Dating
Hinge
A dating app designed to get you off dating apps — with an increasingly sophisticated business built around what happens before you leave.
Hinge is the most difficult case in the CHI dating investigations, because the evidence for the defence is as strong as the evidence for the charge. Hinge documents mechanics that actively work against the incentives of its own category: Your Turn Limits stops daters — including paying subscribers — from sending or receiving new Likes while too many matches wait on a reply. Help Centre Signals awards a visible badge for participation behaviour and states plainly that it cannot be bought. Help Centre We Met asks whether a match became an actual date. Help Centre Those are not cosmetic. They are the opposite of engagement maximisation, and this assessment credits them. Analytical inference The charge is what Hinge built alongside them: Standouts identifies the profiles catching the most attention, replaces ordinary Likes with purchasable Roses inside that feed, and warns that a Standout may never reappear in Discover. Help Centre HingeX sells being seen sooner, seen by more people, and staying at the top of someone's Likes You page for seven days. Help Centre
CFS = CVI − CHI (81 − 57 = +24). Assessed under CHI/CVI Methodology v2.0: dimensions score the underlying customer harm, patterns describe the behaviour that produced it. A CHI of 57 alongside a CVI of 81 is not a contradiction and should not be read as one — it is the split framework doing exactly what it exists to do. Hinge delivers substantial, measurable customer value and operates a meaningful hostility architecture. Both are true. Methodology →
The defining tension
Hinge appears to have built a genuinely thoughtful dating product, then layered an increasingly aggressive attention marketplace around parts of it.
There is a real difference between building a product to solve a customer's problem and then working out how to finance it sustainably, and building a marketplace and then continuously finding new places where desire, competition, scarcity and uncertainty can be converted into revenue. Descriptive framing
Monetization itself is not the finding. A dating company needs revenue, and this assessment does not treat charging money as hostile. The CHI concern begins at a narrower point: when the friction inherent in accomplishing the customer's objective becomes something the company has an incentive to preserve, manufacture, or price. Analytical inference
Its strongest product mechanics are built to get customers into real relationships. Its strongest revenue mechanics sell customers better position in the race to get there.
Read this first · Evidence against hostility
What Hinge actually got right.
This section leads rather than trails, because burying it would misrepresent the finding. These are the strongest customer-aligned dating mechanics identified anywhere across the CHI dating investigations, and they are the evidence behind a CVI of 81 — the highest CVI awarded to a dating company in this index.
Your Turn Limits. Hinge documents that "when too many people are waiting for a response, daters will not be able to send or receive new Likes until they take action." Help Centre This attacks match hoarding, endless accumulation, poor responsiveness and quantity-over-quality behaviour directly. Critically, Hinge+ and HingeX subscribers are not simply exempted — Hinge states the limit is not a cap on how many Likes a subscriber may send, but subscribers still have to respond or close conversations before engaging with new ones. A monetization-first company would have sold the exemption. Hinge did not.
Signals. The Signals badge is awarded on participation behaviour — looking before liking, sending comments, reviewing incoming likes, messaging matches, confirming dates — meeting three of five criteria. Hinge states flatly that "daters cannot purchase a Signals badge" and that eligibility is "based on requirements related to participation on Hinge, not your subscription status." Help Centre A visible status marker that money cannot buy, in a category where almost every other status marker is for sale, is the single strongest customer-aligned mechanism found in these investigations.
We Met. A few days after numbers are exchanged, Hinge asks whether a date actually happened and whether the dater would want to see that person again, and says it uses the answer to improve recommendations. Help Centre This matters structurally: it moves the measured definition of success downstream from app open → like → match toward real human interaction. Most of the category never attempts that measurement at all.
Profile-specific interaction and mutual-preference matching. Hinge's core interaction requires liking or commenting on a specific photo or prompt rather than a whole-person swipe, and recommendations are built around stated preferences on both sides. The design cost is deliberate: it is slower and more effortful than swiping, which is the point.
Pausing and notification control. Daters can pause their profile — becoming invisible in Discover and Standouts while keeping existing matches — and control notification categories. Voluntary, reversible invisibility is the opposite of the retention architecture CHI usually documents.
Geographic relevance controls work. Daters set where they want to date and adjust distance. The investigation specifically tested for a Hinge equivalent of the Luxy finding — an expensive premium subscription repeatedly producing geographically impractical inventory — and did not establish one. Not established This is recorded as a positive distinction, not merely an absence of harm.
Stronger verification. Selfie Verification and the biometric Face Check scan address real identity fraud, catfishing and ban-evasion problems in a category that has historically tolerated all three. Help Centre The privacy cost is real and is scored separately below; the safety justification is also real.
Refund handling, observed firsthand. A CHI investigator whose account was deactivated received a refund of a recent premium purchase without resistance. Fact — directly observed Hinge's published terms say virtual items are non-refundable and that unused value is not returned on an involuntary closure; the conduct observed was more generous than the contract required. That is credited here explicitly.
CVI 81/100 = Core Product Value 25/30 + Feature & Capability Improvements 22/25 + Technology & Performance 16/20 + Trust, Safety & Reliability 9/15 + Innovation 9/10. Trust, Safety & Reliability is the one materially depressed CVI dimension, and the reason is documented in the safety section below — not a general scepticism about the product.
Supported patterns
What the evidence actually supports.
Primary pattern
Major supporting patterns
Moderate / moderate–high patterns
Artificial Scarcity · Hero exhibit
The same person. Two feeds. Two prices.
What Hinge's own documentation says happens when a profile is "catching the most attention."
In Discover
Freean ordinary Like
- Like a specific photo or prompt
- Comment to add context
- Lands in their Likes You, in order
In Standouts
Roses only"you can only send Roses instead of Likes"
- One free Rose per week, non-accumulating
- More sold in packs of 3, 12 or 50
- A Rose "moves to the top of their Likes You"
Hinge states that "while some Standouts may later appear in Discover, it's not guaranteed." Help Centre So the customer faces a refreshing feed of profiles the platform itself has identified as high-attention, a currency requirement to act inside it, and no assurance the profile returns. Every element of that sentence is Hinge's own documentation, not inference.
Scored once, across Revenue Extraction and Behavioral Manipulation. The underlying fact that some profiles receive more attention than others is real and is not scored — what is scored is the decision to separate that inventory into a currency-gated feed. Analytical inference
Evidentiary limit · Stated explicitly
Where this page stops, and why.
Established
Artificial scarcity and monetized priority. Hinge's own documentation establishes that Standouts is a curated high-attention feed, that ordinary Likes are unavailable inside it, that Roses are the required currency, that Roses confer top-of-queue positioning, that additional Roses are sold, that the feed refreshes daily, and that reappearance in Discover is not guaranteed.
No inference is required to reach any part of that. It is quoted from Hinge's Help Centre.
Not established
Deliberate suppression of a dater's best matches to force Rose purchases. This page does not claim that Hinge hides every user's best matches from Discover in order to compel purchases. That is a materially stronger allegation and the investigation did not establish it. Not established
The difference matters. One claim describes a documented commercial architecture. The other describes a deliberate deception, and CHI does not assert deliberate deception without evidence of it.
Mission–monetization tension · Principal finding
Customer success ends the relationship. Corporate success requires more payers.
Hinge built its identity around successful relationships and the idea that a successful customer eventually leaves. Its owner reports to public markets on payer growth and revenue per payer. Both of those things are true at the same time, and neither requires bad faith.
What the product measures
Whether a match became a real date (We Met). Whether a dater is participating rather than accumulating (Signals). Whether existing matches are being answered before new ones are collected (Your Turn Limits). Help Centre
These are the metrics of a product trying to finish the job.
What the owner measures
Match Group reported that "Hinge grew overall revenue 22% Y/Y" in Q2 2026 with global MAU up 13% Y/Y, and reaffirmed that "Hinge is still expected to reach $1 billion in revenue in 2027, driven by continued product innovation, international expansion, and monetization gains." Company disclosure, Q2 2026 results
These are the metrics of a business that needs the customer base to keep growing and keep paying.
The conclusion this page does not draw: that Hinge deliberately wants customers to stay single.
There is no evidence establishing that, and asserting it would be both unfair and analytically lazy. Not established The supported conclusion is narrower and more durable: customer success ends the individual customer's relationship with Hinge, while corporate financial success requires continuing payer growth and increasingly effective monetization of the overall customer base. Those objectives can coexist. They also create a genuine structural incentive tension, and structural incentive tension is precisely what CHI exists to document. Analytical inference
Ownership & development
How Hinge got from there to here.
This is a chronology of evolving incentives, not a corporate villain story. Match Group supplied capital, scale, distribution expertise and growth infrastructure that materially contributed to Hinge's success, and the evidence does not support "Match Group acquired Hinge and ruined it."
- 2016Hinge reboots around an explicit philosophy of intentional relationships rather than engagement and retention — the swipe is removed, the interaction becomes profile-specific, and the product is positioned as the anti-Tinder. Contemporaneous reporting
- Sep 2017Match Group takes an initial investment position following the redesign, after which Hinge's user base grows sharply. Company announcement
- Jun 2018Match Group announces acquisition of a 51% controlling stake, stating it "has the right to acquire all the remaining shares of Hinge within the next 12 months." Company announcement, 20 June 2018
- 2018We Met launches — the first serious attempt in the category to measure whether an app interaction produced a real-world date. Trade reporting
- 2019Match Group completes full ownership of Hinge. Company disclosure
- 2021–22Standouts and Roses arrive: a curated high-attention feed in which ordinary Likes are replaced by a purchasable currency. Help Centre
- 2022–23HingeX and the expanded subscription architecture follow — priority likes, "seen sooner," subscriber-exclusive profile boosting, advanced filtering. Help Centre
- 2024–26Hinge becomes one of Match Group's principal growth assets. Match Group names monetization gains explicitly as a driver and reaffirms the $1bn 2027 revenue objective. Company disclosure, Q2 2026 results In parallel, Hinge ships Your Turn Limits, Signals, Selfie Verification and biometric Face Check. Help Centre
Read the last two rows together. The same period that produced Hinge's most aggressive monetization architecture also produced its most customer-aligned product mechanics and its strongest safety controls. Any account of Hinge that describes only one of those halves is describing a different company. Analytical inference
Safety / Integrity Failure · Historically severe
The curation system elevated a user its safety system had already received a rape report about.
This finding is treated separately from monetization and is not equivalent to a complaint about Rose pricing. It is also not sensationalised: what follows is the documented sequence, with the evidentiary status of each step stated.
Factual status of the Standout appearance — stated precisely, because it is frequently mischaracterised.
The Standout elevation is not merely an allegation contained in the later civil lawsuit. It was documented in February 2025 by an 18-month investigation — the Dating Apps Reporting Project, conducted in partnership with the Pulitzer Center, The Markup/CalMatters, The Guardian and The 19th — which reported the classmate's encounter with his promoted Standout profile in 2022. Investigative finding A civil suit filed in Denver on 17 December 2025 by six women adds further allegations against Hinge, including that he was reported in September 2020 and January 2021, was said to have been permanently banned, and reappeared using identical information. Allegation · unadjudicated Those pleaded claims are labelled as allegations throughout this page. The Standout finding does not rest on them.
Matthews was convicted on 35 counts involving drugging and sexual assault and was sentenced in October 2024 to 158 years to life. Trial record
Why this is not simply "a dangerous person slipped through"
Platforms of any scale will be used by dangerous people, and CHI does not score a platform for the existence of bad actors. The sequence here is more serious than a screening miss: serious conduct reported → purported enforcement → continued platform presence → proprietary algorithmic elevation. Analytical inference
Standouts is not a neutral chronological database. Hinge selects those profiles and presents them as catching the most attention.
What it implicates
Because Standouts is a curated, platform-authored recommendation surface — and because interaction inside it runs through a purchased currency — the failure reaches further than trust and safety alone. It implicates recommendation integrity, platform-implied desirability, algorithmic curation, and monetized Rose interaction at the same time. Analytical inference
A ranking system that says "this person is catching the most attention" is making a statement. It made that statement about him.
This is the principal reason CVI's Trust, Safety & Reliability dimension is scored 9/15 rather than in the range Hinge's other product work would otherwise support, and it is a material component of the Trust & Transparency CHI dimension. It is recorded as a historical severity, not as a claim that today's systems are unchanged — see immediately below.
Current safety · Credit where it is due
Historically severe; current status materially improving but not proven fully resolved.
Current Hinge is not identical to historical Hinge, and this page does not imply that past failures prove present systems are unchanged. Nor does it certify them as fixed.
Face Check, argued honestly — the safety case
Hinge responded to a genuine identity and ban-evasion problem with materially stronger biometric controls. Liveness verification, face matching, identity consistency, age estimation where applicable, and detection of duplicate or returning banned identities are direct answers to documented harms in this category — including the harm documented in the section above. Analytical inference
This page does not write "Hinge uses invasive biometrics for no legitimate reason." That characterisation would be false.
Face Check, argued honestly — the privacy cost
Hinge states that "your facial geometry, which may be considered 'biometric information' in some places, will be used and stored to estimate your age, check your profile photos, reduce fraud, and help detect whether other profiles are using your likeness," and that FaceMaps and FaceVectors "are retained for the lifetime of an account, and are deleted within 30 days of account closure," with audit images kept 90 days after closure. Help Centre
Where the check is mandatory, that is persistent biometric processing as a condition of participating in dating. The correct formulation is the tradeoff itself: Hinge has answered real identity and ban-evasion problems with materially stronger biometric controls, but doing so requires deeper and potentially persistent biometric processing from users in affected markets.
Trust & Transparency · Enforcement and firsthand evidence
Banned, unexplained, and refunded.
Firsthand CHI evidence — enforcement transparency: negative
A CHI investigator was deactivated following an apparent report. The investigator had not used Hinge for well over thirty days before the deactivation, had recently purchased premium, and contacted Hinge asking about the action. No meaningful explanation of the reason was ever provided. Fact — directly observed
This page draws no inference about whether the investigator was innocent, or whether the action was justified. It cannot, and it does not.
Firsthand CHI evidence — refund handling: positive
The recent premium purchase was refunded without resistance. Fact — directly observed Hinge's terms did not require this: they state that no compensation is given for unused virtual items when an account closes, including involuntary closure through a ban. Terms of Service
The conduct was better than the contract. That is a genuine positive and is recorded as one.
The split finding is the point.
A single customer episode produced one clearly negative result and one clearly positive result, and CHI reports both at equal weight rather than selecting whichever supports a thesis. Enforcement transparency — negative. Refund handling — positive.
Live-audit limitation, stated as a boundary rather than a gap. Because the investigator's account is deactivated, a current longitudinal firsthand Hinge audit cannot be performed. This page therefore contains no claim of a current multi-day feed audit and invents no firsthand observations of Standouts inventory, Rose pricing behaviour in-app, or notification cadence. That is a frozen evidentiary boundary of this assessment, not an unfinished research task. Open item
Information & Privacy · Pricing capability and dual-use data
The same behavioural record that improves your matches improves your price.
Established
Hinge discloses that "our pricing varies based on a number of factors. We frequently offer promotional rates — which can vary based on region, length of subscription, bundle size, past purchases, account activity, and more." Terms of Service
Behaviour-, account- and purchase-sensitive pricing capability: established. Hinge has infrastructure capable of using information about the customer not only to improve matching, but to shape the commercial offer that customer sees. It discloses this rather than concealing it, which is credited — disclosure is why this is scored under Information & Privacy rather than under deception.
Not established
Abusive or discriminatory individualized pricing in practice. Not established The investigation found capability and disclosure. It did not find evidence that the capability is used to charge identifiable individuals unfair or discriminatory prices, and this page does not cross that evidentiary line.
The distinction is deliberate. "Can" and "does, abusively" are different findings, and conflating them would be exactly the kind of overreach CHI is built to avoid.
Hinge's privacy policy documents processing of account, profile, content, purchase, usage, technical, precise-geolocation, biometric and government-ID data, for purposes including recommendations, personalisation, trust and safety, fraud prevention, product development, marketing, promotions and advertising measurement — with data shared among Match Group affiliates for safety, customer care, analytics, targeted advertising, and "developing or improving Match Group Brands' core recommendation systems." Privacy Policy Dating necessarily requires intimate data, and collection alone is not equated with hostility here. The concern is narrower and is dual use: the same record can be used to make a dater's outcomes better and to make that dater more monetizable, and only one of those two uses is the reason the customer handed the data over. Analytical inference
What we refused to count
CHI is not simply compiling complaints.
Evidence & methodology
- Company documentationWhat is Standouts? — Hinge Help CentrePrimary source for the entire Artificial Scarcity finding: "a feed of profiles catching the most attention," "in Standouts, you can only send Roses instead of Likes," daily refresh, and "while some Standouts may later appear in Discover, it's not guaranteed." Quoted directly rather than paraphrased.
- Company documentationSubscription and Purchase Benefits — Hinge Help CentrePrimary source for Paid Visibility: HingeX "See Compatible People Sooner," the subscriber-exclusive feature that "boosts a subscriber's profile towards the top of daters' feeds," Priority Likes staying top of Likes You for 7 days, Boost and Superboost, and Rose pack sizes.
- Company documentationWhat is "Your Turn Limits"? — Hinge Help CentrePrimary source for the strongest customer-aligned finding, including the wording establishing that Hinge+ and HingeX subscribers are not simply exempt from the mechanism.
- Company documentationWhat is Signals? — Hinge Help CentrePrimary source for the five participation criteria and for the explicit statements that the badge cannot be purchased and does not depend on subscription status.
- Company documentationWhat is "We Met"? — Hinge Help CentrePrimary source for the downstream real-world outcome measurement, its timing, its privacy handling, and its use in recommendations.
- Company documentationFace Check™ Scan — Hinge Help CentrePrimary source for the biometric tradeoff: facial-geometry use and storage, the stated purposes, FaceMap/FaceVector retention for the lifetime of an account with deletion within 30 days of closure, 90-day audit-image retention, and phased rollout where local regulations permit.
- Company policyHinge Terms of Use AgreementPrimary source for pricing variation by "region, length of subscription, bundle size, past purchases, account activity, and more"; for termination without a refund and the statement that Hinge "may not share information with you regarding your account termination"; and for the non-refundability of Virtual Items including on involuntary closure.
- Company policyHinge Privacy PolicyPrimary source for the data categories processed — including precise geolocation, biometric and government-ID data — and for the stated purposes spanning recommendations, safety, product development, marketing and cross-brand Match Group recommendation systems.
- Investigative reportingDating app cover-up: how Tinder, Hinge and their corporate owner keep rape under wraps — The 19th, February 2025The controlling source for the safety finding. An 18-month Dating Apps Reporting Project with the Pulitzer Center, The Markup/CalMatters and The Guardian. Establishes the 28 September 2020 report, continued platform presence, and the 2022 encounter with his profile promoted as a Standout — independently of the later civil suit.
- LitigationSix women file civil suit against Hinge over Stephen Matthews — filed in Denver, 17 December 2025Used only for pleaded allegations, all labelled as allegations on this page: reports in September 2020 and January 2021, a stated permanent ban, and reappearance using identical information. Unadjudicated. The Standout finding does not rest on this filing.
- Trial recordDenver cardiologist sentenced to 158 years for sexual assaults — October 2024Source for the conviction on 35 counts and the sentence of 158 years to life. Established fact, not allegation.
- Company disclosureMatch Group announces second quarter results — Investor Relations, 4 August 2026Primary source for Hinge revenue growth of 22% Y/Y, global MAU up 13% Y/Y, and the reaffirmed expectation that "Hinge is still expected to reach $1 billion in revenue in 2027, driven by continued product innovation, international expansion, and monetization gains." Company statements, presented as such.
- Company announcementMatch Group expands portfolio with dating app Hinge — 20 June 2018Source for the 51% controlling stake, the right to acquire the remaining shares within 12 months, and the September 2017 initial investment following the swipe-free redesign.
- Company documentationSelfie Verification — Hinge Help CentreCompany documentation for the non-biometric-scan verification path. Recurring customer complaint themes referenced on this page — opaque bans, Standouts/Roses frustration, synchronisation issues — are treated as clearly labelled anecdotal patterns only, and are not cited as proof of any company practice where Hinge's own documentation establishes the mechanic directly.
Every mechanic scored on this page is quoted from Hinge's own documentation, not inferred from customer reports.
Evidentiary status is preserved throughout and is never upgraded for readability. Two boundaries are enforced explicitly and repeatedly: artificial scarcity and monetized priority are established; deliberate suppression of a dater's best matches to force Rose purchases is not. And behaviour-, account- and purchase-sensitive pricing capability is established; abusive or discriminatory individualized pricing in practice is not. Allegations contained in the December 2025 civil filing are labelled as allegations every time they appear and are never presented as findings; the Standout elevation is not among them, because it was independently documented in February 2025.
A current longitudinal firsthand Hinge audit could not be performed, because the CHI investigator's account is deactivated. No firsthand feed observations are asserted anywhere on this page. That is a stated boundary of the assessment, not an omission to be filled later. The firsthand ban-and-refund evidence is labelled as directly observed and is deliberately reported as a split finding — enforcement transparency negative, refund handling positive — with no inference drawn about whether the deactivation was warranted.
CHI 57/100 (Concerning), CVI 81/100 (Strong) and CFS +24 (Customer-Friendly) are settled analytical values and are not presented as a range. Pricing, feature availability, subscription tiers and Face Check market coverage change over time and should be reconfirmed before republication.
Final verdict
CONCERNING — AND SUBSTANTIALLY VALUABLE
Hinge demonstrates that a dating platform can genuinely design around better dating outcomes while simultaneously developing a business model that monetizes the friction, competition, scarcity and attention inside that same marketplace. Its strongest product mechanics — Your Turn Limits, Signals, We Met, Pausing, mutual-preference matching, profile-specific interaction — appear designed to get customers into real relationships. Its strongest revenue mechanics — Standouts, Roses, HingeX, Priority Likes, Boost, Superboost — sell customers better position in the race to get there.
Neither half cancels the other, and this assessment refuses to flatten the case in either direction. "Hinge is a good app" ignores a documented scarcity-and-priority architecture, a disclosed behaviour-sensitive pricing capability, an enforcement process that can sanction a customer without telling them what for, and a safety failure in which Hinge's own curation system elevated a user its safety infrastructure had already received a rape report about. "Hinge is a hostile dating app" ignores a CVI of 81, a badge that cannot be bought, a limit that paying subscribers do not escape, a product that asks whether the date actually happened, and a refund that was handed back more readily than the contract required. The tension between those two descriptions is not a failure of the assessment. It is the finding.
The issue is not whether Hinge should make money.
It is whether monetizing the path to a successful relationship eventually begins to interfere with the obligation to deliver that outcome as efficiently, safely and fairly as possible.