
Five years at a 10-person startup is different from five years at a 5,000-person enterprise. How company context changes how you read a CV.
Same title. Same years of experience. Same skills listed.
One candidate spent those years at a ten-person startup. The other at a five-thousand-person enterprise.
They are not interchangeable โ and the role you're filling decides which one you want. We've covered reading claims and auditing evidence. This is the layer that decides fit: where the experience happened.
People are shaped by where they work. A company's scale, pace, and structure decide what "good" means there โ and after five years, a candidate's instincts match their environment.
Put the startup engineer in a heavily governed enterprise and they chafe. Put the enterprise engineer in a founder-run business where "we'll sort the process later" and they freeze. The skills are identical on paper. The fit is the opposite.
That's why context predicts what skills lists can't: not just whether someone can do the job, but whether they'll survive and thrive in your client's version of it.


Profile a past employer and you get five answers at once:
Scale
Did they build things at 10 users or 10 million? Scale changes how people think about risk, testing, and what "shipped" means.
Pace
Startup speed or enterprise process? One rewards shipping fast and fixing later; the other rewards getting it right the first time.
Structure
Were they a specialist on a 30-person team, or the whole department? One learns depth; the other learns everything.
Sector
Regulated or fast-moving? Finance and health build compliance instincts. Startups build speed instincts. Your client sits somewhere.
Stage
Joining at series A โ or joining at 2,000 people โ is a different skill set. Early joiners build from scratch; later joiners scale what exists.
Five questions. Three minutes of research. It's the cheapest intelligence in screening.
Take "Software Engineer, 5 years":
Startup context: "Third engineer at a 12-person fintech. Built the payment integration, ran production deploys, picked the tools." Enterprise context: "Backend engineer in a 40-person platform team. Owned one service in a 200-service estate. Changes went through two review gates."
Both excellent. But put the first into a heavily-governed environment โ or the second into a founder-run business โ and both struggle. The CV structure is identical. The fit is where the difference lives.
โ ๏ธThe two failure directions
Startup person โ enterprise client. Brilliant, but unused to governance. They'll fight the process and leave. Enterprise person โ startup client. Solid, but unused to ambiguity. They'll wait for the process and stall. Either way: a placement that boomerangs, and a client who remembers.
Our running role: Senior Cloud Security Engineer, energy-sector client. Two candidates, and on the evidence audit they look equal โ both have real Azure security trails, similar years, similar outcomes.
Candidate A did it at a five-person e-commerce startup. Fast pace, no governance, security was "whatever we can afford this sprint."
Candidate B did it at an energy trading firm. Every change reviewed, regulators in the building, security is a licence to operate.
For this client, B is the stronger fit โ not because B is the better engineer, but because B's environment is the client's environment. A might be the better hire for a fast-moving fintech. That's the whole point: context doesn't rank candidates absolutely. It ranks them for this role.
A keyword matcher can't see any of this. A recruiter who reads context sees it in the employer names before they finish the first page.
Fit isn't a nicety โ it's the retention stat that pays your desk. A placement that boomerangs in three months costs the client their confidence and costs you the fee, the refund conversation, and the repeat business. A placement that fits lasts โ and lasting placements are what turn one-off clients into retained relationships.
You already know this from your own placements. Context reading just makes it a deliberate part of the decision instead of a gut feel after the interview.
Context reading has its own failure modes:
Speed matters, same as evidence reading. For each shortlisted candidate:
Three minutes per candidate, done while you read. The answer goes into the same one-sentence defence you're already building: "Eight years Azure security, energy sector, led the migration." The sector half of that sentence is context reading โ and it's the half that predicts whether the placement lasts.
Skills tell you what someone can do. Context tells you where they've done it โ and whether your client is that kind of place. Read both, and your shortlists stop being lists of qualified people and become lists of people who'll actually stay.
AI reads CVs like a recruiter: skills with evidence, career trajectory, employer context. Three-stage matching from code gates to full evaluation.

A candidate listing 'Python' tells you nothing. Used it at BP for 4 years building REST APIs? That tells you everything. How to read CVs for evidence, not keywords.
Read article โ
Full recruitment process, done for you. How Expera works โ from job ad to professional shortlist.
Read article โ
Most CV screening tools match keywords. Here's why reading CVs with context โ skills with evidence, career trajectory, employer context โ finds better candidates.
Read article โ