Building loyalty programs that drive repeat stays

# Building loyalty programs that drive repeat stays

It is 20 November and a Hilton Honors member sits at 57 nights, three short of the 60 that renews Diamond for another year. She books three nights she does not really need, at a rate she would have refused in March. The gap between 57 and 60 is the whole design problem in miniature: everything she has already earned is a cost Hilton has committed to, and only those last three nights are behaviour the program actually bought.

A loyalty program runs two ledgers. On the earn side you decide what you hand out, in which currency, at what accounting cost. On the burn side members take it back, at a moment you do not control and often at a property you do not own. The spacing between those two ledgers is what separates a retention instrument from a discount channel with a progress bar.

Why loyalty programs exist (the real reason)

Two forces sit behind every program, and both are assumed here rather than re-argued: the channel-shift arithmetic the direct-booking lesson works through, and the retention value of a guest who consolidates stays under one brand, which the lifetime valuelifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition → lesson models. What this lesson owns is the machinery in between.

Start with who pays, because it shapes every later choice. Hilton franchises or manages nearly all of its hotels, so it does not own the rooms it is giving away. Hotels pay a loyalty fee on member room revenue into a central program fund, and when a member burns points for a free night the hotel is reimbursed out of that fund at a formula rate. The program therefore has two audiences: the member, who wants the reward night to feel free, and the owner, who compares that reimbursement against the cash rate she just declined. In a sold-out market the comparison goes badly and owners push back on award availability. In an asset-light structure you are negotiating with both sides at once.

The switching-cost machine

A tiered program creates switching costs: the value a member forfeits by moving to a competitor. A Diamond member who books an Accor property starts at zero, so she filters the world down to Hilton first.

The edge case matters more than the rule. Switching costs only bite for members who stay often enough to re-qualify. Below roughly ten nights a year, no tier is reachable, the handcuff is loose, and the member is held (if at all) by the currency in her account rather than by status. That is why a serious program needs two instruments: a tier ladder aimed at the top few percent of stayers, and a spendable currency for the long tail who will never see an upgrade.

Anatomy of a tier structure

Hilton Honors ascends Silver, Gold, Diamond, with qualification available through nights, stays or base points, so a high-rate, low-frequency guest can still climb. Accor's ALL runs Classic through Diamond and keeps a top tier by invitation. Singapore Airlines splits the ladder differently: KrisFlyer Elite Silver and Gold sit on miles flown, while PPS Club qualification runs on spend in premium cabins, which points the best benefits at the passengers who actually pay for them rather than at whoever flies the longest cheap sectors.

The threshold effect

Tiers are spaced to keep members always "almost there", exploiting the goal-gradient effect: people accelerate effort as a goal comes into view. Hilton makes it visible in the app with nights earned and nights remaining. The bar is the marketing.

The qualification window does more work than the perk list. A calendar-year cutoff manufactures a December deadline and a measurable booking surge from members within a few nights of a tier. A rolling twelve-month window is fairer and generates almost no urgency, because there is never a date after which the effort is wasted.

Recognition vs. hard benefits

• Hard benefits cost real money: food and beverage credit, suite upgrades, bonus points.
• Recognition benefits cost close to nothing: priority check-in, a dedicated line, a welcome note, late checkout on a room that would sit empty.

Load the lower tiers with recognition and reserve hard benefits for the scarce top. Note also how the hard benefits get capped: Hilton's US brands give Gold and Diamond members a fixed daily food and beverage credit instead of an open-ended buffet, which converts an uncontrolled cost per guest into a known one.

Recognition stops being cheap under pressure. At a resort running 95 percent occupancy, a 4pm checkout blocks a same-day arrival and can cost a full room-night plus a housekeeping shuffle. Write those perks as subject to availability, and train front desks on when to say no, or the cheap tier becomes the expensive one exactly when rates are highest.

Points, liability, and breakage

Points are a liability

Issued points are a liability: a future obligation to deliver a night or a reward, carried as deferred revenue until redeemed or expired. For a large chain the loyalty liability runs into the billions of dollars and sits in both current and non-current form on the balance sheet. Issuing points is borrowing from guests.

Breakage is the profit

Breakage is the share of points never redeemed. It is pure margin, because the obligation disappears without delivery. Too little and the program is expensive; too much and members feel cheated.

Expiry design decides most of it. Hilton Honors points lapse after two years of account inactivity, so any earn resets the clock. Singapore Airlines KrisFlyer miles expire three years after they are earned regardless of activity, with a paid extension option. The two rules produce opposite behaviour: activity-based expiry keeps members engaged and pushes breakage towards zero among co-brand cardholders, whose monthly card earn resets the clock indefinitely; hard-dated expiry forces periodic redemption spikes, complaints and award-inventory crunches. Growing the card portfolio quietly grows the liability, which is a conversation the CFO will have with you eventually.

Under current accounting rules (ASC 606 in the US, the revenue-recognition standard), companies must estimate breakage and recognise that revenue over time. For deeper context on the standard, the FASB revenue recognition overview is a useful starting point.

The co-branded credit card engine

The most profitable part of a hotel program is often not hotel stays. Hilton sells points in bulk to American Express, which awards them to cardholders on everyday spending. That is recurring, high-margin revenue, and it keeps non-travellers attached to the brand between trips.

The currency design underneath is a real fork. Accor's ALL Rewards uses a fixed rate, 2,000 points for a €40 reward, so members always know what a point is worth and devaluation risk is near zero. The cost is that a fixed-value point behaves like a cash discount and inspires no aspiration; nobody hoards for the suite. Hilton dropped its published award chart in 2022 and lets redemption cost float with the cash rate, which protects margin and matches supply, at the price of continuous quiet erosion that award-tracking blogs publicise for you.

🎬 [VIDEO: "How Hotel Loyalty Programs Actually Make Money" - youtube.com - a clear breakdown of points economics, breakage, and co-brand card revenue in the hospitality sector]

Converting one-time travelers into repeat guests

Step 1: Cheap entry, instant status

New members get a starter tier or an immediate small perk, so the first booking already pays something back. Low friction, instant gratification.

Step 2: Visible progress

Nights earned, next threshold, distance remaining. Progress has to look achievable from where the member actually stands.

Step 3: The status trap

Once a member holds a tier with upgrades and food credit, the perceived cost of switching jumps. She books the brand by default and defends the status each year.

Step 4: Channel lock-in

Full points earning and elite-night credit are reserved for bookings made through the brand's own channels, for the reasons the direct-booking lesson sets out. The failure mode is collateral: stays booked through a corporate travel agency or a wholesaler can fall outside the rules, and the member who loses credit blames the hotel, not her employer's mandated tool. Since mandated bookers are often the highest-frequency cohort, most large programs carve out corporate channels for elite credit even where earning is reduced.

SegmentingSegmentingDividing a market into distinct groups of customers who share similar needs, characteristics or behaviours, so each group can be served with a tailored approach.View full definition → the member base

• High-frequency business travellers: worth heavy recognition and upgrade investment; they drive base volume and rarely pay out of pocket.
• Occasional leisure guests: keep cheap and engaged with points, card offers and partner earning.
• Dormant members: reactivate before expiry, since a lapsed account is a lost owned channel.
• Card-only members: they earn without staying. Their points are a liability with no room revenue attached, so the offer has to convert them into a first stay.

Where loyalty programs go wrong

Devaluation

Raising the points needed for a free night reduces the liability and, if members notice, destroys the belief that makes the currency work. Dynamic redemption pricing spreads the same effect across every date, which is harder to announce and just as easy to detect.

Award availability

Points that cannot be spent on the dates members want are worth less than the arithmetic suggests. Tight award inventory and blackout dates create resentment, and in a franchised system they are often the owner's decision, not the brand's.

Status inflation

Top-tier status attached to a credit card (Hilton's premium Amex grants Diamond outright) fills hotels with members entitled to upgrades that physically do not exist. Upgrade capacity is fixed, so bought status dilutes earned status, and the 60-night guest you built the program for is the one who stops getting the suite.

Overpaying for loyalty

If members chase status on stays they would have booked anyway, the perks are cost without lift. Comparing member revenue to non-member revenue proves nothing, because members self-select. Test with a holdout: withhold the bonus-points offer from a random slice of the eligible segment and measure the difference in stays, not in opens.

A quick framework for evaluating any program

1. Does it produce incremental stays, proven against a holdout rather than a member-versus-non-member cut?

2. Is the points liability matched by realistic breakage plus partner and card revenue?

3. Do the tiers put cheap recognition at the bottom and scarce, capped hard benefits at the top?

4. Is perceived value stable enough that members still believe in the currency after the next repricing?

5. Do the hotels that deliver the rewards find the reimbursement acceptable at peak occupancy?

Key takeaways

• A program buys behaviour at the margin. Only the stays a member would not otherwise have made are worth the perks; everything else is a discount you gave to loyal customers for free.
• Points are a liability; breakage is the profit. Expiry rules decide how much breakage you get, and activity-based expiry combined with a large co-brand card portfolio drives it close to zero.
• Currency design is a real fork. Accor's fixed 2,000 points for €40 kills devaluation risk and aspiration together; floating award pricing protects margin and invites detection.
• Recognition is cheap until the hotel is full. Late checkout and upgrades cost nothing on a soft night and cost a room-night on a sold-out one, so write them as conditional.
• Bought status breaks earned status. Handing top tier to cardholders dilutes the benefit for the high-frequency guests the ladder was built to hold.