometimes name the changed artifact itself, such as a consolidated procedure, rather than summarizing an effect you cannot observe.
Rewrite around action and a bounded result
Build a sentence from an accurate verb, the work object and the supported outcome. Add context only where it helps a reader assess complexity or relevance. “Created a single reference guide from separate equipment instructions, reviewed with the facilities lead” is more informative than “dramatically improved efficiency.” It gives you something concrete to discuss and leaves the benefit at a level you can defend.
Review neighboring bullets as well. If every entry relies on a dramatic metric, the whole document may imply a measurement culture that never existed. A mixture of well-supported outputs, task scope and measured results can represent the work more faithfully than a uniform numerical style.
If an estimate matters, keep it in your private notes until you can establish its basis and whether sharing it is permitted. Never reverse-engineer a percentage from a desired story. Also resist replacing an invented number with an equally unsupported adjective such as substantial or transformational. A precise account of an ordinary contribution is useful career evidence; it does not become weak simply because the organization did not measure its impact.
Fictional example: an onboarding pack without a speed claim
A fictional laboratory administrator helped refresh an onboarding pack. An old CV draft says the pack “cut onboarding time by 40%,” but the administrator has no timing records. Their actual work consisted of collecting equipment instructions, removing superseded pages and asking the safety coordinator to review the resulting checklist. They cannot determine how long new staff spent reading it or whether other onboarding changes affected the experience.
The replacement reads: “Consolidated equipment instructions into a reviewed onboarding checklist and replaced superseded pages in the shared starter pack.” The example still demonstrates organization and document control. In interview preparation, the administrator notes that easier navigation was the intended benefit and that no before-and-after timing study took place. A future version could add a measured outcome only if suitable evidence becomes available; the current CV does not borrow certainty from that possibility.
A checklist for your next step
- Identify what a proposed metric would mean.
- Use only counts with a defined unit and scope.
- Keep feedback claims tied to the actual feedback.
- Separate intended benefits from observed results.
- Remove unsupported intensifiers as well as numbers.
Common questions
May I estimate a percentage from memory?
Avoid presenting a remembered impression as a measured result. If an estimate is appropriate in a particular context, label it and explain its basis; otherwise describe the supported work without the percentage.
Will a deliverable sound too ordinary?
Its relevance depends on the role. Explain the problem it addressed, your contribution and any review or adoption you can substantiate. That gives a reader more to assess than an invented performance figure.
Further reading
These primary resources provide related background. They do not validate the illustrative example or endorse AliveCV.
AI-assisted drafting and automated technical checks. Examples are illustrative; no independent expert review or real-world outcome validation is claimed. Read our editorial policy.