Connected TV advertising: a UK ad fraud checklist

What to take away

For connected TV advertising in the UK, fraud often means spoofed app inventory, mislabelled placements or inflated impressions rather than fake clicks.
Independent verification should be agreed before the first impression serves, not after the invoice arrives.
IAB UK's Gold Standard is built around three pillarsreducing ad fraud, improving brand safety and improving ad experience. JICWEBS' Digital Trading Standards Group (DTSG) Good Practice Principles address supply-chain transparency and brand safety. Neither replaces your own campaign checks.
Data processed to detect fraud can be personal data, so it needs a lawful basis under the UK GDPR.

Why connected TV fraud looks different

Connected TV sits awkwardly between broadcast and programmatic. Inventory clears through an exchange, the creative serves inside an app, and the evidence arrives as a report you cannot audit line by line. That gap is where fraud settles.

A single UK campaign can pass through a broadcaster sales house, an agency trading desk and a long tail of smart TV apps. Each layer may apply its own verification rules, and none of them owns the final number.

The advertising codes enforced by the ASA still apply to what you claim about performance. If you publish fraud figures, they need to be defensible.

The four patterns worth checking

Most CTV fraud is not dramatic. It hides in the supply path, in labels that flatter the inventory, and in counting differences between the seller and your measurement vendor.

Four CTV fraud patterns

Fraud pattern

App spoofing: Impressions on premium app
Mislabelled environment: Brand-safe labels mismatch
Impression inflation: Reach outruns audience
Non-household traffic: Viewing unlike domestic audience

How it shows up

App spoofing: App bundle IDs
Mislabelled environment: Supplied app list
Impression inflation: Platform and vendor counts
Non-household traffic: Invalid traffic rates

What to ask for

App spoofing
Mislabelled environment
Impression inflation
Non-household traffic

Fraud pattern	How it shows up	What to ask for
App spoofing	Impressions credited to a premium app you never bought	App bundle IDs, matched against the placement list
Mislabelled environment	Brand-safe labels that do not match where the ad served	The supplied app or channel list from the seller
Impression inflation	Reach and frequency that outrun the platform audience	Platform-side and vendor-side counts for the same dates
Non-household traffic	Viewing patterns unlike any domestic audience	Invalid traffic rates by vendor, before and after the campaign

A pre-campaign checklist

Ask which verification partners the seller accepts and what invalid traffic filtering rate they report.
Agree in writing which vendor counts impressions, and who pays for that measurement.
Request app bundle IDs or channel lists before the first impression serves, then compare them with the delivered report.
Set a maximum invalid traffic threshold and a named contact for credits or make-goods.
Define a completed view, then test the definition against CTV ad viewability standards UK.
Diary a mid-flight review at the halfway point, while there is still budget to protect.

Illustrative arithmetic helps. A campaign buying 1.5 million impressions at a £22 CPM spends £33,000. If 8 per cent of those impressions fail verification, about £2,640 of that bought nothing, and the clause you agreed decides whether it comes back.

Pre-campaign CTV fraud checklist

Ask which verification partners seller accepts
Agree in writing which vendor counts impressions
Request app bundle IDs before first impression
Set maximum invalid traffic threshold and contact
Define completed view and test definition
Diary mid-flight review at halfway point

How to check a delivered report

Compare total impressions with the platform-side count. A gap beyond a few per cent deserves an explanation.
Look for app or channel names you never approved.
Check that reach and frequency are plausible against the audience size you were quoted.
Ask for the invalid traffic rate by vendor, then compare it with the pre-campaign estimate.

What the data rules require

Detecting ad fraud often means processing device identifiers, IP addresses or hashed email addresses. Under UK law, that is personal data.

You need a lawful basis for it. Legitimate interests usually fits fraud prevention, but it needs a documented balancing test, not an assumption.

The ICO's guidance on tackling scams and fraud explains that data protection does not prevent proportionate fraud prevention. The ICO's Guide to the UK GDPR provides guidance on documenting your lawful basis and handling subject requests.

The UK GDPR does not set one fixed retention period for fraud-detection identifiers. Under the storage limitation principle in Article 5(1)(e), keep them no longer than necessary for the stated purpose, document your rationale and delete or anonymise them when they are no longer needed. The UK GDPR principles cover retention and accuracy too.

Working with platforms and agencies

Ask for a named person who owns verification rather than a shared inbox. When numbers diverge, you want one contact accountable for the answer.

If you buy through a demand-side platform (DSP), check whether invalid traffic filtering is on by default or only when you ask. Examples of DSPs used for digital and CTV buying include The Trade Desk, Google's Display & Video 360 (DV360) and Amazon DSP; ask the platform which CTV inventory and verification controls are available for your campaign. Some settings sit behind a support ticket.

Smaller advertisers often get better control by buying a defined set of apps or channels directly. Our guide to the best connected TV DSPs for UK small businesses covers those trade-offs.

Common questions

Does CTV ad fraud affect small UK advertisers?

Yes, though usually in smaller amounts. The common patterns are misrepresented apps and inflated impression counts, and they scale with whatever budget you spend.

Is third-party verification always necessary?

Not always, but you should know what you are relying on instead. Platform self-reporting is a starting point, and independent verification matters more as spend grows.

What should a make-good clause include?

It should define invalid traffic, name the measurement source, set a threshold for remediation and give a timescale. For example: “If [named measurement provider] reports invalid traffic above [agreed threshold] in the campaign, the seller will, at the buyer's election, replace the affected impressions with equivalent valid impressions or credit their value within [agreed period].” Fill in the brackets in the insertion order; vague wording is the main reason credits go unclaimed.

Who pays for verification in a UK CTV deal?

Usually the buyer, though sellers sometimes absorb the cost to win the deal. Pin it down in the insertion order so nobody argues later.