What the language of "adjustment" reveals about who our products are built for
Home • About • Projects • Blog • Resources • Contact
I’ve always loved creating and building things, and when I chose to leave engineering this left a gap in my life that I’ve since tried to fill with various creative hobbies, and of these the one I’ve become most passionate about is sewing.
I started learning to sew because I wanted to make the clothes I couldn't find in shops, in the size and shape I actually needed. Ready-to-wear clothing generally fits me really badly however carefully I choose the size. Shopping was never really a joy, and I found it hard to find clothes that both expressed my personality but also actually fit. Making my own seemed the obvious solution.
However I ended up finding the same problem waiting for me inside the pattern envelope. I measured myself carefully, chose my size from the chart on the back and cut exactly along the lines, and the first shirt I made had many of the same faults as the clothes I'd been trying to escape. When I went looking for help, I found that the sewing world has a whole vocabulary for this: the full bust adjustment, the sway back adjustment, the narrow back adjustment, the lengthen and shorten lines printed on almost every pattern piece. Most sewists I know make two or three of these adjustments on every garment they sew, which led me to question:
If almost everyone has to adjust the pattern, in what sense is the pattern the standard?
What’s a ‘normal’ body shape and size?
In 1951 the UK Clothing Industry Development Council measured around 5,000 women (whilst, it’s worth mentioning, food rationing was still in force) and the results became the basis of British Standard sizing. In the US, standard sizes were built on a survey from 1939-40 of almost 15,000 (exclusively white!) women. Standard pattern sizing hasn’t really changed since and are still drafted for a B cup, even though most of the evidence we have suggests that this is FAR from standard.
When SizeUK scanned 11,000 people between 2000 and 2004, it found that the average UK woman was a size 16 and that 60% of shoppers struggled to find clothes that fit. There hasn't been a national survey since, and most of what we now know about women's bust sizes comes from companies that sell bras. I think that gap is part of the issue:
The people a standard fails are rarely the people anyone goes back to measure.
The language of adjustment
What has struck me most about the vocabulary of adjustment is where it places the problem.
The pattern is the default, and your body is the deviation that requires an adjustment.
The philosopher Georges Canguilhem once argued that what we call normal is a norm someone chose, which then comes to be treated as a description of nature (The Normal and the Pathological). Disabled activists in Britain have reached the same conclusion from a very different starting point; in 1976 the Union of the Physically Impaired Against Segregation argued that people are disabled by environments built for someone else (something those of us who are neurodivergent feel only too acutely).
Tech has its own vocabulary of adjustment: we talk about edge cases and ‘personas’. UK employment law talks about ‘reasonable adjustments’. Each of these is well meant, and reasonable adjustments in particular have made working life possible for a great many people; however, each one also locates the mismatch in the person rather than in the design, and I think that framing shapes what we go on to build.
Poor fit hides itself
The part that took me longest to see is that poor fit hides itself. When clothes didn't fit me, I never concluded that a survey from 1951 was the problem; I concluded that my body was. I saw how widely that instinct is shared on a course about altering trousers, when we went round the room and each participant explained what needed fixing. Every one of them, including people most would consider to have enviable bodies, described what was wrong with their body, with a clear sense of shame; nobody described what was wrong with the trousers or the pattern. Most women I know could write a similar list of their defects and flaws; and we deal with it by sizing up, altering, or giving up on certain shapes altogether, none of which registers anywhere as a complaint.
When I learned in my forties that I have ADHD and autism, I realised I'd spent forty years doing the same thing to my mind. I had treated every mismatch between how I work and how workplaces expect people to work as a personal flaw: something to apologise for, compensate for and make smaller. Understanding the mismatch didn't make it disappear, but it moved the problem to where it actually sat, which changed how I understood both my difficulties and what I have to offer.
I think products fail people in exactly the same way. Users who don't match the assumptions we built for rarely file a bug report; instead they decide they're not very techy, find a workaround, or quietly stop using the product. Because the people a standard fails tend to blame themselves rather than the standard, the signal never reaches the people who could fix it, which is how a set of measurements taken in 1951 can shape what women wear for more than seventy years.
Hearing the workaround
This is why I think the way we listen to customers matters more than any single piece of research. In Continuous Discovery Habits, Teresa Torres recommends story-based interviewing: rather than asking people what they want or what's wrong, you ask them to walk you through the last time they did something. When people are asked for opinions, their answers pass through the same self-blame that stops them complaining; when they're asked for a story, they show you their workarounds, often without realising that's what they are.
A workaround is a product's full bust adjustment.
It’s gap between the default and the person that the person has quietly absorbed, and I've come to see it as one of the most valuable things a product team can hear.
What AI changes
For most of the last century, I think standardisation was forgivable because variety was expensive: mass production needed a small number of sizes, and product teams needed a small number of personas. AI is changing those economics, in clothing through body scanning and made-to-measure production, and in software through products that can adapt to each person using them. As variety becomes cheap, poor fit stops being a cost constraint and becomes a choice about whose data you collect and whose stories you hear; personalisation built on the 1951 survey would simply automate the 1951 survey at scale. For that reason, and as I’ve argued before, I think AI makes discovery more important, not less.
Drafting my own block
I've since moved on from adjusting commercial patterns to using a personalised pattern block a basic pattern built from your own measurements rather than someone else's average. I drafted mine on a Made to Measure course, which was an amazing experience, and it ended with a toile, the test garment you make before cutting into good fabric, that fitted my body perfectly. I’m now learning about Pattern Drafting. Drafting from scratch is going to be slower and fiddlier than buying a pattern and adjusting it; however, I know that everything I build on that block starts from me rather than from a standard, and I’ll be left with something that meets my needs and makes me feel great.
It's also the best description I've found of what I think good product work looks like: building from the people actually in front of you, rather than patching the gaps in a standard that was never drawn for them.
What's the full bust adjustment in your product, the workaround your users make quietly that should really be telling you something?
