How to curate a portfolio from years of finished work

Most designers, photographers and illustrators who have been working for ten years are not short of good work. They are drowning in it. The problem a decade in is never finding enough pieces, it is that everything from the last ten years is sitting in one undifferentiated pile, and a portfolio built by scrolling back through all of it and grabbing what still looks decent is not a portfolio. It is an archive with a nicer font.

Curating one down is a different exercise from making the work in the first place, and it rewards a different kind of honesty.

How many pieces should actually be in a portfolio?

The honest answer is fewer than most people bring themselves to publish. A portfolio with thirty pieces does not read as thirty separate proofs of skill, it reads as one long scroll that a viewer stops paying real attention to somewhere around piece six. AIGA Baltimore's own advice to design students is blunt about this: it is better to have five to seven amazing projects than ten to fifteen that are not so great. A hiring manager, an art director or a client giving you a minute of attention forms an opinion from your strongest handful of pieces regardless of how many you show them, so anything past that handful is only diluting the average.

Somewhere between ten and fifteen finished pieces is a reasonable target for most careers, closer to ten if you are early and closer to fifteen if you genuinely have several distinct ranges to prove. The exact number matters less than the discipline behind it. Every piece included has to earn its place against the others in the set, not just against the empty space where a portfolio used to be thin.

What you are proud of is not always what got you paid

Run two honest lists side by side. The first is the work you are actually proud of, the pieces you would show a peer without wanting to explain away a weak brief or a client who fought you the whole way through it. The second is the work that paid well, kept a client coming back, or built your reputation in a niche. For most people ten years in, the overlap between those two lists is smaller than they expect walking into the exercise.

That gap is not a flaw in the work. It is the ordinary shape of a career. The project you are proudest of might have paid badly, or not at all, because that was the client who gave you room the paying ones never did. The project that funded two years of rent might be technically excellent and creatively forgettable, executed cleanly against a brief that never asked for anything interesting. A portfolio built only from the paid list looks competent and says almost nothing about you. A portfolio built only from the pride list can look thin if there is not yet enough of it. Pick mostly from the overlap, fill any remaining gaps from the pride list, and only reach for something that purely paid the bills if it happens to be technically exceptional as well.

Should you group by client, by date, or by something else?

Chronological order tells a viewer what year you made something. Client order tells them who signed the invoice. Neither answers the question they are actually asking, which is what kind of person you are to work with and what you are good at. Group your survivors instead by the story each cluster tells: a run that shows range across a single brand voice, a set that proves you can hold complexity together without the layout falling apart, a pair that shows the same problem solved two different ways. A viewer scanning a portfolio arranged this way draws a conclusion in the first thirty seconds, which is roughly all the time most portfolios are given.

This has a useful side effect once you start doing it. A piece that does not obviously belong to any story you are currently telling is a piece you probably should not be including, no matter how proud of it you still are or how much it once impressed somebody. If it does not support an argument, it is decoration, and decoration is exactly what is crowding out the pieces doing the real work.

Retire work even if it was once your best

This is the part most people resist. A piece that won you a client, an award or a following two years ago earns a kind of loyalty that has nothing to do with whether it still represents your current level. Keep it in out of habit and it quietly starts doing the opposite of its old job. Instead of proving how good you are, it shows a viewer the ceiling you have already moved past, and pulls the average of the set back down toward it.

A portfolio is not a trophy case. It is evidence of where you are right now, today, not evidence of what once impressed someone. Review it on a schedule, once a year is enough for most people, and ask one plain question of every piece in it: would you make it this way now? If the honest answer is no, take it out, whatever it once meant to you. Retiring a piece is not the same as deleting it. It simply stops being part of the dozen you are showing, and goes back to sitting in the full archive with everything else you have made.

A free method that genuinely works, for a small set

You do not need any software to do this well, provided the set stays genuinely small. Make one folder for the portfolio itself, kept separate from wherever you store everything you have ever finished, and number what goes inside it in the order you want it seen rather than the order you made it.

1. Create one folder called Portfolio and number each piece inside it 01, 02, 03 and so on, in viewing order.
2. Add a one line note next to each piece naming the story it is there to support, so the reason survives even once you have forgotten it.
3. Review the folder once a year, moving anything that no longer earns its number back out to the main archive.

This works because a numbered folder is honest about what it is, a small, deliberately ordered set, not a home for everything you have made. It stops working the moment you try to use the same folder to also hold the other few hundred finished pieces sitting behind it, because one folder can only be sorted by one logic at a time, and a portfolio and a full archive genuinely need two different ones running at once.

Where a private working library fits in

The tension underneath all of this is that a portfolio needs two things a single folder cannot give you at the same time: a complete private archive of everything you have finished, and a small curated public set drawn from it. Keep those as two collections in Muse and one item can sit in both without being copied. File a piece into your full archive the day you finish it, then file that same item into a Portfolio collection once it earns a place in the dozen. Dropping it back out later, because it no longer supports a story you are telling, removes it from the public set without touching the archive underneath.

That is the honest fit here. Muse does not do the curating for you, the exercise above still has to happen, and no tool can decide which of your last ten years is your best. What it removes is the file management around that decision, the copying, renaming and re-saving that normally makes promoting or retiring a single piece feel like far more effort than the decision itself deserved.