When inflation comes to loyalty points

The case of the inflationary loyalty points

ClubCard wasn’t Tesco’s first loyalty points system. In 1963, the grocer signed on to a loyalty scheme called Green Shield Stamps. The introduction of these loyalty points was so momentous that thousands descended upon the shops when they were launched, and a dozen people fainted. It was a big deal.

A newspaper excerpt at the time:

In Leicester yesterday, the giant Tesco store was besieged by thousands of housewives. Twelve women fainted. The staff was completely overwhelmed. Finally store manager John Eastoe cleared the shop and closed all the doors. Mr Eastoe said, ‘I have never seen anything like it in my life…’

— Humby, Clive; Hunt, Terry; Phillips, Tim. Scoring Points (p. 44). Kogan Page. Kindle Edition.

Here’s how the system worked: Shoppers received a number of stamps proportional to the amount they spent at the supermarket. They collected the stamps in a booklet, which when filled, could be exchanged for products from a catalogue.

The Green Shield stamps were one of a number of stamps systems at the time. These were operated by independent companies who signed on merchants like Tesco to their schemes. Merchants brought in the spend; the stamps companies provided the catalogues for redemptions.

But as a new wave of discount retailers entered the market, Tesco began to shore up its offering with more stamps. The stamps began to subsidise the prices of goods on the shelves. Tesco issued more and more stamps to retain their customers as a result. The supply of stamps grew, driving up the redemption thresholds. Inflation was setting in.

The Wikipedia entry for Green Shields stamps notes that at one point, petrol stations were aiming at company drivers, whose employers footed the fuel bill. Therefore, their corporate spend was enticed with ever increasing numbers of stamps.

Tesco’s chairman at the time, Lord MacLaurin, described the escalating inflation thus:

The ludicrous game of having to collect a barrow load of [stamp] books to obtain a TV set.

— Humby, Clive; Hunt, Terry; Phillips, Tim. Scoring Points (p. 45). Kogan Page. Kindle Edition.

The knock-on effects were interesting. The scheme was costing Tesco £20 million a year to run. This led to arguments at the Tesco board level over keeping or killing the scheme. The fear was that removing the loyalty programme would lead to a drop in sales far worse than the cost of running it.

Meanwhile the Green Shield stamps company was starting to allow redemptions partly in cash to offset the inflationary effects of a rising stamp supply. The inflation can be seen—just as with paper bank money—in the denominations of stamps issued.

Wikipedia tells us that the stamps were originally issued for every 2.5 pence of expenditure at a merchant, leading to large quantities of stamps required for redemption. But as more stamps were required for redemption, because more stamps were being issued, stamps began to be issued in a new denomination, worth 10 2.5 pence stamps. This went on until a stamp worth 40 ordinary ones was put in circulation.

Eventually the whole catalogue redemption business was restructured and rebranded. The stamps were abandoned. Customers could simply pay cash for goods from a catalogue. The overhauled business still exists today as the large high street chain Argos, which books billions in revenue annually, with its unique catalogue, delivery, and storefront model.

The velocity of money, liquidity wars and the Banana Man of Worcester

From the inflationary dynamics above, it’s clear that loyalty point design can suffer from some of the same problems that currencies, or cryptocurrencies, endure. Here are some other features and parallels with currencies and crypto coins, and by extension, social tokens.

One of the things Tesco did with its Clubcard system, which replaced Green Shields, was to ensure the points expired. Expiring points ensure that there isn’t a huge pile of liabilities out there in the Clubcard economy waiting to be settled in. If points didn’t expire, you could imagine millions of hoarded points being taken out of the system and never spent. These hoarded points would end up as liabilities on Tesco’s balance sheet since they would have to be honoured at some point in the future. Here’s the passage explaining the logic behind it:

Tesco did not want to end up in a mess like the world’s airlines, with millions of unredeemed frequent flyer miles accumulating as a growing liability. So every Clubcard voucher carried an expiry date for a year ahead, long enough away for shoppers to feel they were not being coerced into spending them before they wanted (many customers saved up their vouchers to put towards their Christmas shopping, for example) but soon enough that Tesco wasn’t building up financial exposure for the future.

Humby, Clive; Hunt, Terry; Phillips, Tim. Scoring Points (p. 72). Kogan Page. Kindle Edition.

The notion of money with a sell-by date is a little similar to demurrage, or the idea of rotting money, which we covered in our local currencies explorations in earlier Pacenotes. The point of a local currency is to be spent; hoarding it can potentially distort future prices and reduce utility to the community it’s supposed to serve. It’s one reason why bitcoin’s commodity-backed, deflationary design disincentives holders from spending it.

In designing a social token or other type of local money, then, it’s important to bake in some features that encourage monetary velocity. If the money or tokens or points are hoarded, that local economy is left holding the figurative bag. This is why some token communities have implemented the idea of “seasons” within each Discord server. The balance of tokens held rises each season, ensuring that there is a regular movement of tokens.

Another interesting episode from Clubcard history presents itself in the 1990s. At this stage, Tesco is embroiled in a price war with Safeway, which has launched a hugely successful loyalty programme of its own, signing up 9 million users to its ABC Card.

But by the late 90s, Safeway is losing money, a victim of the ongoing price wars. Its loyalty points system is costing it £50 million a year. Its top management is losing faith in the ability of points to convert new customers, and to wring out data for the company to use on existing ones.

Tesco makes an aggressive move. It offers 250 Clubcard points for any Safeway user who trades in their ABC card for a Clubcard. The result? Tens of thousands made the switch on the first day. By the end of the campaign’s first week, 100,000 Safeway loyalty users had switched to Tesco. Ultimately, Safeway decided to end its ABC card programme and focus on slashing prices in stores.

The ABC-Clubcard swap campaign reminds me of yield farming or liquidity rewards. The supermarket shoppers are liquidity providers. Their spend is what the various loyalty schemes want to attract. And they are highly price sensitive, willing to shift their expenditure to whoever offers the best deal.

Loyalty points bonuses are therefore like liquidity mining rewards. By moving your spend to a new liquidity pool—Clubcard—you get rewarded. In the process, other liquidity pools can suffer, as in the zero-sum game played by Tesco and Safeway. These sorts of liquidity wars play out in DeFi all the time: witness the funds sloshing between Uniswap and Sushiswap at the height of “DeFi summer” last year.

Finally you have yield farmers, or degens. These are users or traders who have figured out an arbitrage and are going to exploit it to the hilt. In the case of Clubcard, this was the banana trade.

The greatest exponent of this trade was a 28-year-old physicist named Phil Calcott who lived in Worcester. Tesco awarded 25 Clubcard points to each three-pound bunch. Calcott figured out that the points were worth £1.25 in total, while the bananas only cost £1.17—an eight pence difference. Effectively, Tesco was paying customers to buy bananas.

Unfortunately for Tesco, the yellow fruit was Calcott’s favourite. He began buying huge quantities of bananas, filling up his Peugeot hatchback with hundreds of pounds worth of the stuff on each trip. Eventually he amassed nearly one ton of bananas in his home. He also collected 7,850 Clubcard points by the end of his campaign.

By the time Tesco mailed out Clubcard vouchers, which they did on a quarterly basis (and also so that they could get customers’ addresses), Calcott received a payout of nearly £400 for his efforts. Here’s how the press chronicled the Banana Man of Worcester’s heroics:

‘We thought he had gone bananas,’ said Tesco’s local customer services manager Helen Williams. So did the press. The Sun printed a picture of him, buried in bananas. The Independent ran a concerned editorial on the amount of petrol and work it took to buy the bananas, warning that ‘there is no such thing as a free lunch’.

Humby, Clive; Hunt, Terry; Phillips, Tim. Scoring Points (p. 104). Kogan Page. Kindle Edition.

One of the underlying points the Humby book makes is that loyalty programmes are not an end in themselves, and they are not simply a subsidy to the customer. Instead, they are a way for the merchant to thank the customer and build a relationship, not a bribe. This is a salient point for creators issuing rewards to fans.

Humby also argues that the effects of a loyalty programme are not always evident in simple top or bottom-line figures. Instead, a holistic framework for measuring the impact of such points on the overall business—social tokens issuers might think of this as the community or the token economy—needs to be used in order to judge the effectiveness of a points system’s design.

Humby devised the “loyalty cube” framework, which is what I will cover in the next Pacenotes.