Starbucks Customer Loyalty Statistics: Rewards Members and Tier Data

Starbucks reported 35.5 million 90-day active Rewards members in the US in January 2026, alongside a rebuilt tier structure. The mechanics of that rebuild say more about the economics of the programme than the headline member count does.

The context is a company with $37.2 billion in fiscal 2025 net revenues whose North America comparable store sales grew 0% in the fourth quarter. A large, active loyalty base and flat domestic sales are not a contradiction — but they are the reason a redesign happened.

The tier structure

Three tiers, with the qualifying threshold measured over a rolling twelve months:

Tier	Qualifying threshold	Stars per $1
Green	Entry level, no minimum	1.0
Gold	500 Stars in 12 months	1.2
Reserve	2,500 Stars in 12 months	1.7

The earn rate rises 70% from entry to top. That is a deliberate slope: the programme pays materially more per dollar to the customers who were already spending most.

What a Star is actually worth

Starbucks publishes the earn rates and the redemption thresholds separately. Put together, they give an effective reward rate — the share of spend returned as value. This calculation is ours, not a figure Starbucks states.

Using the newly introduced 60-Star redemption for $2 off:

Tier	Spend needed for 60 Stars	Value returned	Effective rate
Green	$60.00	$2	3.3%
Gold	$50.00	$2	4.0%
Reserve	$35.29	$2	5.7%

A Reserve member gets back about 73% more per dollar than a Green member. The programme is not a flat discount dressed as loyalty; it is a graduated bid for frequency, and the bid rises steeply with how much you already spend.

The expiry rule is the actual mechanism

The tier benefit that gets least attention is the one doing most of the work.

• Green: Stars valid six months, extended by one month with qualifying activity.
• Gold and Reserve: Stars never expire.

Read that as two different instruments aimed at two different problems. For light users, expiry manufactures a deadline — the Stars you already earned become a reason to return before they lapse. For heavy users, removing expiry removes a reason to feel penalised for a quiet month, which protects the relationship rather than prodding it.

The one-month extension at Green is the more interesting detail. It converts every visit into a small renewal of the whole balance, which makes the marginal visit worth more than its own transaction value. That is a well-designed mechanism, and it is invisible in any member-count statistic.

The redemption ladder

Earn rates are only half the mechanism. The redemption thresholds determine what the Stars convert into, and Starbucks rebuilt those too.

The published ladder runs 25, 60, 100, 200, 300 and 400 Stars, with the 60-Star tier newly introduced. That addition is the one worth attention: it sits between the smallest add-on rewards and the substantial ones, and it is the first rung at which a member gets a straightforward cash-equivalent discount rather than a modification or an extra.

The design consequence is a shorter distance to a first meaningful redemption. A Green member reaches 60 Stars after $60 of spend; at Reserve it takes $35.29. In both cases the member experiences a completed reward cycle sooner than the old structure allowed, and a completed cycle is what converts a card in a wallet into a habit.

Against that, the larger redemptions still exist, which keeps a reason to accumulate. A programme with only small rewards trains members to cash out constantly; one with only large rewards leaves them stranded. The spread between 25 and 400 covers both behaviours.

The perks are cheap and deliberate

Alongside the Stars, each tier carries non-monetary benefits:

Tier	Included benefit
Green	Free Mod Mondays — a monthly beverage modification at no charge
Gold	At least four Double Star Days a year
Reserve	At least six Double Star Days a year, plus exclusive experiences

Each of these costs Starbucks materially less than its perceived value, which is the whole design intent. A syrup modification has a marginal cost of cents against a perceived value of a dollar or more. A Double Star Day costs nothing at all unless the member shows up — and if they do, the company has bought an incremental visit at the price of doubling the Star accrual on that visit alone.

Note the direction of the ladder here too. Green gets a recurring monthly nudge; Gold and Reserve get periodic multiplier events. The light user is being given a reason to come this month; the heavy user is being given a reason to concentrate spend that was going to happen anyway into a specific day.

Why the redesign, and why now

Starbucks does not state a reason. The disclosed numbers around it are suggestive.

In the fourth quarter of fiscal 2025 the company reported $9.6 billion in net revenues, with comparable store sales up 1% globally, up 3% internationally, and flat at 0% in North America. Full-year revenues came to $37.2 billion.

A loyalty programme with 35.5 million active US members sitting on top of zero comparable growth in its home market is the signal. The membership base is not the constraint; what those members do with it is. A redesign that raises the top-tier earn rate, introduces a smaller and more attainable 60-Star redemption, and removes expiry pressure from the best customers reads as an attempt to convert existing membership into visits rather than to recruit more members.

Whether it works is a fiscal 2026 question, and the answer will show up in comparable sales rather than in the member count.

What Starbucks does not publish

Two absences matter for anyone quoting these figures.

Share of sales from Rewards members. This is the number that would show whether the programme drives the business or merely tracks it. The releases we reviewed do not contain it. Figures have circulated at various levels over the years; we do not republish them, for the reasons set out in our methodology.

A consistent quarterly member count. The 35.5 million figure appeared in a loyalty announcement in January 2026. The Q4 fiscal 2025 results release, three months earlier, does not carry the metric at all. Selective disclosure is normal, but it means any time series assembled from these numbers is being stitched together from different document types with different purposes.

The liability nobody quotes

Unredeemed Stars are a balance sheet item. Every Star earned and not yet spent is a deferred obligation, and the accounting treatment shapes programme design more than most coverage acknowledges.

This is where the split expiry rule does double duty. Stars that lapse at Green after six months extinguish the associated liability — the company keeps the revenue and discharges the obligation. Stars that never expire at Gold and Reserve do the opposite: they accumulate indefinitely against the balance sheet.

That Starbucks accepted the second condition for its highest-value members, while retaining expiry for the lightest, indicates the relationship is judged worth more than the accounting relief. It is a reasonable trade, and it is only visible if you read the expiry terms as a financial instrument rather than as a customer-service detail.

We cannot quantify it: Starbucks does not break out the deferred revenue attributable to unredeemed Stars in the material reviewed here. Anyone quoting a figure for it is estimating.

Where this sits against the alternatives

Loyalty economics and discovery economics buy the same thing — a transaction — through opposite mechanisms, and comparing them is instructive.

A graduated rewards programme pays a known percentage to a known customer for a repeat visit. Discovery commerce pays an unknown acquisition cost to reach a stranger once. The channel-mix and seller-concentration data in our TikTok shopping statistics report shows what the second model looks like at scale: enormous aggregate growth, and a median participant earning nothing.

Between them sits the information problem. Our product photography statistics report covers the cost of failing to tell a first-time buyer what they are getting — a cost that a retention programme largely sidesteps, because returning customers already know.

And the contrast with disclosure is sharp. Our Amazon Alexa statistics report covers a channel where the installed base is published and the commercial behaviour never is. Starbucks at least publishes an activity-qualified member count, even if only when it chooses to.

What to watch next

The redesign gives a clean test, because it changes the mechanism without changing the membership base. Three things will show whether it worked, and only one of them is the number most coverage will quote.

North America comparable store sales. This is the metric the redesign appears aimed at. Membership was already 35.5 million while comps sat at zero; if the new structure converts existing members into visits, it shows up here first. If comps stay flat while the member count rises, the programme is growing its roster without changing behaviour.

Transaction count against average ticket. Comparable sales can rise from more visits or from higher spend per visit, and a loyalty redesign built on frequency should move the first. A comps recovery driven entirely by ticket would mean price, not loyalty, did the work.

Whether the member count reappears in the results release. Starbucks disclosed 35.5 million in a dedicated loyalty announcement, not in the quarterly numbers. If the metric starts appearing in results releases, it is because it supports the story. If it stays in standalone announcements, that is a signal about which direction it is moving.

Note the asymmetry in what we will be able to check. The sales metrics are mandatory disclosures and will be there every quarter. The member count is voluntary, and voluntary metrics have a way of becoming scarce exactly when they would be most interesting.

Reading loyalty statistics generally

Three habits make loyalty figures usable, and they apply well beyond this company.

Prefer activity-qualified counts. “90-day active” is a far more honest metric than cumulative sign-ups, which never go down. When a programme quotes total members without a recency qualifier, assume the active number is much smaller.

Derive the effective rate yourself. Earn rates and redemption thresholds are usually published in different places precisely because the combined figure is the commercially sensitive one. Combining them is arithmetic, not analysis, and it is the fastest way to compare two programmes.

Watch the expiry terms. They reveal who the programme is actually designed around. A scheme that expires points aggressively for everyone is managing a liability; one that expires them only for light users, as here, is running a behavioural mechanism.

Related reading

• TikTok shopping statistics — acquisition economics against retention economics
• Product photography statistics — the cost of a first-time buyer not knowing what they are getting
• Amazon Alexa statistics — a channel that publishes reach and hides behaviour
• All research — everything currently on record
• How we source figures — what we will and will not publish

Frequently asked

20 entries

How many Starbucks Rewards members are there?

35.5 million 90-day active members in the US, as stated by Starbucks in January 2026.

What does "90-day active" mean?

A member who has used the programme at least once in the previous 90 days. It is a stricter measure than total sign-ups, which Starbucks does not emphasise.

Does Starbucks publish the member count every quarter?

Not reliably. The Q4 fiscal 2025 results release does not contain it; the 35.5 million figure came from a separate loyalty announcement three months later.

What are the Starbucks Rewards tiers?

Green at entry level, Gold at 500 Stars earned within 12 months, and Reserve at 2,500 Stars within 12 months.

How many Stars do you earn per dollar?

1 Star per $1 at Green, 1.2 at Gold, and 1.7 at Reserve.

What is a Star actually worth?

Using the 60-Star redemption for $2 off, roughly 3.3% of spend at Green, 4.0% at Gold and 5.7% at Reserve. Those rates are our calculation from the published figures, not numbers Starbucks states.

Do Stars expire?

At Green they are valid for six months, extended by a month with qualifying activity. At Gold and Reserve they never expire.

Why does the expiry rule differ by tier?

Expiry is a pressure mechanism on light users and a removed friction for heavy ones. It pushes Green members toward the next visit and rewards Gold and Reserve members with the freedom to save.

What is the smallest redemption?

25 Stars. A newly introduced 60-Star tier gives $2 off, sitting between the smallest add-on rewards and the larger 100 to 400 Star redemptions.

Why did Starbucks redesign the programme?

The company does not say directly, but the timing is informative: North America comparable store sales were flat at 0% in the fourth quarter of fiscal 2025.

How large is Starbucks overall?

$37.2 billion in net revenues for fiscal 2025, with $9.6 billion in the fourth quarter.

What share of sales comes from Rewards members?

Starbucks does not disclose it in the releases we reviewed. It has been reported at various levels over the years, but we do not republish figures we cannot source.

Is a 5.7% reward rate high?

It is high for grocery or general retail and unremarkable for high-frequency food service, where margins support it and visit frequency is the metric being bought.

What do the non-monetary perks do?

Free Mod Mondays at Green, at least four Double Star Days a year at Gold and at least six plus exclusive experiences at Reserve. They add perceived value at low marginal cost, which is the point.

Does a bigger member base mean more sales?

Not automatically. Membership grew while North America comparable sales sat at zero, which is precisely the gap the redesign appears aimed at.

Are unredeemed Stars a liability for Starbucks?

Yes, in principle — every Star earned and unspent is a deferred obligation. Starbucks does not break out the amount attributable to unredeemed Stars in the material reviewed here, so any figure quoted for it is an estimate.

Why let Gold and Reserve Stars never expire if that increases the liability?

Because the relationship is judged worth more than the accounting relief. Expiry is retained precisely where it does behavioural work — on light users — and dropped where it would irritate the best customers.

What is the fastest route to a first reward?

The new 60-Star tier for $2 off. It takes $60 of spend at Green and $35.29 at Reserve, which shortens the distance to a completed reward cycle.

What should be watched to judge whether the redesign worked?

North America comparable store sales first, then transaction count against average ticket. A recovery driven by ticket rather than visits would mean price did the work, not loyalty.

What would make these statistics better?

A consistent quarterly member figure in the results release, and the share of transactions attributable to members. Both are within the company's power to publish.