← all systems · see it in The System
Romance scam
A confidence trick in which a false romantic relationship is built, usually online, to get the targeted person to send money or commit fraud on the offender's behalf (src:wp:romance-scam); 'the deployment of fake romantic relationships primarily for material ends' (src:lazarus-2023-systematic). Walked from the victim-support and researcher's side: the stages researchers describe, who is targeted and what is known about risk, the measured harm (financial and emotional, the double hit, under-reporting), the organised forms, the platforms' and banks' part, what prevention and support are measured to do, and the debates. No scripts or steps for doing it.
Properties published scienceour proposal
Sourced fields, and our own readings.
| born at | the 16th-century 'Spanish prisoner' scams (romantic deception for gain); the online form 'became apparent in about 2008' (Whitty & Buchanan 2012); IC3 counted it among the top 5 internet scams in 2011 |
| born precision | centuries (kind); about a decade (online form) |
| boundary | Inside: the false relationship and its run as researchers describe it (preparation and persona, target selection, first contact, the shift to private channels, grooming, the requests (the sting), the payout, the squeeze, the aftermath), between the offender side and the targeted person, including the targeted person's appraisal, doubts, checks and exit. Outside: the platforms it runs on, the payment rails, banks, police and support services that answer it, the family and friends round the targeted person, and the criminal economies (fraud factories, laundering) it feeds. |
| nested level | Runs on people (Human L5-L7: the need for companionship, trust, love, decisions) and on the internet's services (dating sites and apps, social media, messaging); a kind of confidence trick and of advance-fee fraud, and one of social engineering's forms; feeds and is fed by organised crime. |
| scale | social |
Measures (32) published science
| Grooming phase duration | 6-8 months (typical) time |
| Time to a declaration of love | within a week to a couple of weeks of first contact time |
| Bond phase duration in pig-butchering scams | 3 to 11 months (26 interviewed victims) time |
| Squeeze phase in pig-butchering scams | 1-7 months of continued large investments time |
| Longest reported scam duration | up to 8 years (extreme cases) time |
| Fastest large payment in one study | about £70,000 over a weekend after about a month GBP, time |
| FTC romance-scam reports, 2023 | 64,003 reports; about US$1.14 billion; median US$2,000 reports, USD |
| FTC romance-fraud reports, 2021 | over 56,000 victims; over US$547 million; +80% on 2020 victims, USD |
| IC3 confidence/romance complaints, 2025 | 23,159 complaints; US$929,287,469 (2024: 17,910; US$672,009,052) complaints, USD |
| IC3 confidence/romance victims and losses, 2012-2022 | 2012: 4,476, $55,991,601; 2017: 15,372, $211,382,989; 2019: 19,473, $475,014,032; 2021: 24,299, $956,039,739; 2022: 19,021, $735,882,192 victims, USD |
| UK romance-scam losses, 2023 | £93m, up by a third on 2022 (UK Finance) GBP |
| Australian romance-fraud losses, 2022 | AUD$210 million (ACCC) |
| Netherlands registered cases, 2023 | 494 cases; 264 with financial loss; €7,644,870; average nearly €29,000 per victim cases, EUR |
| Estimated victims in Great Britain (survey) | almost 230,000 people |
| Self-reported victimisation, general population | general population: 6.17% once and 2.81% more than once (one study); 0.65% victims and 2.28% knowing a victim (another); summed by the review as 1-3%. Self-selected samples: 63% of one Facebook-user sample (29% more not sure); 88.55% of users of victim-awareness websites share |
| Share of scams reported (global) | 7% (Global State of Scams Report 2022) share |
| Share of elderly fraud reported (FTC) | 2% - 6.7% share |
| Share of fraud ever reported | less than one third (estimates) share |
| Victim panel (Whitty 2018) | 11,780 UK panel members; 200 romance-scam victims; of victims 63% middle-aged, 21% young, 16% older; 60% women; model Nagelkerke R2 = 0.354 people, share |
| Human accuracy at sorting fake from genuine profiles | 13.51 of 20 profiles correct on average (SD 2.63): 6.74 of 10 fake, 6.77 of 10 genuine; above the 10 of 20 a coin would score |
| Automatic scam-profile detection accuracy (research system) | 97% accuracy |
| Scam-profile demographics (one dating site) | about 60% male, same as real profiles; scam female profiles about 30, male about 50; widowed over-represented; military and engineering occupations share, age |
| Pig-butchering losses per victim (interviews) | US$600 to US$185,000 USD |
| Pig-butchering victims approached by recovery scams | 11 of 26 people |
| Pig-butchering victims finding the term offensive | 69.2% share |
| Pig-butchering complaints with wallet addresses (California DFPI) | 9.2% of 273 (96.7% actionable) share |
| Dating-app users exposed to pig butchering (US) | 12% in October 2023, up from 5% in 2018 share |
| Scammer location (one study) | about 50% from African countries, 16% from Asian and English-speaking countries share |
| Firefly femme fatale hunting success (comparison) | 9 of 92 observations (9.8%) share |
| Convicted offenders' case files (Nigeria, EFCC) | 50 cases: victims 56% US; offenders posed as Caucasian American men 46%, military 12%, Caucasian European men 10%; targeted women 70%, men 14%; 74% university students; Facebook 46%; iPhone 58% share |
| Military narrative in Australian romance-fraud reports | 3,259 ACCC reports (Jul 2018 - Jul 2019); starting on a social networking site OR 1.843 for a military narrative; military narrative not significant for financial loss reports, odds ratio |
| Romance-scam victims in a UK online panel (Whitty 2018) | 200 of 11,780 (about 1.7%, derived; lifetime, self-reported, online panel) people |
Open contentions (6) open contention
Where sources or readings disagree. Both stay on the map.
- C-3964 Is 'pig butchering' a romance scam or an investment fraud?
booked: The map keeps the competing readings; none is ruled.
the other reading: investment fraud (the ACCC records romance baiting as investment fraud); its own category, a deliberate amalgamation of the two (Cross 2023) - C-3974 Is the romance scam an internet-age crime or centuries old?
booked: The map keeps the competing readings; none is ruled.
the other reading: rung 12: the online dating romance scam, its name and its research field belong to the internet era (the review literature starts at 2000); the online romance scam 'became apparent in about 2008' (Whitty & Buchanan 2012): when it became visible to researchers and police, not necessarily when it began - C-3975 Why do romance scams work: persuasion tricks, something like abuse, or wider social causes?
booked: The map keeps the competing readings; none is ruled.
the other reading: psychological maltreatment and economic abuse, as in domestic violence (Cross et al. 2018); a sociological and regional account: the model reproduces the older 419 fraud thesis and leaves out contextual drivers, offender worldviews and cyber spiritualism (Lazarus et al. 2023); an affective economy where emotion and trust become currency (Feng 2026); offenders' own accounts: neutralisation (reparative justice for colonial exploitation, Lazarus et al. 2025) and the fraud as a game (Abubakari & Oseh-Ovarah 2025) - C-3976 Is anyone who falls for a romance scam partly responsible, or does saying so only add to the harm?
booked: The map keeps the competing readings; none is ruled.
the other reading: an ethics reading: the more counterevidence a victim disregards, the more responsible they become for later losses (Sorell and Whitty 2019, via Bilz 2023); detection is hard for anyone: participants sorted only about two thirds of profiles right, which the author reads against the view that victims are stupid (Whitty 2019) - C-3977 Who is most at risk of a romance scam?
booked: The map keeps the competing readings; none is ruled.
the other reading: middle-aged, well-educated women, high urgency, sensation seeking and addictive disposition (Whitty 2018, UK panel); romantic idealisation predicts; loneliness, neuroticism and sensation seeking do not (Buchanan and Whitty, via Coluccia 2020); older adults: fewer cybercrime reports but more financial loss (Havers 2024, all cybercrime); targeted for assets (encyclopedia); younger adults: the COVID rise was sharpest at 20-39 (Buil-Gil 2021); romance baiting reaches 25-34 (ACCC via Cross 2023); vulnerability is complex and changes over time (Lazarus 2023); the source disagrees with itself: Whitty 2018's abstract says victims are 'more trustworthy'; its Table 1 and Table 2 show them lower (17.37 vs 19.39; beta -0.512, p = 0.015); its results text lists H8 as not significant; Coluccia 2020's abstract and discussion list neuroticism as a victim risk; its body ties neuroticism to presenting deceptive profiles and to distress after discovery, and reports no neuroticism effect on victimisation in study [18] - C-3978 Are romance scams rising or falling?
booked: The map keeps the competing readings; none is ruled.
the other reading: IC3 showed a fall in 2022 (19,021 victims, $736M against 24,299 and $956M in 2021) while web searches and news rose (Herrera 2024); the fall may be categorisation: romance baiting counted as investment fraud (Cross 2023)
Cross-scale patterns seen here (20) our proposal
Shapes that recur when the same kind of problem is solved at different scales, and how each shows in Romance scam. Each links to every other system that shows it.
- red-queen (structural)
Each payment rail disrupted (remittance agencies) is replaced (money mules, then cryptocurrency); prevention messaging aimed at emergencies is met by investment-framed asks; detection improves and profiles are recycled. Losses keep rising (IC3 2012 $56M to 2025 $929M) while defences improve: both si - same role, different builds (structural)
The job, in plain words: a false courtship signal draws a partner close so the sender can take from them. Built in people from profiles and months of messages; built in Photuris fireflies (aggressive mimicry, the biologists' term) from mimicked female flash replies that draw males of other species t - templating (structural)
One module reused across victims and kept working: recycled profiles with minor edits, stock personas (the soldier, the engineer, the widower), stock crisis stories, playbooks; and the older 419 advance-fee template reused inside the romance scam (Lazarus: the model 'has all the hallmarks of the Nig - who opens the door (structural)
The targeted person opens their own door: they send the money themselves, often by methods that cannot be recalled; police told one victim it was not theft because she transferred it willingly. The port (the person's trust and their access to their own money) is installed by the receiver and opened - borrowed-then-held (partial)
Moving the money was first borrowed from outside services (Western Union, MoneyGram), then partly held: offenders' own mule networks and their own fake investment apps and platforms. - lightning (partial)
The revelation: a held false belief released at once, after which the person is not the same (devastation 'akin to experience the death of someone close'); but often it is not sudden: many cannot believe it even with strong evidence, and some go back. - nested ladder (partial)
Crime scripts nest by abstraction: metascript (fraud), protoscript (dating fraud), script (online romance scam), track (one modus operandi); each lower rung a more concrete form of the one above. - same shape, opposite sign (partial)
One process, a lure under false pretences, pointed both ways: the people who run pig-butchering scams in fraud factories were often lured abroad under false pretences and forced to work; and the recovery scam turns the helper's role into the exploiter's. - The carry (partial)
The wait: the request comes only when grooming has filled the relationship to the point the scammer judges ready ('until they feel that the victim is ready to part with their money'), typically 6-8 months; and overflow passed on: when a victim can pay no more, they are passed on (sucker lists, recov - trade one property for another (partial)
Fraud prevention against online inclusivity and privacy: stricter identity verification would remove fake profiles and harm users who need anonymity (people exploring gender identity, people in sensitive situations). - Umwelt: each system carves its own slice of one shared medium (partial)
The scam lives in the slice of the shared medium the text channel carries: no face, no voice or a disguised one, several operators as one name. Each side's channel set differs: the offender reads the target's public traces and disclosures; the target sees only the crafted persona. The rotation 'woul - barrier-gates-complexity (surface)
The Shift to private channels seals the relationship away from platform monitoring and reporting, and grooming builds behind that seal; no source shows the without-the-seal case directly. - causal ceiling (surface)
What the targeted person can know is bounded by the text channel: people sort fake from genuine profiles at about two thirds (13.51 of 20), better than chance but far from sure. A practical limit, not a structural one (the dark twin). - holding / entrainment (surface)
The relationship sets the targeted person's daily clock: good-morning and good-night calls, hours on chat each night, demands for constant availability. - namespace completeness (the map predicts its gaps) (surface)
The crime script's scenes-by-actions grid implies tracks not yet documented (LGBTQIA+ targeting and platforms; offender-side decisions), named as gaps by the authors. - partner-becomes-part (surface)
Not a gain/gain fusion: the targeted person is drawn into the offending system as a money mule or co-offender. Booked as the reversed case. - runaway from an unstable balance (surface)
After a first small payment, each payment feeds the next (foot-in-the-door, near win); no balance point is shown. - same toolkit, independent builds (surface)
West African romance fraud and Chinese/Southeast Asian Sha Zhu Pan build the same structure (persona, grooming, money ask) from the same parts (platforms, messaging, payment rails); independence of the two lines is not shown. - symmetry breaking (name the cut) (surface)
Self-disclosure runs one way: the victim discloses far more than the scammer; scammers ask many questions and share little. A symmetric before (reciprocal disclosure in real relationships) is not shown in the sources read. - the whole sets the local reference (surface)
The statistics' category scheme fixes what each local report counts as: romance baiting is read as investment fraud in Australia, so the romance figures read low.
Valence: what it trades, and with whom (15) published science
Each exchange with a neighbour, read from both sides: who gains, who pays, and what carries it.
- Human (Homo sapiens) via sys:sensory-environment · Romance scam gain
- Family (the human family as a system) via sys:sensory-environment · Romance scam loss
- Police via sys:human-society · Romance scam loss
- Mental health via sys:sensory-environment · Romance scam even
- Social media via sys:internet · Romance scam gain
- Threat actor via sys:internet · Romance scam gain
- Fraud factory via sys:organized-crime · Romance scam gain
- Attachment (the infant-caregiver bond) via sys:sensory-environment · Romance scam gain
- Workers forced to run scams (in fraud factories) via sys:organized-crime · Romance scam gain
- Online dating service via sys:internet · Romance scam gain
- Money transfer operator (Western Union, MoneyGram) via sys:payment-system · Romance scam gain
- Peer support groups via sys:human-society · Romance scam loss
- Business email compromise via sys:human-society · Romance scam gain
- Cryptocurrency via sys:payment-system · Romance scam gain
- Banks via sys:payment-system · Romance scam mixed
Partners and parts published science
Built from / builds (6)
- Human (Homo sapiens)
- Threat actor
- The fabricated persona and profile
- Crisis narratives and other requests (the sting)
- Workers forced to run scams (in fraud factories)
- Grooming (the relationship built before the asks)
Sits inside / holds (5)
expressed by (24)
- Generation/Synthesis
- Recombination/Composition
- Scaffolding/Enablement (Meta-Tooling)
- Templating / Standardization
- Niche Construction (alter environment)
- Amplification/Scale-Up (Manufacture & Distribution)
- Sensing/Scanning
- Search/Foraging
- Active Experimentation / Probing (Optimal Experimental Design)
- Multi-Scale Data Fusion / Triangulation
- Trail-Marking / Externalized Memory (Stigmergy)
- Optimal Stopping & Patch-Switching
- Social Learning / Public Information Use
- Evade Detection
- Speed
- Precision Attack
- Subversion & Co-optation
- Prepositioning & Latency
- Leverage Host/Amplification
- Divide & Isolate
- Validation of Identity/Intention
- Deception
- Resilience & Repair
- Redundancy & Diversity
floor filled by (19)
- Human (Homo sapiens)
- Internet
- Emotion (feelings)
- Payment system
- Electronic mail
- Police
- Social media
- Threat actor
- Persuasion
- Money mule
- Fraud factory
- Attachment (the infant-caregiver bond)
- Telephone
- The fabricated persona and profile
- Crisis narratives and other requests (the sting)
- Online dating service
- Peer support groups
- Grooming (the relationship built before the asks)
- Instant messaging
Kinds (6)
- Pig-butchering scam (romance baiting, cryptorom, Sha Zhu Pan)
- Military impersonation romance scam
- Pro-dater scam
- Marriage-inheritance (419) romance scam
- eWhoring
- Romance fraud without the internet (lonely-hearts ads, letters)
Notes published science
- Fireflies run a romance scam too: Photuris females answer the courtship flashes of other species' males, then try to eat the ones that come close. It mostly fails: in one study only 9 of 92 watched hunts ended in a catch, partly because the males hover nearby trading signals for minutes before committing, which is very like a person running a reverse image search before they trust a profile (src:maquitico-2024-femme-fatale; src:bilz-2023-tainted-love).
- Many victims say losing the relationship hurt more than losing the money, and some describe it like the death of a loved one. Researchers call it the 'double hit', and a review of 21 studies found the psychological harm tracks the betrayal more than the size of the loss (src:whitty-2018-do-you-love-me; src:bai-2026-fraud-mental-health).
- When police and banks got good at spotting Western Union transfers, scammers moved to money mules, then to cryptocurrency, and then stopped asking for help at all: in 'romance baiting' the partner offers an investment instead of a sob story, so in Australia those losses are counted as investment fraud and romance fraud looks smaller than it is (src:cross-2023-romance-baiting).
- Spotting a fake dating profile is genuinely hard: shown 10 fake and 10 real profiles, 261 people got about 13.5 of 20 right on average, better than a coin but far from safe. The author reads it against the popular view that only foolish people fall for these scams (src:whitty-2019-who-can-spot).
Layers
The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.
the boardroom (shared floors)
- L4: The false relationship (grooming): where the offender side's persona, tailoring and pressure meet the target's needs, disclosures and hopes; a hyperpersonal online relationship of daily contact, declarations of love, a shared future, secrecy and isolation
- L5: The targeted person's appraisal: trust and love for the persona; ambivalence, denial or rationalisation of doubts; or suspicion and active checking. Traits measured in victims after the fact (urgency, sensation seeking, an addictive-disposition score in one UK panel; romantic idealisation in another) are correlates of having been a victim, not shown causes
- L6: The sting: requests for money (foot-in-the-door, crisis with deadline, door-in-the-face, an investment), reinforced by fake third parties and promises of repayment; the decision to pay, refuse or verify
- L7: Outcomes: money to the offenders (and sometimes the target made a mule); the squeeze (sextortion, extortion); the double hit; revelation, shame, silence, under-reporting; revictimisation (recovery scams, sucker lists, the scammer's return); or exit, report, support and recovery
the romance scam as victimologists and criminologists describe it: online spaces where people look for love, up to money lost, harm done and help found (reader's view)
- L1: The online spaces where people look for partners and talk: dating sites and apps, social media, email, messaging apps, forums and games; and the public traces people leave there
- L2: The persona and the first contact (the Setup, the Hunt and the Hook): a fabricated profile with stolen or AI-generated photos (soldiers, engineers, the widowed), targets chosen by wealth, nationality or signs of vulnerability, or contacted en masse
- L3: The Shift: the talk moved off the platform to email, instant messaging and phone, away from monitoring and reporting; daily routines of contact; the face kept out of sight (no meeting, one-way webcam)
- L4: The false relationship (grooming): love declared within weeks, intense daily contact, backstory and shared future, mirroring of the target's disclosures, love bombing, secrecy and isolation; often 6-8 months
- L5: The targeted person's appraisal: love and trust in the persona, idealisation, hope, urgency, the near-win; doubts denied or rationalised, or suspicion and checks
- L6: The sting: requests for money (small to large, crisis with deadline, decreasing amounts, or an investment), third parties who confirm the story, and the decision to pay, refuse or verify
- L7: Outcomes back on the world: money moved through banks, remittance, gift cards and cryptocurrency to the offenders (and the target sometimes made a mule); the double hit of money and relationship lost; shame, under-reporting and revictimisation; or exit, report, support and recovery
way in: Emotion and need: the targeted person's own situation (inner channel)
- L1: An unmet need for companionship, a wish for economic security, a lack of purpose; divorce or widowhood; loneliness (rising in lockdown)
- L2: Felt as longing and hope (Human emotion channel)
- L3: Romantic beliefs and idealisation of the partner
way in: Hearing: calls and voice
- L1: Telephone and VoIP calls; good-morning and good-night calls
- L2: Heard (Human hearing channel)
- L3: A voice taken as the persona's; voices can be disguised with apps or modulation software
way in: Images and identities taken for the persona
- L1: Photographs of models, soldiers, ordinary social media users and earlier victims; AI-generated images
- L2: Taken and reused; profiles recycled with minor edits
- L3: A persona with gender and backstory chosen to fit the target sought (a widowed man about 50, a woman about 30; military, engineering, medicine)
way in: Platform detection (an outside influence that removes personas before contact)
- L1: Profile data: demographics, descriptions, images; behaviour
- L2: Spam-style filters fail; machine-learning classifiers on profile features (one research system 97% accurate)
- L3: Profiles flagged or removed before they reach targets (the system's stated aim: early detection)
way in: Sight: profile photos, messages, poems, documents on a screen
- L1: Profile photographs, daily messages, love poems (often plagiarised), documents and photos sent
- L2: Read by eye (Human sight channel)
- L3: The persona read as a real person; the scammer's own face is never seen (no meeting, one-way webcam)
way in: Social conditions: the pandemic, online dating's growth
- L1: COVID-19 lockdowns: romance fraud reports rose after April 2020 above the pre-COVID trend
- L2: More internet use among older adults, more loneliness among young adults
- L3: More people meeting partners online: about 23% of UK internet users had met a romantic partner online (2007 figure cited)
way in: Targets' profiles and public traces (offender side)
- L1: Dating profiles, social media accounts and other online sources about people
- L2: Target selection: targeted (wealth, Western nationality, divorced or widowed), scattergun (mass messages, mass liking), sucker lists of past victims, or a past victim again
- L3: Reconnaissance: what is gathered is used to tailor the persona and the approach
way in: The group, its training and playbooks (organised forms)
- L1: The organisation: lone scammers, small teams, large networks; fraud factories where many workers are trafficked and forced
- L2: Induction and training (targets, conversation, digital tools, financial requests); playbooks with scenarios
- L3: Persona, backstory and false documents prepared before contact (the Setup)
way in: The target's doubts and refusals (read by the offender side)
- L1: The target questions a request or the story, or refuses to pay
- L2: Noticed by the offender side
- L3: Answered by returning to grooming, or by guilt, threats to end the relationship, silent treatment, threatened self-harm; amounts lowered (door-in-the-face)
way in: The target's messages and disclosures (read by the offender side)
- L1: The target's messages: self-disclosure, often far more than the persona gives back
- L2: Read by whoever is on shift: several scammers can answer as one persona at any hour
- L3: Used to emphasise similarities and personalise the persona (a feedback arrow in the script)
way in: The target's own checks
- L1: Search engines, reverse image search, scam-awareness sites and the persona's other online traces (or their absence)
- L2: An internet search of the persona, the commonest response to a suspicious profile
- L3: Image reuse, identity mismatches and warnings found; even no hits can raise suspicion (no digital footprint)
- L4: Active, iterative detection under emotional, social and relational pressure
way in: Thought: the story as it is taken in (inner channel)
- L1: The narrative: a shared future, a backstory, a crisis or an investment
- L2: Taken in as part of a relationship believed to be real (victims describe the grooming stage as therapeutic)
- L3: Each stage perceived as a near win, like gambling
way in: Warnings from outside: family, friends, banks, police, platforms
- L1: Warnings from family and friends, banks (which can lock accounts), police early-intervention, platform warning messages
- L2: Contextual warnings shown when money is asked for raised awareness and lowered risky behaviour
- L3: Often met with denial, anger or secrecy; police intervention most promising when the victim already doubts
- L4: Weighed against the relationship: counterevidence disregarded in some accounts, and the persona works to discredit outsiders (secrecy)
way out: Disclosure: personal details, documents and intimate images handed over
- L1: Material later used for identity fraud, sucker lists or sextortion (the squeeze)
- L2: Sent over private channels
- L4: Intimate self-disclosure; documents; intimate photos in some cases
way out: Help-seeking and recovery
- L1: Social ties rebuilt, or not; distress can persist beyond financial resolution
- L4: Shares the story; gains solidarity, acknowledgement and an understanding of how it happened
- L7: Seeks help: peer support groups, counselling, mental health care (most in one case study did not)
way out: Payment: money sent
- L1: Money leaves the target's accounts, savings, loans or second mortgages for the offenders
- L4: Bank transfer, Western Union or MoneyGram, gift-card codes, cryptocurrency, or a fake investment app
- L6: Persuaded to send money
way out: Proceeds to the offender side
- L1: Funds the groups and fraud factories that run more scams
- L4: Laundered through mule accounts and crypto platforms
- L7: Money siphoned off overseas through mules, remittance or cryptocurrency
way out: Reporting: police, bank, platform, fraud hotlines
- L1: A record in official statistics, or nothing; police often say little can be done when offenders are overseas
- L4: Report made (IC3, FTC, Action Fraud, Scamwatch, local police); fraud reported less than any other crime type, under one third
- L7: Decides to report, often after delay; shame and fear of judgement hold many back
way out: The target made a money mule
- L1: Others' money laundered through the target; the target may be treated as a co-offender
- L4: Forwards money to other accounts, empties accounts into cash or property, reships goods
- L6: Asked to forward money or packages for the persona
way out: Withdrawal from family and friends
- L1: Fewer people round the target to notice or warn; networks hard to rebuild after
- L2: Time and resources turned to the persona; contact with loved ones reduced
- L4: The relationship kept secret at the persona's request