Research written to be cited.
Methodical write-ups on experimentation and CRO, with every claim traced to a published source. Each paper ships with full references and a PDF.
The Conversion Cost of Deception: Why Dark Patterns Fail as a CRO Strategy
Dark patterns are interface choices that steer users toward decisions they would not otherwise make. They are widely treated as a conversion tactic. We argue they are a poor optimization strategy: they inflate the immediate metric a CRO program is judged on while decoupling that metric from the profitable, retained customers it is meant to proxy for. Drawing on the empirical literature, we map dark patterns onto the conversion funnel, explain why they win short-horizon A/B tests, and describe the measurement trap that lets a deceptive variant report a win while the business absorbs an unmeasured loss. We review the regulatory turn against dark patterns and propose an honest optimization alternative that captures the legitimate psychology without the deception.
Bhaskar Roy Sarkar · August 3, 2026 · 7 references · 5 min read
Sarkar, B. R. (2026). The Conversion Cost of Deception: Why Dark Patterns Fail as a CRO Strategy. cazyweb Research. https://cazyweb.com/research/dark-patterns-in-cro
Papers carry Highwire citation metadata, so reference managers pick them up automatically.
Augmentation over Automation: People, Process, and Infrastructure as Preconditions for Organizational AI
Why AI succeeds as leverage on human specialists, not as their replacement: the field-experiment evidence, the productivity J-curve, and the capability stack organizations must build bottom-up.
13 references · 8 min read
July 17, 2026AI SystemsTool-Using LLM Agents in Production Business Workflows: A Practitioner Review
A practitioner review of tool-using LLM agents in production: constrained tool design, human-in-the-loop gates, failure modes and evaluation, grounded in a live agency platform.
15 references · 10 min read
July 17, 2026AI SystemsMulti-Agent LLM Systems: Orchestration Patterns, Failure Modes, and When More Agents Help
A practitioner review of multi-agent LLM orchestration - orchestrator-workers, pipelines, debate - with a failure-mode taxonomy and evidence on when adding agents pays.
14 references · 7 min read
June 28, 2026ExperimentationAlways-Valid Inference in Conversion Testing
A practitioner review of sequential testing methods and the statistical cost of peeking at A/B test results.
19 references · 10 min read
June 28, 2026ExperimentationBayesian vs. Frequentist Inference in Conversion Testing
A practitioner's guide to Bayesian and frequentist A/B testing: what p-values and posteriors really mean, the role of priors, and when each fits CRO.
13 references · 11 min read
June 28, 2026CRO ProcessWebsite Performance and Conversion: A Review of the Evidence
A review of the evidence linking Core Web Vitals and page speed to conversion, bounce, and search ranking, with practical guidance for CRO teams.
21 references · 11 min read
June 28, 2026ExperimentationStatistical Power, Sample Size, and Minimum Detectable Effect in A/B Testing
A practitioner's review of statistical power, the 80% convention, minimum detectable effect, and the sample-size formula that sets how much traffic an A/B test needs.
15 references · 10 min read