The loyalty program sits in marketing. Its economics land on the chief executive’s desk.
Every point the program issues is a promise the company has not yet kept, and the company carries that promise on its books until the member uses it. When a bank buys those points to give its cardholders, the card deal becomes a contract that runs for years and pays in cash. The consent a member gives decides what the company can do with her data. None of that fits on a campaign calendar.
A frequent flyer books a trip with the airline’s own credit card. Marketing records an engaged member. Finance records the miles the airline now owes her, and the partnerships team records the card spend. The data team adds a signal to her profile. Four accurate reports, and not one of them sees all of her.
A loyalty program strategy is a decision about the whole company
The question most loyalty teams bring to planning is how to make the program more engaging, with a richer reward or a new tier. Those are fair questions, and they are the last ones a strategy should answer.
A loyalty program strategy is not a plan for the program. It is the company deciding what the customer relationship has to earn, and who answers for it.
The difference shows up in the questions each one asks. A plan for the program asks how many members joined and how many redeemed. A strategy asks what the program did to revenue and margin, and what the company now owes because of it. Those questions belong to the people who answer for the whole business, which is why the strategy has to start with them.
Loyalty Juggernaut describes the shift in three stages. Loyalty began as an earn-and-burn engine and grew into a customer experience engine. Now it is an enterprise growth engine, and in the company’s words, “that makes loyalty a CEO decision, not a retention chore.” A strategy written from the enterprise asks the program to prove it on the income statement.
Lenders fund loyalty programs as businesses
In 2020, United raised $6.8 billion secured by substantially all of the assets of its MileagePlus program. The lenders were not financing a campaign. They were financing a business with its own members and its own card income. United valued that business at about $21.9 billion, twelve times its 2019 EBITDA, a measure of its operating earnings. By July 2025, United had repaid the debt, and the program’s assets were free of the lenders’ claim again.
A program that can carry $6.8 billion of debt is an asset in its own right. Its strategy belongs with the people who answer for the balance sheet, and the airline loyalty economics guide shows where that value comes from.
United’s case is the most dramatic, and the same pattern runs through what three more companies have told investors. Marriott members filled 75% of its room nights in the United States and Canada in 2025, and Marriott books the program’s revenues and expenses at the corporate level, outside its reported segments. American Airlines received $6.2 billion in cash from its co-branded credit card and other partners in 2025. Starbucks leaders introduced a redesigned program at the January 2026 Investor Day, and the company called its March launch “a key milestone in the Back to Starbucks strategy.” In all four companies, the economics pulled the program up to the people who run the business.
The same program carries an obligation, and that is where the view from marketing gets too narrow. A point keeps costing money after the campaign ends, because it sits on the balance sheet as deferred revenue, money the company has received for something it still has to deliver, until she redeems it. United expects some miles to go unredeemed, and its 2025 annual report says that value becomes revenue in step with the miles members do redeem. The points liability accounting guide explains how that promise is measured, and finance should put a price on it before marketing issues the first point.
A loyalty program strategy starts with outcomes and ends with mechanics
A loyalty program is the offer members see: how they earn and what they redeem, with the tiers and partners that come with membership. A loyalty program strategy decides what that offer has to do for the business, so it has to be written before the offer is.
Start with the outcomes. Name the revenue the program should move and the margin it can spend to move it. Decide what data members will agree to share and which partners should pay to reach them. Then set the obligation finance can carry, and agree on how anyone will know what the program caused.
Next comes the value exchange, which is what she gets in return for her spend and her data. Hard value is money she can count, such as a discount or a free night. Soft value is recognition, such as early access or an agent who greets her by status. The strategy sets the mix between the two, and finance puts a cost on both.
Mechanics come last. Points build frequency, and tiers concentrate her spend with the brand. A paid membership asks her to commit up front, and it works when the benefits clearly beat the fee.
Build it in the opposite order and the program still launches, only with mechanics that answer to no outcome. A tier is not a strategy. It is a tool, and a tool needs a job. Before any mechanic is approved, the loyalty team should be able to name the outcome it serves.
Loyalty touches six lines of the business, and at least four departments hold them
The program never touches only one line. United’s card agreement gives Chase “permission to market to the Company’s customer database,” according to the same annual report. One term of one contract is both a data decision and a revenue decision.
Count the lines and there are six: revenue, margin, customer data, partner income, the liability on the balance sheet, and attribution, which is the work of proving what the program caused. Put all six on one page and the chief executive has a scorecard, with a metric for each line and an owner who reviews it on a set schedule.
Each owner manages the line in front of her. Marketing wants more members, finance wants a smaller liability, and the partnerships team wants more card spend. Each is right about her own line, and none of them owns the trade between the lines.
That is not a failure of the people. It is a failure of where the program sits. The seat that sees all six lines at once belongs to the chief executive, so the scorecard belongs on that desk.
Six decisions come before the first point is issued
A strategy that starts from the enterprise makes six decisions, in this order.
- Name the outcomes and put a number on each. Choose the lines of the scorecard the program must move, and set targets the CFO would sign. When revenue and margin pull in different directions, the strategy should already say which one wins.
- Write the value exchange before the rules. State what she gets and what the company asks of her in return. If the offer needs a paragraph to explain, she will skip the math and treat it as a coupon.
- Choose mechanics that serve the first two decisions. Each mechanic spends money, so each one needs an outcome from the first decision to answer to. A tier no member is chasing is not a benefit. It is a cost with a logo on it.
- Price every reward before launch. Finance should approve the cost of each reward and the redemption rate the plan assumes. The obligation starts the day the point is issued, and it grows with every point after it.
- Treat partner deals as long contracts. Co-brand remuneration arrives in cash under a contract that runs for years, so the deal deserves the scrutiny of any major contract. American’s own annual report warns that proposed limits on card transaction fees and card interest rates could fundamentally alter the profitability of its card agreement. A risk of that size belongs in the boardroom.
- Give the chief executive one scorecard and a test for cause. Review the six lines together every quarter. Keep a comparable group of members out of each major offer, a group operators call a holdout, and count only the difference between the two groups as growth the program caused. Growth that would have happened anyway does not count.
GRAVTY® gives the chief executive one view of the program
In GRAVTY®, that one booking reaches all four teams as a single member record, with the flight, the card spend, and the partner activity together. Patented multi-dimensional behavior tracking keeps adding to it every time she flies or spends.
The rules that paid her miles run in GRAVTY Visual Rules, the patented no-code rules engine. When finance reprices a reward, the loyalty team changes the rule without writing code.
Agentic AI Compass, launched in October 2025, works like a team of expert analysts. It reads program performance next to member sentiment and benchmarks, and it simulates an offer’s outcome before launch. Each recommendation is explainable and auditable, so the CFO can see how it was reached. AI-Trust watches earn and redemption activity for fraud as it happens, which protects the currency her bank pays for.
WestJet shows what a strategy-level redesign looks like on the platform. Steve McClelland, its vice president of loyalty and strategic partnerships, describes “a multi-year journey to completely re-architect WestJet Rewards, from a flight-centric program to a full points-and-status ecosystem,” with GRAVTY® as “the platform driving this vast transformation.”
The four reports now draw on one record. The chief executive sees one set of numbers for the program, and the holdout groups from the sixth decision show which growth it caused.
She judges the program by the promises it keeps
She never sees the scorecard. She sees whether the upgrade clears and whether the miles are there when she needs a seat.
Each of those moments was decided somewhere above the program. The reward pricing finance approved decides what her miles can buy, and the consent the company asked for decides whether the next offer feels like help or like surveillance.
A program run as a campaign counts her as a response rate. A program run as company strategy treats her balance as a promise and her data as a trust. Those promises reach every line of the business, so their owner has to sit where the lines meet.
Marketing can run the program. Only the chief executive can own what it promises.
The numbers behind loyalty program strategy
These are the disclosures and working tools behind this essay. Each figure comes from the company’s own filing or release.
Four companies that took the program to investors
| Company | What it disclosed | Where it lands |
|---|---|---|
| United Airlines | Valued MileagePlus at about $21.9 billion in June 2020, twelve times its 2019 EBITDA. Had raised $6.8 billion in notes and a term loan by July 2, 2020, secured by the program’s assets. Repaid the last $1.52 billion on July 7, 2025 | Financing and the balance sheet |
| Marriott | Members filled 75% of room nights in the United States and Canada in 2025, and 68% worldwide. The annual report books its card fees and the program’s revenues and expenses at the corporate level, outside its reported segments. In August 2026, chief executive Anthony Capuano reported new long-term US card agreements with JPMorgan Chase and American Express | The demand plan and corporate results |
| American Airlines | Received $6.2 billion in cash from its co-branded credit card and other partners in 2025. Signed a 10-year deal that makes Citi the exclusive US issuer of its cards from 2026. The same report warns that proposals to cut the fees on card transactions or cap card interest rates “could fundamentally alter the profitability” of its card agreement. In July 2026, chief executive Robert Isom named “lead in loyalty” as one of four commercial pillars | Partner income and contract risk |
| Starbucks | Leaders including chief executive Brian Niccol introduced a redesigned program at the January 2026 Investor Day. It launched in March with three levels of membership, and the company called it “a key milestone in the Back to Starbucks strategy” | The turnaround plan |
A scorecard for the six lines
| Outcome | The metric | The owner | Review cadence |
|---|---|---|---|
| Revenue | Member share of sales: member sales ÷ total sales, and its trend | Marketing and sales | Monthly, and with the CEO each quarter |
| Margin | Margin per member after reward cost: (member gross margin minus reward cost) ÷ active members | Finance and merchandising | Monthly |
| Customer data | Share of revenue from known, consenting customers | Technology and privacy teams | Quarterly |
| Partner income | Partner cash per active member: partner cash ÷ active members, with contract terms | Partnerships team | Quarterly, and before each renewal |
| Balance-sheet liability | Growth in points owed against growth in revenue | Finance | At each quarterly close |
| Attribution | Holdout lift: growth in the offer group minus growth in the holdout group | Marketing analytics | After each major offer, and each quarter |
Table stakes for each outcome
A strategy at the top does not excuse a weak foundation. Each building block serves an outcome from the scorecard.
- Revenue: a program type that fits how she buys. Points reward frequent everyday purchases. Tiers concentrate spend. Paid membership fits customers who buy often enough to earn back the fee, and games and challenges give her reasons to return between purchases. The loyalty program design guide sets out the structures in order.
- Margin: simple rewards with a price on each. She should understand the offer at a glance and reach a first reward early. Finance signs off on the cost of each reward before launch.
- Customer data: consent that powers personalization. Recognize her on every channel she uses, from the app to the store. Record what she agreed to share, and use it to choose offers that fit her.
- Partner income: partners from her ordinary week. Pick brands she already buys from. Set what each partner pays per point and when it settles.
- Balance-sheet liability: published terms and fraud controls. Publish the rules for earning and redemption, including when points expire. Set the redemption rate the plan assumes, and watch earn and redemption activity for fraud.
- Attribution: a holdout on every major offer. Keep a random group of eligible members out of each offer, and compare the two groups. The loyalty program ROI guide shows the math.
Frequently asked questions
What is a loyalty program strategy?
A loyalty program strategy is the plan for what a loyalty program must deliver for the whole company. Its outcomes run from revenue and margin to partner income and the liability every point creates. The loyalty program is the offer members see: how they earn, what they redeem, and which tiers and partners come with membership. The strategy sets the outcomes first, then the value exchange, then the mechanics.
Should a loyalty program charge a membership fee?
Yes, when members can see that the benefits clearly beat the fee. Every renewal tests that promise again. In a McKinsey survey published in October 2020, members of paid programs were 60% more likely to spend more on the brand after joining. Free programs raised that likelihood by 30%. Consumers expected at least a 150% return on the fee. Half of the members who canceled did so in their first year, so the first renewal is the one to price the benefits for.
What is the best example of a loyalty program?
For a brand designing its own program, Marriott Bonvoy and McDonald’s are the best examples to study, because both disclose how much business runs through members. Bonvoy members accounted for 75% of Marriott’s room nights in the United States and Canada in 2025. McDonald’s recorded $40 billion in systemwide sales to loyalty members in the twelve months to June 2026, up over 20%. For any one company, the best model is the program whose economics look like its own. The loyalty program examples guide picks one per industry.
What are the top 10 loyalty programs?
For a brand building its own program, the top 10 are the ten whose companies publish enough of the economics to learn from. Each one teaches a single lesson:
- Marriott Bonvoy (hotels): more than 295 million members at the end of June 2026. The lesson: at that scale the program is the demand plan, so report the share of business members bring next to the member count.
- Hilton Honors (hotels): 260 million members, per Hilton’s July 2026 results. The lesson: a member count measures reach, and a strategy still needs a revenue or margin target beside it.
- Delta SkyMiles (airline): $8.2 billion in remuneration from American Express in 2025. The lesson: the card contract is a revenue line in its own right, and it belongs in front of the CFO.
- American Airlines AAdvantage (airline): $6.2 billion in cash from its card and other partners in 2025. The lesson: partner income carries contract and regulatory risk, and the strategy has to model both.
- United MileagePlus (airline): backed $6.8 billion of debt in 2020, all repaid by July 2025. The lesson: lenders value a program as an asset, so its strategy belongs with the balance sheet.
- McDonald’s (quick service): $40 billion in sales to loyalty members in the twelve months to June 2026. The lesson: member sales show reach into revenue, and only a holdout shows how much of it the program caused.
- Starbucks Rewards (coffee): relaunched in March 2026 with three levels, from Green to Reserve. The lesson: a program redesign can anchor a turnaround plan that leaders present to investors.
- Costco membership (warehouse club): 84.1 million paid memberships at the end of its fiscal year in August 2026. The lesson: in a paid program the fee is the commitment, and each renewal tests the value.
- Amazon Prime (retail): Prime fees sit in subscription services revenue, which was $49.6 billion in 2025. The lesson: membership fees are revenue on the income statement, which changes how finance judges the program.
- American Express Membership Rewards (cards): a $16.5 billion rewards liability at the end of 2025. The lesson: every point issued is an obligation, so finance prices the rewards before marketing issues the points.
Who should own a loyalty program strategy?
The chief executive should own the outcomes, because the program touches lines that report to different leaders. A loyalty leader runs the member experience and the calendar. Finance owns reward cost and the liability. The partnerships team owns card and partner contracts. Technology owns identity and consent. One scorecard, reviewed together each quarter, keeps all six lines working toward the same targets.
How do you know if a loyalty program strategy is working?
Check the outcomes the strategy named, and measure them against a holdout. Keep a comparable group of members out of each major offer, and count only the gap between the groups as growth the program caused. Then review member share of sales, margin after reward cost, and partner income per member together each quarter.
