
Web3 Loyalty Program Lessons: Why Starbucks, Nike and Reddit Pulled Back, and What Works Now
The big brand web3 loyalty program experiments of 2022 have mostly closed. The ones still running share one habit: they hide the blockchain and reward people for coming back, not for collecting.
Key takeaways
Starbucks closed its Odyssey beta on 31 March 2024, Nike wound down RTFKT in January 2025 and retired the .SWOOSH site on 5 May 2026, and Reddit shut its Vault wallet on 1 January 2026.
The programs that survive treat the blockchain as plumbing. Blackbird's restaurant rewards run on a chain built on Base, yet the consumer pitch is about points per dollar and better tables.
A web3 loyalty program works when it rewards repeat behaviour the brand already wants, keeps rewards useful without a resale market, and has a plan for what happens to holders if it ends.
A web3 loyalty program was supposed to be the friendliest way for big brands to bring customers on-chain. During the last NFT boom a coffee chain, a sportswear giant, a social network and a major airline group all launched one. By the end of 2026 most of them have closed or been quietly handed off, and the few that remain look very different from the launch announcements.
At AnkhLabs we think the closures are useful rather than embarrassing. They show which parts of the idea were hype and which parts still make sense for a brand or a crypto project that wants customers to come back. This article walks through what happened, what survived and how we would design a program today.
What is a web3 loyalty program?
A traditional loyalty program records your points in the brand's own database. A web3 loyalty program records some or all of the rewards on a blockchain, usually as NFTs (unique tokens that prove ownership of a digital item) or as fungible tokens that work like points. In theory that gives customers real ownership. They can keep a badge in their own wallet, trade it, or use it with partner brands that read the same chain.
In practice the pitch had three parts. Brands hoped to build a community around collectibles, earn fees from resales, and attract new customers through the novelty. The first part worked for a small core of fans. The second turned out to be modest. The third faded as the NFT market cooled. Our guide to NFT marketing covers what still works for collectibles on their own.
What happened to Starbucks Odyssey?
Starbucks Odyssey was the flagship. It added collectible NFTs called Stamps to Starbucks Rewards, on the Polygon network, and ran as an invite-only beta for about 18 months. Its first paid drop, the Siren Collection, sold out in 18 minutes, according to CoinDesk, and Stamps sold for as much as $100 each.
The economics were thinner than the headlines. CoinDesk reported that Starbucks likely earned a little over $250,000 from its cut of secondary sales, based on on-chain data, and that Polygon Labs had paid Starbucks $4 million in 2022 to build and host the program, according to two people familiar with the matter. Starbucks declined to discuss the financials. Its later First Store Collection, priced at $100 a token, did not sell out in April 2023.
Starbucks announced the closure on 15 March 2024. The beta ended on 31 March, the Discord server closed on 19 March and the marketplace moved to Nifty Gateway, where holders can still keep and trade their Stamps, The Block reported. The lesson is simple. A collectible layer bolted onto a program people use for free drinks attracted collectors, but it never became a reason for ordinary customers to visit more often.
What happened to Nike's .SWOOSH and RTFKT?
Nike went further than anyone. It bought the digital sneaker studio RTFKT in 2021 and later launched .SWOOSH, its own platform for virtual products. RTFKT was a commercial success by NFT standards. The Block counted nearly $50 million in lifetime NFT earnings, over $45 million in royalties and about $1.5 billion in trading volume across its collections, mostly from CloneX.
None of that saved it. RTFKT announced its wind-down on 2 December 2024 and planned to close by the end of January 2025. Nike then sold the brand to an undisclosed buyer on 17 December 2025, Hypebeast reported, adding that investors in an April 2025 class action had described the shutdown of RTFKT's web3 services as a "rug pull".
The .SWOOSH website followed. Nike told users to move their digital collectibles to another wallet before the site closed on 5 May 2026, with future releases moving to its SNKRS app. In its goodbye post Nike said its goal had been "to explore what creativity, ownership, and community could look like". The change lined up with chief executive Elliot Hill's push back toward retail and in-person events. When the champion of a program leaves, the program needs a business case that stands without them, and this one did not.
What happened to Reddit's collectible avatars?
Reddit's version came closest to mass use. It sold collectible avatars on Polygon, gave each buyer a built-in wallet called Vault, and paid artists 80% of primary sales and 50% of resale royalties. Reddit said the program onboarded "millions" of people, according to Crypto Briefing.
In August 2025 Reddit announced it was ending the creator program. The avatar shop closed on 11 November 2025 and the Vault wallet closed on 1 January 2026, with users told to export their recovery phrase to keep control of their avatars. People can still wear the avatars they own on Reddit, and all secondary sale fees now go to the creators. The report linked the decision to falling sales and the earlier resignation of the head of the program Bianca Wyler.
Reddit is the most instructive case because the onboarding worked. Millions of people got a wallet without noticing. What failed was the reason to keep buying. Once avatars stopped being new, there was little else to do with them.
Which web3 loyalty programs are still running?
The survivors are quieter and more practical.
Lufthansa Group's Uptrip. Launched in spring 2023 with Miles & More, Uptrip lets passengers scan a boarding pass to collect NFT trading cards and redeem full sets for lounge access, upgrades, status or bonus miles. The Lufthansa Innovation Hub says it was handed over to Miles & More in 2025 and is fully run by them. It works because the reward is something a frequent flyer already wants.
Blackbird. The restaurant loyalty and payments app founded by Ben Leventhal raised a $50 million Series B led by Spark Capital in April 2025, with Coinbase and a16z crypto taking part. Diners earn its FLY currency at more than 600 restaurants, and its Flynet chain is built on Base. Yet a Points Guy guide from 10 September 2026 explains points per dollar, tiers and perks across nine destinations without mentioning a blockchain once.
That last detail is the whole story. The programs that last sell the reward, not the technology.
Web3 loyalty programs compared
Program | Chain | Status | What drove the outcome |
Starbucks Odyssey | Polygon | Beta closed 31 March 2024 after about 18 months | Collectibles sat beside rewards, not inside them |
Nike RTFKT and .SWOOSH | Ethereum (RTFKT) | RTFKT sold December 2025, .SWOOSH site closed 5 May 2026 | Strong trading volume, weak fit with the core business |
Reddit collectible avatars | Polygon | Shop closed 11 November 2025, Vault closed 1 January 2026 | Onboarding worked, repeat buying did not |
Lufthansa Uptrip | NFT cards in user wallets | Launched spring 2023, run by Miles & More since 2025 | Rewards tied to flights people already take |
Blackbird | Flynet, built on Base | Operating in nine destinations | Chain hidden, rewards tied to visits and spend |
Sources: CoinDesk, The Block, Hypebeast, Sole Retriever, Crypto Briefing, Lufthansa Innovation Hub, Fortune, The Points Guy. Status as of 6 November 2026.
Why did most brand web3 loyalty programs fail?
Read together, the closures point to four causes.
The reward was the collectible, not the behaviour. A good loyalty program pays people for something the business wants more of, such as visits, flights or spend. Odyssey and Reddit paid people for buying collectibles, which is a different and much smaller audience.
Resale was treated as a revenue line. Royalties looked attractive when NFT prices were high. When volume fell, so did the reason to run the program. Starbucks' secondary income was small next to what Polygon paid to host it.
Programs depended on one executive. RTFKT and .SWOOSH faded after the CEO who backed them left. A program with no measured link to repeat purchase is easy to cut in the next budget review.
Nobody planned the ending. Every closure left holders to export wallets or move assets on a deadline. That damages trust in the brand, and in Nike's case it ended in a lawsuit.
How to design a web3 loyalty program in 2026
If you are a crypto project or a brand thinking about on-chain rewards, these are the principles we would start from.
Start with the behaviour you want. Write down the one action that makes the business healthier, such as a second purchase, a weekly trade or a monthly login, and reward that. Our article on crypto user retention covers how to pick and measure it.
Hide the chain until it helps. Use an embedded wallet so customers sign in with an email, and only show the on-chain side to people who ask for it. We explain the options in our guide to embedded wallets. For UFC's Web3 collectible programme the onboarding was built for fans' first wallet experience, and the launch was designed around fandom rather than technology.
Make rewards useful without a market. A reward should be worth having even if nobody will buy it from you. Upgrades, early access, better tables and real discounts all pass that test. A collectible whose only value is resale does not.
Use points before tokens. A non-transferable points system lets you test what people respond to before you create anything that trades. Our piece on crypto points programs explains the trade-offs. For Mythical Forest we ran a points-driven engagement programme with rewards tied to sustained activity, which reached 1M+ users onboarded.
Plan the exit on day one. Decide what happens to holders if the program ends. Assets in self-custody wallets, rewards that stay usable after the program stops and a long notice period all protect the brand later. Transferable tokens can also raise securities and consumer protection questions in some countries, so take advice before anything trades. This is not legal advice.
What we'd tell a founder
Treat a web3 loyalty program as a retention product, not a launch campaign. The novelty that sold out a drop in 2022 is gone, and what remains is the old loyalty question of whether customers come back more often because of it. Measure returning users and repeat spend, not mints or wallets created.
Keep the first version small. Pick one reward, one action and one audience, run it for a quarter, and only add tokens or trading once you can show the program changes behaviour. If you want help with that, our product marketing team works on exactly this kind of design, and you can see more of the programs we have supported in our case studies.
If you are planning a web3 loyalty program and want a second opinion on the design, talk to AnkhLabs.
The bottom line
The first wave of brand web3 loyalty proved that big companies could put millions of people into wallets. It did not prove that collectibles make people loyal. The programs still running reward the visit, the flight or the meal, and keep the blockchain in the background. That is a less exciting story, and a much more durable one.
This article is for information only and is not investment or legal advice. Program status is as of 6 November 2026.