Skip to content
Gamify It.
Connect

Case · M4-M5 × Pay / Stay (food service)

Starbucks Rewards: stars, tiers and devaluation risk

Starbucks Rewards pays Stars per dollar, with personalized challenges and mobile ordering. At the end of fiscal 2025, 34.2M US members were active over 90 days and their spend was 58% of tender in US company-operated stores. A 2026 tier relaunch drew backlash from casual members.

M4-M5 ×StayPayService business

The outcome · 2025

34.2M US 90-day active members, and member spend at 58% of US company-operated tender (data as of Sept 28, 2025; 60% a year earlier). The three-tier relaunch on March 10, 2026 drew mostly negative reaction from casual members (press).

Case grade for the number: B. Evidence grade for the mechanism link: D.

Analyst inferenceCorrelationalMany causesCompany-reported

Read the source

Stage cell: M4-M5 × Pay / Stay (food service).

What they did. Stars per dollar, personalized challenges and double-star days, mobile order and pay, and from March 10, 2026 a three-tier system (Green, Gold at 500 Stars, Reserve at 2,500 Stars in 12 months) with different earn rates.

The mechanism. [Analyst inference] Exchangeable points GT#75 (Actionable Gamification Ch.8, p.186) and status points GT#1 (Actionable Gamification Ch.6, p.116) turn each visit into progress, and a visible balance is stored value HK-INV27 (Hooked Ch.5). Changing earn rates for existing members revokes something they already counted on INF-T-SC-19 (Influence Ch.7, p.184), which triggers loss aversion and a sense of a broken deal.

The outcome. 58% of US tender (2025) Case grade B. Company-reported numbers. from 34.2M 90-day active members, down from 60% a year earlier. The 2026 relaunch lowered the Gold earn rate versus the old flat rate and drew mostly negative reaction (press, grade C); Starbucks kept existing Stars.

What it means at your stage. For M4-M5 loyalty programs, the risk is devaluation shock: grandfather balances, lead with what members gain, and avoid cutting the earn rate for the median member. Track member share of spend and 90-day actives before and after any change.

Caveats. Company dashboard; many causes behind spend share. Mechanism link Evidence grade D. Practitioner framework or anecdote: plausible, untested. [Analyst].

Would you have called it?

This case is also a round of Predict the Outcome: read the situation, pick what you think happened, then see the evidence.

Play this round

Playbooks that cite it

Try this

The techniques behind it

Each one is a starting point, not a recipe. Gamify It turns a technique into an experiment card for your product: the change, the metric, the guardrail and how long the test needs at your traffic.

  • TechniqueEvidence grade D. Practitioner framework or anecdote: plausible, untested.

    Exchangeable Points (GT#75)

    spendable credits

    A currency users earn and then spend or trade for things they value, unlike progress points that cannot be spent. Chou warns that running an economy is hard: the currency must stay scarce, earn rates need constant rebalancing, and assuming an economy alone creates engagement is a mistake.

  • TechniqueEvidence grade D. Practitioner framework or anecdote: plausible, untested.

    Status Points (GT#1)

    progress score

    A score that rises with meaningful actions, telling the system how close each user is to a Win-State and telling the user where they stand. The point values themselves announce what the product cares about, so weight them by what matters to the user.

  • TechniqueEvidence grade C. Limited evidence: few or small studies.

    Make stored value visible, and exportable

    show what users have built

    Show users what they have built in the product (their library, history, notes, reputation, working setups) in summaries and anniversaries, and let them export it. Visible value raises the felt cost of leaving; export keeps it a gift rather than a hostage.

  • TechniqueEvidence grade B. Moderate evidence: solid studies with some limits.

    Grandfather when you take something back

    notice, choice and grandfathering on removals

    When you raise prices, shrink a free tier or remove a feature, protect existing users: grandfather current terms for a period, give long notice, offer a choice (keep the old version for 90 days, opt in early) and tie the change to a clear gain. Taking back a freedom people already have triggers far more resistance than never granting it.

  • TechniqueEvidence grade D. Practitioner framework or anecdote: plausible, untested.

    Status tiers that grant real power

    tiers with visible criteria and real perks

    Ranks and tiers change behavior only when they bring powers or standing others recognize, and when the criteria for the next tier are visible. Chou's audit of a navigation app found its top rank barely changed behavior because it gave no real power, while earned seller tiers on a marketplace did.

Limits

What we don’t claim

  • The Hook Model is a practical design tool, not a validated or complete theory of behavior.Hooked, Introduction
  • Octalysis is a practitioner framework built from observing games, not a validated scientific model.Actionable Gamification Ch.15, pp.399-400
  • The evidence is correlational until it is tested in your product.
  • No success is attributed to a single technique: real outcomes have many causes.
  • Company names are used editorially; no endorsement is implied.

Sources

Source IDs: CASE-M12GT#75GT#1HK-INV27INF-T-SC-19

Book content is paraphrased, never quoted at length. Last verified October 24, 2026.

Your product

Find the stage your product is stuck in

Your product has its own version of this story. Find the stage it’s in and the move that fits it. Connecting is free; a Pro subscription starts it.

Add Gamify It to your assistantTake the 2-minute check-up