It started last summer. Clothes adverts flooded my Instagram feed, and it was uncanny. One minute, I’d be envying a friend her Agnès B boiler suit (it has a bumflap, so you don’t have to take it off when you go to the loo: so French); the next, I would be served an ad for a more or less identical boiler suit, except without the flaps, and only £24.95. Had Instagram accessed the microphone on my phone and listened to my chitchat? But I hadn’t even spoken that envy out loud. They had seen inside my head.
I wanted dungarees, because they reminded me of the 80s, when I had a babysitting racket and spent all the money on them. I also wanted a maxi dress, because it seemed to meet my newly acquired aversion to tights. These things, while not profound, were intensely personal, or so they seemed. And yet there they were, telepathically, on my feed.
Steadily, I also started to see people wearing these clothes in real life. The world and her dog suddenly had the dungarees. When I went into the kind of bar whose uniform would once have been tight-jeans-fancy-top, I was confronted with what looked like the warm-up act for the Nitty Gritty Dirt Band. And even though this did blow a hole in my presumption of idiosyncrasy, now that everyone had my stuff, I wanted it more.
Eventually, I overcame my luddite objections (I still dislike buying anything I haven’t touched, ever since I bought two really cheap light sabres on eBay that turned out to be Darth Vader chopsticks) and clicked. As a sort of experiment, for a month, each time a perfectly targeted ad popped up with something I definitely really wanted, I caved and clicked.
I ended up buying six things, always on Instagram. Fashion brands confirm this is how the business model works: Instagram for buying; Facebook for “building community”; Twitter for complaining. The aforementioned dungarees (weirdly priced at £23.21 by the time I hit “buy”), plus a fringed shirt that looked like cheesecloth (£38.10), a pair of leggings (£24 for two pairs) and a denim dress that looked flattering (£22). Also, a maxi dress (£24.76) and a boiler suit (priced in dollars, for some reason: $41.60). Weeks went by, with nothing through the door. I realised all the items were being manufactured in Asia and shipped very slowly (something the brands obscured until after I’d added my details). This has no quality implications (plenty of brands manufacture in Asia) but it did seem to cast us back to the 19th century, in delivery terms; they seemed to be coming by slow ship. There seemed to be no middlemen, no distinctive visual signals of a marketing or import platform, no “thank you for your purchase”.
Three things never arrived (the money went out, and came back in again, and no fraud was perpetrated) and three did: one stripy maxi dress; the dungarees; and the boiler suit. The brands weren’t proper words, they were strings of letters (think Koioiloaboy) and, as it turned out, the clothes weren’t proper clothes, either.
The dress came first. There hadn’t been a huge amount of detail in the picture, but it looked as if not much could go wrong: vivid stripes, floor length, simple shirt-shaped top. I’d pictured it going with everything, too bold to clash. My initial worry on opening the package, however, was the fire hazard. I’ve never seen material like this used to clothe people; it looked more like the stuff you’d use to mask the electric wires at a festival. I didn’t even notice how thin it was, how much like being naked it was, until I left the house.
I wore it to interview the Brexit vote campaigner Gina Miller, live on stage. The tailoring was so amateur that the space between the buttons gaped when I sat down. “You look so colourful,” Miller said. “I wish I’d worn more colour.” It was an extreme test of English manners: the person you’ve just met looks so absurd, so like a clown, that the only way to cover your astonishment is to pretend to wish you looked like a clown, too.
The boiler suit came a fortnight later (the manufacturers know something you don’t: waiting for them is better than having them). It was khaki, quite close fitting, and there was no obvious way in which it differed from the photo. Except the material was thinner than any fancy dress costume, fractionally less well-made, and completely unadorned. It looked like prisoner-issue gear in a North Korean camp. After a very simple journey – home, tube, bar – there was a hole on the seam and you could see my pants. It went beyond the point of poor quality; wearing it was like exposing yourself as a loser in the game of life, a veneer of almost-respectability that would be instantly unmasked by anyone who looked too closely.
The dungarees were thicker but simpler, involving nothing so complicated as a button, so you had to tie yourself into them as if you’d made an outfit from a tablecloth and a piece of string. They were made of a very cheap black twill; if they were used as curtains at an event, it would diminish your entire experience of it. Better not to have curtains than cover the windows in this.
I’m not exaggerating: none of these items could survive being worn once; all they were good for was a photograph. It was like a bizarre syllogism: these clothes are in photographs, the photographs look good, ergo these clothes look good. Except they really, really don’t. They were entirely designed to look good in an Instagram advert; then to look good in your own Instagram feed. And while they were cheap, all more or less £25, they would have cost about a tenth of that to produce. I’d reached the logical endpoint of virtual commerce, the point at which life mattered only in two dimensions.
Instagram has had a democratising effect on fashion: there are plenty of small brands that are built for the platform, or rather, Instagram built them. They can swerve all the things you usually need to set up a shop (major finance, infrastructure, bricks and mortar) and focus on a good targeted ad strategy. Some have the authentic tang of artisan fashion. Digital natives can discriminate quite easily between a genuine insurgent brand and my non-brand, non-fashion items, partly by reading reviews, partly because they can read the visual language. Paris Starn, creative director of Paris 99, a reputable Los Angeles-based brand, tells me that “a way for designers to use Instagram thoughtfully is to put the same creativity they have into designing into conceptualising photographs”. Live your values, photograph your stuff in the same spirit as you created it, in other words, and people will want it. Starn explains: “Our SS19 lookbook shoot used friends of the brand to model and photograph the clothing, and took place in a decades-old diner, referencing the designer’s love for Americana culture and baking.” You do not have to be a genius to see the difference between that and a super-airbrushed studio shot of some dungarees.
Dancing Leopard, a UK/Ibiza label that chases most of the Guardian fashion desk round their Instagram feeds, also relies on the social network to take its clothes out of a standard e-commerce environment and into everyday life. According to its marketing director, Rosie Middleton, it differs from traditional fashion marketing in the sense that Instagram users don’t want fantasy; they “want to see models they can relate to, and also other customers wearing the product”. The dud brands I bought into have a different business model; to predict who would like a certain thing so closely that when you produce an unwearable simulacrum of that thing, they will still buy it.
How, exactly, do they do this with such apparent sophistication? When you’re packaging disappointment and sending it halfway across the world (the care labels, where they existed, were in Korean), what’s the business model? Mo White, CEO of brand consultancy business Hyperwild, worked on an early online sales algorithm for fashion, which cannot be named because of a non-disclosure agreement. It could suggest products to people who had already shown a taste for that kind of thing, a far more sophisticated process than the classic cookies model that makes an online retailer think you’ll buy a Hoover for two full years after you’ve just bought a Hoover. “There was an assumption that fashion couldn’t have any structured data applied to it,” White says. “That fashion is creative, that nobody really understands it, it’s all subjective, nobody knows why yellow is in fashion. But to the insider, there are tons of rules. So you know exactly why yellow is in fashion, and why Alexander McQueen is using ruffles.”
If you can teach a machine to recognise and understand different types of airlines and flight routes, then you can teach it to recognise the right pink from the wrong one. Once you’ve built a machine’s understanding, it can triangulate the garment’s attributes with the customer’s, on fixed references: psychographics (personality, values and opinions), demographics and adjacents (what defines that person and where else this person shops). “These are exactly the tools that Cambridge Analytica used, applying commercial values to political contexts,” White says. (There is an immediate symmetry, there: flogging something that sounded attractive to people idling on Facebook that, when it turned up six weeks later, was unbelievably bad.)
This expertise was, in the first place, used by major online retailers: if someone bought something from Next, would they be interested in anything from Burberry? And in the intervening years, this has expanded. Could you tell that a person of a certain age, who also liked dogs and cheese, would also, in this season, like dungarees? Are we all just tremendously obvious? Apparently the answer is yes.
No algorithm can excuse the low quality of the products, though. “The basic rules remain the same: if it’s a crappy product, nobody’s going to use it or want it,” White says. And our overall instincts for telling in advance whether or not something’s crap seem mixed: a recent report from Wrap, a charity promoting sustainable waste management, found that the average lifespan of a garment in the UK is only 2.2 years. Every week, we buy 38m items and 11m go to landfill. In an ideal world, we’d be cherishing an item for a couple of decades; but even if that’s unrealistic, nobody shops for anything to wear once. Do they?
“You should have known what the quality was like from the picture,” my sister observed. “Look at the overlocking on the neck.” I looked again, and saw wonky stitches, rogue bits of thread. It’s possible that a digital native would know instinctively how to distinguish between the bargain and the scam. Yet, Aaron Balick, author of the Psychodynamics Of Social Networking, says that understanding the artifice of an image does not immunise you against it. “Fifteen, 20 years ago, you would speak to young women and teenagers, and you would tell them about airbrushing in magazines, and they would already know it. They would look at those magazine images, know they were airbrushed, and still compare themselves negatively to them. The primitive part of the brain still reacts.
“You present an ideal self online,” Balick continues, “and feel in life that you’re not actually like that. And people are observing you and thinking, ‘I need to be more like that’, so they feel a lack. It’s a contagious lack.”
Of course, that contagion is more about how we present ourselves to one another than how we’re marketed to: you’re going to feel much more inadequate confronted by the perfection of a friend on Instagram than that of a model, and it will be a more textured inadequacy, relating to the whole package: why does she get such great weather on her holiday. Why is her dog so much better behaved?
Yet the sales element represents the beginning of the unravelling: because once you’ve had some seam slippage and felt half naked while out and about, the take-home is not “next time, I’ll check the overlocking on the neck”, but rather, “I will never buy anything on Instagram ever again.”
But that isn’t the effect it had on me: I would still buy anything from a brand I’d found for myself. I would never follow a sponsored ad, just as I would never willingly watch an ad that came before a video I’d clicked on, or indulge an Amazon recommendation. We talk so much about the evil genius of online marketing, the way you’re talking about bras one minute, and the next you have 15 adverts all along your Facebook sidebar. But maybe it isn’t as clever as we think; it found the holy grail of what kind of person wants what kind of thing, but it forgot an advertising staple as old as time: the hardstop of reality, where you use the thing, and it needs to be good.
• If you would like a comment on this piece to be considered for inclusion on Weekend magazine’s letters page in print, please email email@example.com, including your name and address (not for publication).