The Verification Files
Countermeasures

From problem to proof

Each case file, the problems it highlights and the countermeasures needed to combat them.

Countermeasures · TVF-008Mobile data

27 Things a Mobile Number Reveals: Signals for Fraud, KYC and Leads

Problems identified
  • A phone number is stored as a contact field and checked only for format.
  • Where a network lookup is run, it is used for live or dead and nothing else.
  • Out of network numbers are accepted and dialled indefinitely.
  • A passed check is taken as proof that the number is valid.
Countermeasures
  1. Treat the number as evidence
  2. For lead generation: live status and subscriber name
  3. Before KYC: SIM swap, exact age, subscriber name
  4. For fraud: SIM swap, out of network, subscriber name
  5. Run the cheapest check first
  6. Recheck out of network numbers, then reject
  7. Step up after a recent SIM or device swap
  8. Recheck stored numbers before using them
  9. Cover the gaps in network KYC
  10. Read the batch
Countermeasures · TVF-007Identity verification

When Identity Verification Becomes the Risk: Five Ways Checks Fail

Problems identified
  • Checks the law requires before access are run after the point of entry.
  • Old breached records pass as established ones.
  • Data is collected and kept with no use for it.
  • A customer is verified once and never again.
  • AI agents make purchases for people nobody has verified.
Countermeasures
  1. Complete legally required checks before entry
  2. Sequence the rest on cost and friction
  3. Test the flow with everyone who has a stake
  4. Let the law set the floor and the board set the appetite
  5. Judge breach exposure by pattern, with provenance
  6. Keep data only if you use it
  7. Re-check on events and signals
  8. Make the cost case as insurance
  9. Treat an agent as someone acting for a person
  10. Set limits before giving an agent payment details
Countermeasures · TVF-006KYC

How Fake Leads Become KYC Data: From Breach to Identity Check

Problems identified
  • Bots submit breached records to lead forms and are paid per lead.
  • The details are real, so validation passes them.
  • Collectors resell the records into KYC without the provenance behind them.
  • A bot’s metadata trail is accepted as proof of a visit.
  • The affiliate and KYC industries share nothing.
Countermeasures
  1. Look for the pattern, not the record
  2. Know the tells in a batch
  3. Run fraud detection on the form
  4. Reject the cost argument at volume
  5. Investigate any file over 70% breached
  6. Pass provenance on with every record
  7. Require the same at the KYC end
  8. Treat volume as the first warning
  9. Check new phone numbers with the mobile network
  10. Cross-check traffic sources against known fraud
Countermeasures · TVF-005Affiliate fraud

How Affiliate Fraud Gets Paid: Ad Hijacking, Cloaking and Sub-Networks

Problems identified
  • Commission is paid on sales the brand was already going to make.
  • Lead payments go to bots submitting stolen data.
  • Sub-networks hide who is sending the traffic.
  • The network sees masked metadata, not the real submission.
  • Evidence of fraud is held privately and never shared.
Countermeasures
  1. Ask for direct publishers only
  2. Have every sub-network labelled
  3. Turn off automatic approval
  4. Never send affiliate traffic to your own lead form
  5. Capture the real metadata
  6. Check timing, details and sources
  7. Use monitoring for ad hijacking
  8. Withhold payment on suspicion
  9. Make fraud controls a condition of buying
  10. Push for a central evidence register
Countermeasures · TVF-004B2B lead gen

Are Content Syndication Leads Real? The Content Swindlefication

Problems identified
  • Real contacts are sold as people who downloaded your content.
  • The supplier controls the landing page, so the buyer cannot see what happened on it.
  • Buyers are told not to call, so nobody asks the lead.
  • QA, certification and a perfect profile match stand in for proof.
  • The lead is used as a consent record it cannot support.
Countermeasures
  1. Ask three questions before you buy
  2. Control the page or track it
  3. Check call recordings for splicing
  4. Ask for delivery in real time
  5. Without proof, call every lead at once
  6. Reject any lead that denies the download
  7. Ignore “do not call, nurture”
  8. Do not rely on the lead as consent
  9. Put the budget elsewhere
Countermeasures · TVF-003KYC

Where Commercial KYC Data Comes From, and Why a Match Is Not Proof

Problems identified
  • A match is treated as proof, whatever the reference data is.
  • Reference files hold old and breached records with no traceable origin.
  • The same record sits in several databases and is counted as independent sources.
  • Every party is paid on a match, so nobody is paid to question one.
Countermeasures
  1. Ask how the data was collected
  2. Get the original sign-up date
  3. Treat very large volumes as a warning
  4. Weight the match by its provenance
  5. Test whether sources are independent
  6. Test the data by originating site
  7. Require provenance with every record
  8. Audit the file you already hold
  9. Pay on provenance, not on a match
  10. Check what any persistent key is built on
Countermeasures · TVF-002Identity

Valid Data Is Not a Real Person: The Human Data Illusion

Problems identified
  • Records are counted as people before anything has been proved.
  • Valid details are submitted by someone other than their owner.
  • Checks test the fields in a record, not who sent it or how.
  • Bad data is defended by the people who paid for it.
Countermeasures
  1. Start from unproven
  2. Validate every field
  3. Read the submission metadata
  4. Test the batch as a whole
  5. Cross-check against an independent source
  6. Capture the submission yourself
  7. Pay only for records that pass
  8. Decide from the top