How much process to show, and to whom

Two readers, opposite needs

There is a permanent argument about whether portfolios should show process. It is unresolvable as stated, because two different buyers are reading the same page with opposite requirements.

A studio hiring you as a contractor wants process. They are trying to work out whether you can think, because they already have people who can execute, and they will be handing you ambiguous briefs. Show them only polished finals and they cannot tell your work from a template purchase.

A shop owner commissioning a sign wants the opposite. Process makes them anxious. Twelve rejected routes tell them the outcome was uncertain, and they are already nervous about spending money with a stranger on the internet. They want to see finished things that look like the finished thing they want.

So the question is never “should I show process”. It is “which of these two people am I trying to reach”, and then: what is the smallest amount of process that makes the case to them.

The test for keeping a process image

Keep a process image only if it explains a decision that the final work does not explain by itself.

That is a hard test and most images fail it. A grid of thirty thumbnail sketches does not explain a decision; it demonstrates effort, which nobody is buying. A photograph of your desk explains nothing. A mood board explains nothing to anyone but you, because mood boards are compressed references whose meaning lives in your head.

Images that pass the test:

• A side-by-side of the constraint and the response. The original shopfront photographed at dusk, next to the new mark at the same distance and light. This explains why the letterforms got heavier.
• A rejected route with one line about why it lost. “This version tested badly with older customers, who read the abstract mark as a different shop.” This is the single most persuasive kind of process image, because it shows judgement rather than production.
• A technical diagram the outcome hides. The colour separation, the die line, the responsive breakpoints, the frame where the animation had to hold for the voiceover.

Three to five images that pass the test beat forty that do not, and they beat them on the same page for the same reader, which is the part people find counter-intuitive.

The ordering that works

Put the finished work first. Always, for both readers.

The reason is mechanical: on a portfolio page, most readers decide whether to continue within the first screen, and the finished work is the only thing that earns the scroll. Process-first case studies (opening with research photos and sticky notes) lose the audience before the payoff, and they lose the studio reader too, who has forty portfolios to get through and is scanning for a reason to stop.

So: hero image of the finished thing, a short line of context, then process, then more finished work, then the write-up. Anyone who reaches your process images has already decided the outcome is good enough to be curious about how it happened. That is the only state in which process is interesting.

The specific trap: showing work you do not want

Every rejected route you publish is a route somebody might ask for. This happens more than people expect. You show three directions, note that the client chose the third, and the next enquiry says “I liked the second one in your bakery project, can you do something like that for us”.

Sometimes that is fine. Sometimes the second route was a weak idea you included for contrast, and now you are being hired to repeat your own worst work. If you would not want more work like it, do not put it on the internet. This applies to finished pieces too, and it is the single most common reason an established freelancer’s enquiries drift away from the work they want: the portfolio is a menu, and people order from menus.

Doing it for nothing

Process documentation costs nothing but attention, and attention is the thing you will not have at the end of a project.

The practical habit: at the moment you finish each stage, export one image and drop it in a folder named for the job. Not later. At the end of a project the intermediate files are buried in versioned folders with names like final_v7_ACTUAL.ai, and reconstructing the story takes an afternoon you will never spend.

For screen capture on a phone or laptop, whatever is built in is enough. For desktop screen recording, OBS Studio is free on every platform and records a 1080p timelapse of you working without a watermark or a time limit; the paid alternatives at this task buy you convenience, not capability. For mocking up a printed piece, Blender is free and will render a folded carton or a book in context far better than a flat photo.

One caution on timelapses: a 45-second speed-up of you drawing is satisfying to watch and communicates almost nothing about your judgement. It performs well on social platforms, which reward watching, and poorly with buyers, who are hiring a decision-maker. Make them if you want reach. Do not confuse the reach with enquiries.

The limitation

Process pages have a failure mode nobody warns about: they make ordinary work look considered. A designer who documents heavily can present a mediocre outcome wrapped in enough reasoning that it reads as rigorous, and some buyers will be persuaded. This works until the second project, when the work has to be good rather than well-explained.

The check on yourself is simple. Cover the words. Look only at the images. If the finished work does not hold up in silence, the process pages are not a strength: they are a defence, and you are the person they are defending against.