| Takeaway | Detail |
|---|---|
| A fixed cash reward is the value-safe choice at checkout. | The lucky draw’s expected payout trails the fixed reward by 23%, so the guaranteed option wins on expected value. |
| Most shoppers still choose the lucky draw. | 70% of audit participants picked the draw despite the 23% expected-value gap. |
| One draw feels fair even when the odds are not. | The law of large numbers makes the 23% edge visible over many checkouts, but a single draw can look like an even 50% shot. |
| A waiting period changes how the reward is valued. | A 48-hour delay makes the fixed cash reward’s guaranteed value more salient and reduces the emotional pull of the draw. |
Twenty-three percent is the house edge hidden inside payment-app checkout rewards, according to MIT’s checkout-rewards audit. At the median checkout, a fixed cash reward sat next to a lucky draw with a probability-weighted payout 23% lower. The lucky draw is not a consumer perk; it is a price-discrimination engine that converts optimism into margin.
That math did not stop shoppers. 70% of participants chose the draw, even though the fixed option had the higher expected value. Consumers acted as if a 50% chance of winning felt fair enough to skip a guaranteed payout. The larger pattern is predictable: over thousands of transactions, the law of large numbers makes the 23% gap a reliable transfer from shoppers to the app.
The design works by separating customers according to risk tolerance. A 48-hour delay is often the only way to make the fixed option look rational again; without it, the draw’s upside dominates. At the median checkout, the safe option is the fixed cash reward; the draw is the fee.
The Pricing Trick
Stripe and Adyen do not pick your reward; they hand the decision to your wallet. At checkout, the merchant payment orchestrator passes a reward-decisioning callback to the wallet app, and the wallet chooses between a fixed-cash rebate and a random draw from a prize ladder. The app has already priced both options, and the draw is marked to cost it less.
Probability weighting explains how a product that costs less can feel like more. According to Tversky and Kahneman's 1992 cumulative prospect theory, a 1-in-100 reward is decision-weighted at roughly 13%, not 1%. The consumer is not multiplying 1% by the jackpot; the brain inflates the tail. So the headline prize on the ladder out-feels a higher-EV fixed cash offer before any random draw even resolves.
The pricing engine turns that distortion into margin. It marks the draw's expected payout at 82% of the fixed cash payout: for every $100 in transaction value, the median draw expected value is $0.62 versus $0.76 for fixed cash — a $0.14 gap. That gap is the 23% advantage fixed cash holds at the median checkout, and it sits underneath a button that looks like pure upside.
Fixed cash also clears faster. It lands in the wallet balance — PayPal balance or Apple Cash — and is spendable immediately. Lucky-draw balances are often locked to the app's rewards tab for a campaign window, which effectively discounts the draw by 5-10% in realized value. The nominal gap above is therefore a floor, not a ceiling.
The mystery framing is deliberate. The reward module pre-selects the draw as the default action, so a user who wants the fixed cash must tap away from it. The default is the nudge; the probability weighting is the hook; the lock-up is the back-end margin.
At the median 2026 mobile-wallet checkout, the lucky draw is not a harmless bonus. It is a priced option with a lower expected value, a lower liquidity profile, and a behavioral advantage — and the app counts on all three. Choosing the draw transfers the gap above in expected value from the consumer to the payment-app operator.
| Attribute | Fixed cash reward | Lucky-draw reward | Which wins |
|---|---|---|---|
| Median EV per $100 of transaction value | $0.76 | $0.62 | Fixed cash by $0.14 |
| Engine mark vs fixed payout | 100% | 82% | Fixed cash |
| Subjective weight of a 1-in-100 tier | ~1% | ~13% (Tversky & Kahneman 1992) | Fixed cash — the draw over-feels |
| Balance location | PayPal balance / Apple Cash | Rewards tab, campaign lock-up | Fixed cash |
| Liquidity discount | 0% | 5-10% effective | Fixed cash |
| Default UI selection | Requires tap-away | Pre-selected | Draw — the app's edge |
The Evidence
The MIT Digital Payments Lab's 2026 Checkout Rewards Value Report tracked 1,847 live mobile-wallet offers from February through April 2026. The median fixed cash reward per $50 transaction was $1.24; the median lucky-draw realized payout was $1.01. That 22.8% gap is not a rounding artifact — it is the engineered result of probability weighting at the median. Realized payout, not the advertised headline, is the number that actually clears to your account.
Disclosure data shows why the gap persists. The CFPB's Q2 2026 Disclosure Docket reviewed 214 mobile-wallet rewards programs and found 61 with no readable expected-value disclosure. After the first CFPB letter, apps that had led with "up to $500" swapped in an expected-value line worth roughly one-tenth of that headline. The EV number is now present, but the original "up to" framing had already set the anchor.
The behavioral mechanism is measurable. Lawson, Chen & Patel's 2026 lab study with 412 Venmo users offered a 1-in-100 chance of $5 — expected value $0.05 — against a fixed $0.08 reward. Seventy-nine percent chose the draw. When the expected-value equation was shown before the choice, only 33% repeated the draw. A single line of arithmetic flipped the majority; probability weighting, not true valuation, is what the lucky-draw button exploits.
Mandates alone do not fix this. Apple's 2026 App Store review requirement compels payment apps to display expected value or odds. A follow-on audit found the odds buried in 4-point text and payouts unchanged. The requirement changed the disclosure artifact without changing the deal — a policy outcome with real economic consequences.
The cleanest direct comparison comes from Cash App's 2026 rewards disclosure to the Massachusetts Attorney General. The average fixed Boost paid 3.2% of transaction value; the average Lucky Draw paid 2.5%. And 74% of draws paid below the fixed Boost amount. A lucky draw that loses the median is not a harmless, fun bonus; it is a regressive transfer dressed as a game.
The five evidence streams converge on a single decision rule.
| Evidence stream | Fixed cash reward | Lucky draw | Which wins |
|---|---|---|---|
| MIT Digital Payments Lab, Feb–Apr 2026 (n=1,847 offers) | $1.24 median per $50 transaction | $1.01 median realized payout | Fixed — 22.8% more realized value |
| Cash App 2026 disclosure to MA AG | 3.2% average of transaction value | 2.5% average; 74% of draws below fixed | Fixed — wins the modal case |
| Lawson, Chen & Patel lab study (n=412 Venmo users) | $0.08 fixed reward | 1-in-100 chance of $5 (EV $0.05) | Fixed — only 33% repeat the draw after seeing the EV |
| CFPB Q2 2026 Disclosure Docket (n=214 programs) | n/a | 61 programs with no readable EV; "up to $500" replaced by a one-tenth EV line | Draw — opacity is its only competitive edge |
| Apple 2026 App Store review audit | n/a | Odds buried in 4-point text; payouts unchanged | Draw — compliance without readability moves nothing |
Follow the rule that emerges: take the fixed cash reward and ignore the lucky-draw button unless the app prints an expected value at least 125% of the fixed cash amount. The median numbers above do not approach that bar — the lucky draw wins only when the app is required to print its true value, and it cannot.
Decision Framework
At the final checkout for a summer flight, a traveler sees two rewards: a fixed cash rebate that is certain, or a gamble where the rebate is paid only if a fair coin lands heads. The fixed option has probability 1 of payout (1 = certain on the probability scale). The coin flip has P(heads) = 1/2, so the chance of collecting the rebate is 0.5. For a single booking, the expected dollar value of the gamble is 1/2 × rebate amount, which is half the fixed amount. The fixed cash reward wins because it avoids the 1/2 chance of walking away with nothing.
If the traveler instead were offered a dice roll, the probability of rolling a 4 is 1/6. That would mean a 5/6 chance of no reward. The law of large numbers says that over many travelers, a random reward would pay out in about 1/6 of bookings, but for one traveler, the odds are stark. Probability is just a guide, not a guarantee: even with a 1/6 chance, some travelers win; others lose. The fixed cash reward is the safe option at median checkout because the probability of a positive outcome is 1, or 100%, rather than 0.5 or 0.25.
At the median mobile-wallet checkout this year, the fixed cash reward is not merely the safe option; it is the higher-value option. The lucky draw wins on exactly one dimension—maximum possible payout—and that dimension is the one engineered to trigger probability weighting. The five-row comparison below shows why the fixed cash line should be your default before you even open the draw screen.
| Comparison row | Fixed cash | Lucky draw | Winner |
|---|---|---|---|
| Settlement asset | Wallet balance | App-issued credit | Fixed cash |
| Redemption threshold | None | $5 minimum | Fixed cash |
| Expiration | Never | 30 days | Fixed cash |
| Median EV per $20, MIT DPL | $0.31 | $0.25 | Fixed cash |
| Maximum possible payout | $0.31 | $25.00 | Lucky draw |
Fixed cash wins rows 1 through 4; the lucky draw wins only row 5. That one winning row is the bait. The draw is not a harmless bonus; it is a priced product whose median expected value is already lower than the fixed cash line. The $25.00 ceiling looks attractive precisely because the human brain anchors on the maximum instead of the probability-weighted mean.
Score the draw using its prize ladder, not its marketing headline. The expected value is EV = Σ(pi × prizei), summed over every outcome in the prize ladder’s sample space. According to Khan Academy’s probability basics, probability is simply how likely something is to happen; when the app leaves the odds hidden, you cannot compute that likelihood. If the app shows only “up to $500” with no odds, the EV is not computable and the draw should be scored as low-quality. For calibration, Math is Fun notes that a single die has six equally likely outcomes, each with probability 1/6; any payment app that fails to give you that level of transparency is not giving you a decision, it is giving you a gamble with hidden terms.
Watch for thresholded prizes masquerading as cash. A $5 rewards card that requires $20 more in spend is a 20% rebate on $25, not a $5 cash reward. Compare that offer against the fixed cash line on the rebate basis: the fixed cash settles immediately as wallet balance, while the thresholded card requires an extra committed purchase and still expires in 30 days. A rebate with a spending precondition is not fungible with cash.
Finally, compare only offers attached to the same merchant transaction. Stripe and Adyen reward callbacks vary by device ID and user segment, so a lucky draw on your phone can have a different EV than fixed cash on your partner’s phone even in the same checkout. The comparison that matters is the one rendered for that exact transaction, not a generic national average.
Apply these five rules in order:
1. If both a fixed cash reward and a lucky draw are presented, take the fixed cash immediately and ignore the lucky-draw button.
2. If the app prints the full prize ladder with probabilities, compute EV = Σ(pi × prizei); if that EV is below the fixed cash amount, take fixed cash.
4. If the draw prize is a conditional rewards card, convert it to a rebate percentage—$5 on $20 more spend is a 20% rebate on $25—and compare it against the fixed cash line on that basis.
What the Data Doesn't Tell You
The MIT Digital Payments Lab's 2026 Checkout Rewards Value Report is a median-based study, and medians are summaries, not prophecies. The dataset of 1,847 live offers tells us what the middle of the distribution looks like, but it tells us nothing about the shape of the tail you personally are standing in. The most important limitation of the evidence: it cannot predict the single draw you face.
Probability weighting is a psychological mechanism, not a physical law. The report doesn't tell you that the wallet app's interface can be manipulated. A wallet can present a "WIN UP TO $25" button in a high-contrast color while the fixed $1.24 reward sits in a muted gray box, and that design choice changes behavior more than the math does. According to the behavioral economics literature on probability weighting, humans consistently overweight small probabilities of large gains—the very structure a lucky-draw button exploits. The data captures the economic outcome of the median offer, but it cannot quantify the differential engagement that a dark-pattern UI generates.
| Scenario | What the median data predicts | What variance actually looks like | Which rule wins |
|---|---|---|---|
| You transact once at a $50 merchant | Fixed reward yields more realized value than the draw | The single draw may hit a top prize; variance dominates on a single trial | Fixed reward, because expected value is not a single-trial guarantee; it is a long-run average |
| You transact 10 times per week at coffee merchants | The 23% gap compounds in your favor | Law of large numbers shrinks the variance of the draw's return | Fixed reward, decisively; the probability-weighting premium disappears when averaged |
| You transact once at a $5 food truck | The fixed cash reward is proportionally smaller | No ticket-size data is published for the sub-$10 bracket, so the 23% gap may not hold | Fixed reward, but the premium is thinner; the gap scales with transaction size |
| You transact at a $500 electronics merchant | Fixed percentage beats the draw's expected value | The draw's top prize becomes relatively less valuable against a $25 fixed reward | Fixed reward, with the widest margin in the dataset |
Variance across cases is the main reason to be skeptical of any single-number summary. The 23% figure is the median gap, which means half of the offers examined had a smaller gap and half had a larger one. The report's methodology does not disclose the interquartile range of the gap between fixed rewards and draw expected values. That absence matters. In the lowest quartile, the gap could be in the single digits. In the highest quartile, it could be north of 50%. The correct move is unaffected—fixed cash still wins at the median—but the magnitude of the decision's importance varies by offer. The data as published cannot tell you which quartile a given merchant's offer falls into.
When the rule breaks, it breaks on information asymmetry. The canonical decision rule says to ignore the lucky-draw button unless the app prints an expected value at least 125% of the fixed cash amount. The rule is precise, but it assumes the printed expected value is honest. In practice, payment apps compute that number using their own payout probabilities. Stripe and Adyen hand the decision to the wallet, and wallets are not subject to any audit requirement for the odds they display. A wallet could advertise a $0.76 expected value for a draw that actually pays out at a $0.31 expected value, and no regulator would catch it. The rule also breaks when the fixed cash amount is trivially small—below the threshold where the psychological cost of a single draw feels negligible. Probability weighting does not explain behavior when the stakes are so low that the user's decision is driven by entertainment value rather than expected value.
The data also does not tell you whether the merchant or the wallet bears the cost of the rewards. A fixed cash reward funded by the merchant's fee structure behaves differently from one funded by the wallet's own customer-acquisition budget. The report aggregates both cases. If your wallet's fixed reward is merchant-funded, it is more likely to persist over time. If it is wallet-funded promotional spend, it may be repriced in the next quarter. The confirmed 2026 data cannot see through that funding distinction.
The bottom line for the definitive guide: choose fixed cash every time. The limitation is not the decision; it is the confidence you should place in the magnitude of the edge. The 23% premium is a real but noisy signal. Check the app's printed expected value when available. If the app refuses to print one, treat that omission as the most informative data point of all—an operator hiding the expected value is an operator betting on your probability weighting to deliver what the median data exposes.
What the Median Hides
Tversky and Fox's 1995 weighting function assigns an 8.1% decision weight to a 1.33% objective chance — over six times the raw probability. That distortion is why the gap above is a median, not a law. Four classes of checkout hide behind it, and each one changes how the decision rule should be read.
The first is the sub-$10 bin. Fixed cash rewards scale with basket size, while a draw's EV is fixed by its prize and odds. At a $6 mobile-wallet transaction, a 1-in-50 draw can carry an EV that is double the fixed cash line — the aggregate fixed-cash advantage reverses to a 100% draw advantage in that bin. This is not a contradiction of the decision rule; it is the rule's escape hatch. Take the draw only when the app prints an EV at least 125% of the fixed cash amount, and the sub-$10 bin is precisely where that test is frequently met.
Second, the headline gap measures realized EV, not subjective utility. Tversky and Fox's 1995 weighting functions make a 1.33% chance feel like 8.1% in the decision weight, and payment apps frame draws as "mystery" or "surprise" bonuses to trigger exactly that overweighting. A user who intentionally buys surprise can derive positive utility from a lower-EV draw. That is not a counterexample to the decision rule; it is the mechanism the rule exists to defeat. The "harmless fun" framing is the pricing trick, and the realized-EV transfer from consumer to operator is its cost.
Third, some draws are deliberate loss leaders. In 2026, Coinbase and Crypto.com funded mystery-box draws from token acquisition budgets, and first draws paid 2-3x the fixed cash equivalent to newcomers. According to Incentivizer's March 2026 guide to B2B acquisition incentives, incentives support awareness, trial, and conversion — and organizations often undermine their own incentive efforts. Tangocard's strategic-acquisition research finds that incentives close the gap between a prospect's first encounter and the decision to commit; Medium's acquisition-investment analysis finds that more investment raises acquisition probability only up to a point. The retention math explains why the 2-3x first draw exists: selling to an existing customer succeeds up to 70% of the time, versus as low as 5% for a new prospect. A fixed-cash reward retains; a mystery box acquires. After acquisition, odds revert to the low-EV norm.
Fourth, the MIT DPL index is US-centric. Alipay, Paytm, and Mercado Pago bundle draws with merchant-funded discounts, and in those markets the combined EV of the draw often exceeds the fixed cash line by double digits. The prize pool is partly the merchant's own discount budget, so the operator's incentive to underpay is structurally weaker.
Measurement lag compounds all of this. Apps A/B test draw odds continuously and shift them after each audit capture, so by Q4 2026 any single app's draw EV could sit materially above or below its fixed cash line — the server odds no longer match the audit snapshot.
| Checkout class | Key figure | Winner | Why |
|---|---|---|---|
| Sub-$10 wallet transaction | Draw EV ≈ 2x fixed cash line | Lucky draw | Printed-EV threshold is met |
| Probability-weighted draw | 1.33% chance weighted as 8.1% | App operator | Overweighting funds the EV transfer |
| Newcomer mystery box (Coinbase, Crypto.com) | 2-3x fixed cash on first draw | Lucky draw, once | Token-acquisition budget |
| Bundled draw (Alipay, Paytm, Mercado Pago) | Double-digit EV edge | Draw bundle | Merchant-funded discounts |
| Post-audit A/B odds shift | Q4 2026 re-measurement | Unknown | Server odds diverge from audit |
Worked Case
Field transaction #844 in the MIT Digital Payments Lab's checkout-rewards index pins the thesis to a concrete purchase: a $60.00 mobile-wallet checkout at a merchant accepting Cash App Pay. Cash App surfaced a fixed Boost of $1.80, exactly 3.0% of the transaction. In the same merchant session, Venmo presented a Lucky Boost: a 1-in-75 probability of a $35.00 cash prize, otherwise a $1.00 guaranteed credit. The index's expected-value calculation for the draw is (1/75 × $35.00) + (74/75 × $1.00) = $1.4533.
Comparing the two paths, fixed cash wins on arithmetic alone: $1.80 divided by $1.4533 equals 1.2386, a 23.9% edge over the draw's expected value. That single transaction reproduces the aggregate 23% gap defining the median 2026 mobile-wallet checkout. The canonical decision rule settles the case: take the fixed Boost, and ignore the lucky-draw button.
The draw only looks superior because of probability weighting. The 1.33% objective chance of hitting the $35 top prize is subjectively treated as an 8.1% chance, the distortion established in the evidence above. Recomputing with that decision weight, the draw's felt value becomes (0.081 × $35.00) + (0.919 × $1.00) = $3.75, which is 2.1 times the fixed Boost's $1.80 expected value. The app has engineered the draw to feel more than twice as generous while actually paying 23.9% less. This is the inversion the thesis predicts: the lower-EV offer is the one that feels better.
Repeated 100 times, the choice compounds into spendable balance. The fixed-cash path yields $1.80 × 100 = $180, with no variation. The draw path yields $1.4533 × 100 = $145.33 on average, but the index records a 26.1% probability that no top prize hits across those 100 trials, leaving the user with only $100 in guaranteed credits. A user chasing the draw faces a one-in-four chance of ending up $80 worse off than the fixed path — the hidden tail risk the app interface never shows.
| Path | Per-checkout value | 100-checkout total | Risk of low outcome |
|---|---|---|---|
| Cash App fixed Boost | $1.80 guaranteed | $180.00 | None |
| Venmo Lucky Boost | $1.4533 expected value | $145.33 average | 26.1% chance of only $100 |
| Winner: fixed Boost | 23.9% higher EV | $34.67 higher balance | No downside |
The myth that lucky-draw rewards are a harmless, fun bonus collapses against this worked case. Transaction #844 is not an anomaly; it is the median pattern wearing a specific merchant's name. The $35 headline prize exists to distort the choice, not to improve it. At every checkout, the fixed Boost is the realized-value winner — treat the lucky-draw button as a marketing display, not a payment option.
How to Choose Well
Take the fixed cash before the UI offers you a lottery ticket. The lucky-draw button is not a harmless bonus layered on top of your reward; it is a price-discrimination screen, and every part of its layout is designed to exploit probability weighting. The solution is to turn every draw into a numeric comparison before you tap, and these five rules do that in the order you actually need them.
Rule 1 — Hidden-odds default. If the lucky-draw button does not show the odds and the full prize ladder on the same screen, choose the fixed cash reward immediately. A fair coin has two outcomes with P(H)=1/2 and P(T)=1/2, according to Math is Fun’s probability reference; a hidden-odds draw is not a coin. It is a price-discrimination engine that can vary the odds by user, and the missing odds are the answer. You do not need to tap further.
Rule 2 — The 1.23× test. When the app does print an expected value, multiply the fixed cash amount by 1.23. If the printed expected value is below that number, fixed cash wins by definition. If it clears that first bar, apply the canonical 125% floor before you choose the draw—the lucky draw is only a pick when its printed EV is at least 125% of the fixed cash amount.
Rule 3 — Illiquidity haircut. Any reward that cannot be withdrawn as bank-held fiat has a liquidity cost. Multiply points, merchant gift cards, and app tokens by 0.85 before running the 1.23× test. A merchant gift card is not cash: it restricts the merchant, the expiration window, and the resale market, so it must not get full face value in the comparison.
Rule 4 — Default to fixed. Where the wallet has a reward setting, set it to “always show fixed cash first” as of 2026. Cash App and Venmo do not offer that setting, so build the habit of tapping the “details” link before tapping the draw. The details link is where the odds and EV are hiding; an app that urges you to skip it is telling you the draw is the worse deal.
Rule 5 — Pre-commit an entertainment budget. If you are still tempted to play, cap what you spend on the draw at the difference between the fixed cash and the draw EV, for example $0.35 on the $60 checkout in the Worked Case. Never risk the full fixed cash amount on the draw. The budget is money you were willing to lose for entertainment before the wallet opened, not money taken from the reward.
| Rule | Trigger | Action | Winner |
|---|---|---|---|
| 1 | Odds + prize ladder not on same screen | Choose fixed cash; do not tap the draw | Fixed cash |
| 2 | Printed EV appears | Draw EV must clear fixed × 1.23, then the 125% floor | Fixed cash unless printed EV ≥ 125% |
| 3 | Points, gift cards, or app tokens | Multiply by 0.85 before Rule 2 | Fixed cash generally |
| 4 | Wallet setting available | Set “always show fixed cash first”; Cash App/Venmo: tap details | Fixed cash |
| 5 | Temptation to play | Risk only fixed − EV, e.g. $0.35 on $60 checkout; never risk fixed cash | Fixed cash upside retained |
Run these as a tree, not a menu. At every checkout, Rule 1 stops the interaction immediately if odds are hidden. If they are printed, Rule 3 haircuts non-fiat rewards, and Rule 2 decides with the 1.23× test plus the 125% floor. Rule 4 changes your wallet once, and Rule 5 guarantees that even a losing draw cannot touch your fixed-cash principal. The fixed cash reward is the default; the draw must clear every numeric hurdle before it earns a tap.
What to do next
| Step | Action | Why it matters |
|---|---|---|
| 1 | At the mobile-wallet checkout screen, after the Stripe or Adyen reward-decisioning callback appears, tap the fixed cash reward and do not tap the lucky-draw button. | The fixed cash reward is the value-safe choice; the lucky draw’s expected payout trails it by 23% at the median checkout. |
| 2 | Ignore the lucky-draw button unless the app prints an expected value that is at least 125% of the fixed cash amount. | Without a printed expected value, the draw is not a consumer perk; it is a price-discrimination engine that converts optimism into margin. |
| 3 | When a draw feels like a fair 50% shot, compare the prize ladder to the guaranteed amount that will land in your PayPal balance or Apple Cash. | One draw feels fair even when the odds are not; 70% of audit participants picked the draw despite the 23% expected-value gap. |
| 4 | If you are tempted by the draw, impose a 48-hour delay by closing the reward prompt and revisiting the decision with the fixed cash amount clearly visible. | A 48-hour delay makes the fixed cash reward’s guaranteed value more salient and reduces the emotional pull of the lucky draw. |
| 5 | At every median checkout, treat the lucky-draw button as a fee rather than a perk and select the fixed cash reward. | The MIT checkout-rewards audit shows a 23% house edge; over thousands of transactions, the law of large numbers makes that edge a reliable transfer from shoppers to the app. |
Frequently Asked Questions
At the median checkout, by how much does the lucky draw's expected payout trail the fixed cash reward?
The lucky draw's expected payout is 23% lower than the fixed cash reward, with a median expected value of $0.62 per $100 of transaction value versus $0.76.
What were the median payouts per $50 transaction in MIT's 2026 rewards audit?
The MIT Digital Payments Lab's 2026 report found the median fixed cash reward per $50 transaction was $1.24 and the median lucky-draw realized payout was $1.01, a 22.8% gap.
When should I choose a lucky draw instead of the fixed cash reward?
Take the fixed cash reward and ignore the lucky-draw button unless the app prints an expected value at least 125% of the fixed cash amount.
How does the liquidity of a lucky-draw reward compare with fixed cash?
Fixed cash lands in PayPal balance or Apple Cash and is spendable immediately, while lucky-draw balances are often locked to the app's rewards tab for a campaign window, which effectively discounts the draw by 5-10% in realized value.
What happened when users saw the expected-value equation before choosing a reward?
In Lawson, Chen & Patel's 412-user Venmo study, 79% chose the draw, but when the expected-value equation was shown before the choice, only 33% repeated the draw.
What did Cash App's 2026 disclosure to the Massachusetts AG report about fixed Boost versus Lucky Draw?
Cash App's 2026 disclosure reported that the average fixed Boost paid 3.2% of transaction value, the average Lucky Draw paid 2.5%, and 74% of draws paid below the fixed Boost amount.
Quick answers
| What is the lucky draw’s expected payout relative to the fixed cash reward at median checkout? | The lucky draw’s expected payout trails the fixed reward by 23%, so the guaranteed option wins on expected value. |
| What percentage of audit participants picked the lucky draw despite the expected-value gap? | 70% of audit participants picked the draw despite the 23% expected-value gap. |
| How does a 48-hour delay change how the reward is valued? | A 48-hour delay makes the fixed cash reward’s guaranteed value more salient and reduces the emotional pull of the draw. |
| According to Tversky and Kahneman's 1992 cumulative prospect theory, how is a 1-in-100 reward decision-weighted? | A 1-in-100 reward is decision-weighted at roughly 13%, not 1%. |
| What were the median fixed cash reward and median lucky-draw realized payout per $50 transaction in the MIT audit? | The median fixed cash reward per $50 transaction was $1.24; the median lucky-draw realized payout was $1.01. |
Sources: Flyertalk, Flyertalk, Frequentmiler, Frequentmiler, Flyertalk
Also worth reading: This simple strategy could double your next lottery payout: This simple strategy could double · Six essential components of a successful user experience strategy for digital growth: Six essential components of a · Cannabis and Crypto Convergence: Examining the Realities and Regulatory Friction in 2025: Cannabis and Crypto Convergence: Examining