The interface
Buttons, defaults, wording, timers, disclosures, prompts and visual hierarchy.
Concept suite · Product pages 0.1
DigiData deceptive-design products
Do not count dark patterns.
Reconstruct the choice.
Three focused products examine the parts of deceptive design that a screenshot cannot settle: the route a consumer must follow, the variant the system chooses and the moment at which adaptation becomes exploitation.
One question per product. Reasons before scores. Evidence before badges.
Path. Variant. Moment.Three focused tools
P
PARITYCompare the routes into and out of a commitment.
PathM
MIRRORReconstruct who receives a different experience and why.
VariantM
MOMENTTest when adaptation becomes exploitation.
MomentReasons before scoresEvidence before badges
The problem
A screenshot can show a button. It cannot show the system.
Conventional dark-pattern reviews tend to inspect visible interface elements. That remains useful for false urgency, preselection, disguised advertising, hidden costs and other practices that appear directly on the screen.
Deceptive design often operates across a longer chain. The consumer may face a different route when cancelling than when subscribing. A system may serve different interfaces according to profile, history or predicted value. The same design may acquire different significance when the system identifies fatigue, distress, financial pressure or developmental vulnerability and adapts the interaction around it.
The screen is evidence. It is not always the complete object of assessment.
The sequence
What appears before and after the choice, how often it appears and how much friction surrounds each route.
The architecture
Profiling, optimisation, triggering conditions, experimentation, adaptation and measured outcomes.
The products
Path. Variant. Moment.
Each product answers one bounded operational question and produces a reasoned record. None claims to certify legal compliance.
01 · Path
PARITY
Measure the distance between yes and no.
Compare the route into a commitment with the route to refuse, reverse or leave it. Identify which extra burdens protect the consumer and which merely obstruct the choice.
Explore PARITY →02 · Variant
MIRROR
One service. Many users. Show who sees what.
Reconstruct how profiles, predictions and operating conditions alter the interface, sequence, price, offer, disclosure or friction delivered to different users.
Explore MIRROR →03 · Moment
MOMENT
Do not label the person. Test the moment.
Assess whether a consumer's circumstances, the system's knowledge and the surrounding choice architecture combine to create an exploitable vulnerability event.
Explore MOMENT →The shared method
From feature to outcome.
The suite moves beyond pattern spotting without treating every persuasive feature as unlawful. It reconstructs the mechanism and then applies the relevant legal or governance threshold.
Feature
What does the consumer encounter: a prompt, default, timer, ranking, offer, delay or disclosure?
Sequence
Where does it appear, what surrounds it, how often does it recur and which alternatives remain available?
Objective
What is the system optimising: task completion, conversion, spend, retention, data disclosure or time?
Adaptation
Does the system change timing, content, friction or support according to profile, context or feedback?
Outcome
Does the architecture materially impair choice, obstruct a right, exploit susceptibility or create a defined consumer harm?
The decision lifecycle
Five focused questions. One connected evidence chain.
The deceptive-design products complement FAIRTEST and DELTA. They do not duplicate them.
FAIRTEST
Should this experiment or design variant launch?
PARITY
Are the trader-preferred and consumer-protective routes meaningfully comparable?
MIRROR
Which users or conditions receive different routes, and what causes the difference?
MOMENT
Does the system use that difference to exploit a vulnerability event?
DELTA
Does a later change require any earlier assessment to reopen?
FAIRTEST and DELTA appear here as adjacent DigiData products in development. The three pages in this package concern PARITY, MIRROR and MOMENT.
Shared technical architecture
One local-first engine. Three specialist methods.
The planned applications will use one schema-driven browser engine with product-specific question sets, evidence structures and deterministic outcome rules.
ProcessingLocal by default
Assessment answers and evidence remain in the browser unless the organisation deliberately exports them.
LogicInspectable rules
Generative AI will not determine the outcome class. The method should remain reproducible and reviewable.
RecordsJSON and PDF
Users can save, reopen, version and transfer a structured project file and generate a print-ready reasoning record.
VisualsJourney and relationship maps
SVG diagrams will render matched pathways, decision surfaces and person-system-context chains.
The first release will not be a surveillance platform. It will not crawl authenticated journeys, replay identifiable users or infer hidden system logic from a screenshot. Users must supply or test the evidence on which the assessment relies.
Proposed pricing
Three access levels for each product.
A fictional demonstration explains the method. A single-assessment licence supports one bounded decision. An organisation licence supports repeated internal use.
| Product | Public demonstration | Single assessment | Organisation licence |
|---|---|---|---|
| PARITY | Free | £395 + VAT | £7,500 per year + VAT |
| MIRROR | Free | £695 + VAT | £12,500 per year + VAT |
| MOMENT | Free | £495 + VAT | £9,500 per year + VAT |
| Complete suite | Three fictional demonstrations | Purchased separately | £22,500 per year + VAT |
No per-seat pricing at launch. These assessments require legal, design, product, data science, operations and audit teams to work together.
Professional review remains separate. Software access does not include an independent legal opinion, audit or certification.
What the suite is not
A badge for persuasive design.
The tools do not declare that every nudge, recommendation, personalisation or retention measure is unlawful. Nor do they convert complex legal tests into a single compliance score.
- They do not replace the UCPD, DSA, AI Act, sector rules, professional judgement or user research.
- They do not require architecture-level analysis before a team removes an obvious prohibited practice.
- They do not prove intent merely because a system optimises conversion, engagement or retention.
- They do not infer hidden variants or vulnerability from evidence that the organisation has not supplied.
Bright-line practices remain bright-line practices. A false countdown timer or obstructive cancellation route does not become acceptable because an organisation completes a sophisticated assessment.
Regulatory direction
The market is moving from screens to systems.
EU law already prohibits certain manipulative interface practices under the Digital Services Act. The forthcoming Digital Fairness Act is expected to address manipulative interfaces, addictive design, unfair personalisation and the exploitation of vulnerabilities more broadly. Enforcement in the United Kingdom and United States continues to show the operational significance of cancellation friction and obstructive journeys.
European Commission · Digital Services ActOfficial overview, including the ban on dark patterns.
European Commission · Digital Fairness Act directionAddictive design, dark patterns and unfair personalisation.
Ofcom · cancellation-friction enforcementA current example of journey-level obstruction and organisational incentives.
Legal and market context checked on 16 August 2026. The products remain methods in development, not statements of settled law.
Start with the right question
Path, variant or moment?
Choose the product that matches the operational decision. Use the suite only where the evidence chain genuinely requires all three.