Simon Clements made a promise last year that he has come to regret deeply. After telling his son he could have a PlayStation 5 for his 18th birthday, Clements scoured retailers’ websites daily to track one down. When he confessed he had been unable to buy one, his son, who is autistic, reacted badly.
“We’ve had a lot of problems. There have been tears and tantrums — he doesn’t understand that sometimes you just can’t get things,” said Clements, who owns the Yorkshire Pie bakery in Leeds. His efforts for his son’s birthday this year have once again been fruitless. “All I can do is keep trying,” the 48-year-old said.
Ordinary shoppers such as Clements stand little chance against an unseen army of tech-savvy resellers, who roam the web hoovering up goods in short supply — before flipping them at vastly inflated prices. The key weapons deployed by these 21st-century touts are known as bots — pieces of software that automate the checkout process in less than a second. Resellers set them up to make dozens of checkout attempts simultaneously.
The proliferation of bots is exacerbating the shortages caused by malfunctioning global supply chains. This Christmas, shoppers around the world are being comprehensively outgunned by resellers swiping everything from graphics cards to Pokémon collectibles in the blink of an eye.
The explosion of bots over the past three years is a direct consequence of the marketing tactics employed by Nike and Adidas. The sportswear giants deliberately under-serve huge demand for their trainers by selling limited quantities in hyped “drops” that burnish their brands. That, in turn, has created a booming secondary market where limited-edition trainers sell at prices multiple times the original retail price.
Advertisement
Few have capitalised more spectacularly than 19-year-old Lucas Titus. When he was 14, the London-based Titus bought a few pairs of Ultraboost trainers from the Yeezy line developed by Adidas and rapper Kanye West, and resold them for a profit of up to £150 per pair. And when he started dabbling in automation, things really took off.
“I used Android emulators on my PC to emulate seven different smartphones to tap 1,000 times a second. I wasn’t expecting it to work, but it got me four pairs of Yeezy Zebras which resold for £1,300 [almost ten times the retail price]. That was quite something for me,” Titus said.
He taught himself to code and launched Cybersole, a bot that resellers have used to gatecrash drops, knowing there will always be a collector or “hype beast” willing to cough up more to buy the trainers from them.
Since Titus launched it 3½ years ago, Cybersole has successfully completed two million “checkouts”, buying goods worth $300 million (£225 million). His company had amassed £2.6 million in retained earnings by the middle of last year, allowing Titus, who is now juggling a computer science degree with his own business, to rent a luxury apartment overlooking the River Thames.
He acknowledges, though, that the outlook is uncertain. Hyped drops from Nike and Adidas are becoming more infrequent, spurring resellers to turn to non-fungible tokens (NFTs) and graphics cards, where margins are fatter. PS5s sell for an average of £542 on the resale platform StockX, compared with a retail price of £450. One of Nvidia’s top graphics cards — the GeForce RTX 3090 Founders Edition — resells for £1,838, compared with a retail price of £1,399.
Advertisement
Bot developers charge an upfront fee and a monthly subscription. They have capped the number of users to preserve the value of the product, creating a vibrant resale market for bots where users resell their licence keys for thousands of pounds.
Casey (he did not want to give his full name), the developer behind WhatBot, another sports shoe bot, said some developers have quietly made millions by issuing new licence keys at the resale price. Bot owners and developers also rent their bots temporarily. Acquiring one is just the first step for serious resellers, who go to extraordinary lengths to trick websites into thinking that their purchases are being made by many people in different locations.
Resellers rent proxy servers with different IP addresses for each individual checkout attempt. In a practice known as “jigging”, they configure bots to enter subtle variations on their home address at the checkout (“Rd.” instead of “road”, for example). They use dozens of virtual credit cards with different identification numbers, and link their checkout attempts to shadow Gmail accounts operated by separate bots known as “activity generators”.
Day and night, these bots open emails, send emails, watch videos on YouTube — anything to create the illusion that they are legitimate, discrete accounts. Some resellers go even further by renting high-powered servers from the likes of Amazon and Google, located in close proximity to the retailer’s own server to reduce latency. While some resellers troll the masses by posting pictures of PlayStation 5s stacked high in their living room, most swap tips in closed online chat forums known as “cook groups”.
Curtis Taylor, 26, runs Peachy Pings, a cook group on the social network Discord that charges its 412 members £35 a month. Peachy Pings members reap the benefits of monitors — software that constantly refreshes a retailer’s webpage and pings a notification to the group as soon as sought-after goods are back in stock.
Advertisement
Crucially, notifications arrive about five minutes quicker than publicly available restock alerts on Twitter, providing the edge that resellers need to swipe goods before the masses arrive.
“Some of the people in cook groups are absolute geniuses,” said Ben, a 23-year-old reseller from London who declined to give his real name. “One time, this 14-year-old kid made a script [code used to automate processes] that meant you could buy PlayStation 5s from Argos the day before they went on sale. It took them half a day to figure out what was going on.”
Ben made about £7,000 a year reselling trainers, streetwear and tickets to club nights when he was at university in Nottingham, before getting a job as an actuary after he graduated. Now, he works from home between 9am and 6pm with his cook group open on a laptop next to him. When he hears a ping, he drops what he is doing and quickly snaffles as many PlayStation 5s or graphics cards as he can get his hands on.
Recently, he made an £8,000 profit in a single drop by buying 40 sets of “ultra premium” Pokémon cards, released in limited quantity to commemorate its 25th anniversary.
After reselling goods worth between £250,000 and £300,000 over the years, Ben has set up a limited company and hired an accountant.
Advertisement
“I am not a passionate actuary ... motivation-wise, it’s quite tricky because there are days where I make more money reselling than I do in an entire month from my actual job,” he said.
Developers and resellers acknowledge that their behaviour is controversial, but they claim it helps burnish brands such as Nike and Adidas by exacerbating the scarcity of their trainers. The real problem, they argue, is the lack of supply.
“If you deleted all bots from the internet tomorrow, I don’t think the price of a PlayStation 5 [on the secondary market] is going back to the retail price. The number of people who want a PS5 far exceeds the supply,” said Casey.
For retailers, bots risk alienating loyal customers and cause their websites to crash. Among the most commonly used defences are “captchas”, where customers are asked to click on trains or planes or taxis in a block of pictures. When they identify suspicious behaviour, some retailers ask shoppers to perform object-orientation tasks, such as rotating a picture of an upside-down teapot to its correct position.
“Us developers push an update that improves performance on a certain site, then retailers realise 90 per cent of their stock has gone to bots,” Titus said. “So retailers will change something that breaks the bot, and then us developers have to adjust and it goes back and forth, back and forth like that.”
Advertisement
John Graham-Cumming, chief technology officer at the cybersecurity company Cloudflare, believes the power of machine learning will ultimately mean retailers win out in the game of cat and mouse. He predicts bots will be largely sidelined in the same way that email providers now, in effect, filter spam.
As far as this Christmas goes, though, bots remain a formidable force. Many parents will find that if they truly want their kids to have a PlayStation 5, they will have to pay through the nose for it.
