Reading a Freelancer's Work History in Five Minutes
When a shortlist has ten names on it, you can't spend an hour on each profile. You need a quick, consistent way to read someone's work history and decide whether to talk to them. These are my notes on doing that in about five minutes per person.
The idea is simple: look for patterns, not highlights. One impressive project tells you less than a steady record of finished ones.
Minute one: does the history match the job?
Start with relevance. Ignore everything else for now.
- Scan the list of past projects or contracts.
- Count how many resemble your job in type, size or stack.
- Note the most recent relevant one and how long ago it was.
If nothing in the history resembles your project, that isn't automatically a no, but it moves the person down the list. Someone who has built five booking systems will likely handle your sixth faster than someone who's built none.
Minutes two and three: look for patterns
Now look at the shape of the history as a whole. A few signals are worth more than any single line:
- Finished work. Are projects completed, or does the history show a lot of starts?
- Repeat clients. People who hire the same freelancer again are giving a strong, quiet signal.
- Delivery on time. If the record shows milestone dates, check whether work was delivered when promised.
- Steady activity. A consistent record over months suggests reliable availability. Long gaps aren't bad, but you might ask about them.
- Project size. Does the person usually take short tasks or long engagements? Match it to what you need.
Be aware of where the history comes from. A self-written list of past clients is a claim. A record built from actual recorded contracts is closer to evidence. On some platforms the difference is explicit: a verified history lists only contracts recorded on the platform, past clients can confirm an engagement from their own account, and cancelled contracts don't count toward the totals. If you're curious how that kind of record is put together, you can learn more about how one freelance marketplace approaches it.
Minute four: read one project closely
Pick the single most relevant project and read it properly.
- What was the freelancer's actual role?
- How long did it take?
- Was it delivered in stages, and were the stages on time?
- Did the same client come back afterwards?
- Is there a portfolio piece, demo or document attached?
One close read is enough. You're checking that the summary holds up, not auditing a career. If the project description is thin, that's fine; note it as something to ask about rather than a reason to rule the person out.
Minute five: note questions, then decide
Write down two or three questions for the call, then make a quick call on whether to talk.
Good questions come from gaps you noticed:
- "Your last three projects were mobile. How much web work have you done recently?"
- "This project ran longer than planned. What happened?"
- "You work with a few repeat clients. What kind of work do they come back for?"
Decide with a simple scale:
- Talk now: relevant history, solid patterns, questions are minor.
- Maybe: some relevance, a gap worth asking about.
- Skip: little relevance and no strong patterns to compensate.
Keep notes in the same format for everyone. It makes comparisons fair and stops you from favouring whoever you read last.
A few things to keep in mind as you use this method:
- Newer freelancers will have short histories. Weigh their portfolio and their proposal more heavily, and consider a small first milestone.
- Privacy. Some clients ask to stay anonymous, so a project may show as a private client. That's normal and not a warning sign.
- Earnings figures say something about experience, but little about fit. Many freelancers show a range or hide totals entirely, and that's reasonable.
- Ratings and reviews where they exist can be useful, but they're easy to inflate. Concrete history, like completed contracts and repeat clients, tends to be harder to fake.
Five minutes per person won't catch everything. It isn't meant to. It's meant to get the right three people onto a call, where you can ask the questions that a profile can't answer.


Comments 0
Replies to this note, oldest first.
No replies yet. Found it useful? Say so below.