SIAM News Blog

Rewarding Your Shopping Loyalty with Points

By Michael J. Armstrong and Aaron L. Nsakanda

How many “points” have you collected so far this holiday shopping season? Like millions of consumers, you likely belong to at least one loyalty rewards program. These programs give you “points,” “miles,” or some virtual currency when you buy from specified airlines or retailers. You can later redeem your points for rewards that range from cheaper groceries to exotic vacations.

Of course, such rewards do not represent corporate generosity. Like inspirational brand reputations and slick e-commerce websites, rewards programs encourage you to patronize certain businesses instead of their competitors, although some question their effectiveness. The programs also gather data about your spending habits, which lets businesses promote their products to the most promising prospects.

Some loyalty programs are huge. AccorHotelswhich operates Novotel, Fairmont and other brandshas 27 million members. Marriott International has 110 million. (China could make even those figures look small; its experimental social credit programs might some day cover its entire population. But those reward a different kind of loyalty.)

The most complex programs involve alliances of otherwise unrelated companies. Air Miles Canada members can shop at thousands of participating stores. Aeroplan has partners representing more than 150 brands. In Europe, Miles & More includes 300 businesses. A “host” company runs each of these programs and tracks your points while managing interactions among the partnering businesses.

Hosts Handle the Details

For example, suppose you stay overnight at a participating hotel. The host company will subsequently credit points to your account but also collects revenue from the hotel. In effect, the host sells points to the hotel, which then gives them to you as an incentive for staying there. Later, you might redeem some points to book a flight. The host reduces your point balance while paying the airline on your behalf for the ticket. In between accumulation and redemption, the host company tracks your points and holds the cash. To you, the points are an asset. To the host, however, they are a liability it must some day redeem.

In Canada, the outstanding points balance reportedly approaches CDN$16 billion across all loyalty programs combined. In the U.S., Marriott alone has US$5 billion outstanding.

Some points go unredeemed because, for example, they expire or the owning members quit. This “spillage” benefits the host company, which keeps the previously collected cash. Hosts can consequently boost profits by adjusting programs to increase spillage, but such measures may anger consumers.

Advance Arrangements Needed

Some points go unredeemed because they expire or the owning members quit. Public domain image.
Before any transactions can occur, the host company and its partners must negotiate the number of points earned per purchase or redeemed per reward. They often also commit to specific numbers of rewards redeemed per year per partner.

For instance, Aeroplan promised to buy $581 million of Air Canada tickets in 2017 for anticipated rewards. This presents a problem. Host companies must make these arrangements in advance, but only later discover how many rewards consumers want from each business.

Of course, hosts can simply pre-arrange for more rewards than they expect to need to ensure a safety margin. But this is expensive, as some rewards will go unused. We worked with doctoral student Yuheng Cao and associate professor Moustapha Diaby (University of Connecticut) to study this rewards-supply planning problem. We created mathematical models to simulate point accumulation and redemption in a loyalty program over time, then used these models to evaluate potential host company decisions.

Uncertainty Hurts Performance

One study examined the effects of uncertainty about future demand for rewards. Unsurprisingly, higher uncertainty tends to reduce profits for the host company; this makes it more difficult to accurately plan reward needs in advance.

Less obviously, performance also suffers if hosts try to limit their annual rewards budgets or outstanding points balances. To cap point balances, hosts must purchase more rewards at higher prices, which reduces immediate profitability. Conversely, host companies with restricted budgets must pay out fewer rewards. This increases their future liabilities.

Both problems worsen as demand uncertainty escalates. Hosts progressively lack enough flexibility to handle demand surprises; this implies that they should try to reduce demand uncertainty where possible. They might do this in advance through improved forecasting, or offer discounts on underused rewards as the year unfolds. For example, suppose an airline normally charges 25,000 points per ticket. It might temporarily cut that to 15,000 for flights on unpopular days.

Options Could Compensate

Alternatively, hosts could include “options” clauses in their agreements with their loyalty program partners. With this strategy, host companies would still agree to buy a set number of rewards from each partner. But the contract would also give hosts the option to buy more rewards later if demand ends up being higher, and hosts would pay extra for that privilege. This approach would mean that loyalty program partners absorb more of the program’s risks. In exchange, they would get more of the program’s revenue.

We explored this option contract concept in another modelling study. Our results suggest options contracts should indeed perform better overall than traditional fixed-size contracts.

Options are particularly valuable in large loyalty programs facing high demand uncertainty, as they let host companies spread their greater risks across many redemption partners. By comparison, the common practice of prearranging extra rewards is less effective in these cases and leaves many rewards sitting unused at many partners.

Overall, our research suggests that reducing uncertainty about redemption demandand using reward supply option contractswould both benefit hosts. They could increase their profitability while controlling their rewards budgets and accumulated points liabilities.

This article is republished with permission from The Conversation.

Michael J. Armstrong is an associate professor of operations research in the Goodman School of Business at Brock University. He uses mathematical models to analyze historic battles, such as Gettysburg, Balaclava, and Coral Sea; and to understand modern missile combat, as in naval warfare and Israeli rocket defenses. Aaron L. Nsakanda is an associate professor of management science and supply chain management in the Sprott School of Business at Carleton University.

blog comments powered by Disqus