What 2,000 Australians said dark patterns cost them

The harm was not confined to money

The most useful feature of CPRC’s report is its breadth. Participants did not describe one neat outcome called “dark-pattern harm”. They reported spending more than intended, creating unwanted accounts, sharing excess personal information, feeling annoyed or manipulated and abandoning services they no longer trusted.

That matters because implementation work can become too narrow. A pricing team may watch the final total but miss the pressure that produced the selection. A privacy team may count consent events without seeing the account gate that preceded them. A lifecycle team may optimise retention while support absorbs the frustration created by the exit path.

Ten patterns made the problem concrete

The report tested consumer responses to ten recognisable forms of design, including hidden costs, disguised ads, trick questions, scarcity cues, activity notifications, confirmshaming, forced continuity, false hierarchy, nagging and data grabs.

Those labels remain useful for teaching. They are not the enacted language of section 28B. The portal’s practice library therefore connects each mechanism to a neutral design comparison and keeps the legal analysis in its own layer.

Why self-reported evidence still matters

A survey cannot replay every respondent’s screen. It can reveal patterns in what people remember, how they describe the experience and which consequences they associate with it. That is valuable for choosing research questions and identifying groups or journeys that deserve closer study.

It should be paired with behavioural and product evidence. If a person says cancellation was impossible, preserve the route they encountered, the device, account state, retention prompts and final backend status. If someone reports spending more than planned, reconstruct how the initial price, extras, transaction charges and final total were presented.

The pricing guide and cancellation guide provide the enacted provision-level context for those journeys.

The study and section 28B answer different questions

CPRC published the report while advocating for broader Australian consumer protection. The later 2026 Act now supplies enacted wording, but the survey does not retroactively become an enforcement decision.

For conduct on or after commencement, section 28B requires its complete statutory test. The design must be connected with a consumer supply, possible supply or offer; the conduct limb and detriment limb must be assessed on evidence. A survey category can help a team spot the mechanism. It cannot replace those steps.

The practical lesson

Measure the customer cost as well as the conversion. Complaints, reversals, unwanted accounts, support contacts, cancellation retries and later mistrust can reveal consequences that a single funnel metric hides. Then connect those observations to the actual journey and the law in force at the relevant time.

Source check on 14 September 2026

The CPRC report page still presents the 2,000-person survey and the percentages shown here as findings from June 2022. This review checks their attribution, not a new Australian prevalence estimate. The reported consequences can overlap, so adding their percentages would double-count respondents and would not produce a total harm rate.

Questions for an Australian journey review

1. Which financial, privacy, emotional or trust consequence could arise from the complete journey rather than from one isolated control?
2. How will user research distinguish annoyance, misunderstanding, pressure, unintended action and actual detriment?
3. Can the team connect reported harm to preserved interface states, timing and configuration without treating the research label as a legal conclusion?