Let’s talk about Growth, Research, and Optimization

CRO Roundtable 255

Share to:

Winners will receive notification via email. It’s the CRO Roundtable Roundup!

Thanks to Matt BeischelIqbal Ali, Shiva Manjunath, and Craig Sullivan for joining us. Want to get in on the action and talk with other cool CRO people? Then you should probably…

One Sentence Takeaway

I’m a busy person; give me the TL:DR

Structured context, smart metrics, and disciplined testing turn noisy data into reliable optimization insight.

The Notes

What we discussed this week

  • Context engineering for AI decision-making
    • tailor scraped content before model ingestion
    • AI needs structured data to make informed decisions
    • generic tools often fail to prioritize important information
    • context engineering is not solely a technical function
    • segment metadata to guide model focus
  • Rising AI hallucination rates
    • AI hallucination can propagate errors, and chain-of-calls compound errors
    • using the analogy of phantom road traffic to illustrate AI halucination propagation
    • multiple AI calls increase chances of hallucination
  • Price-ending and discount psychology
    • ending a price in a non-zero digit is called “charm pricing” and it’s effects are culturally dependent
    • round prices signal quality but not strongly supported
    • testing different price endings can yield varied results
    • Messaging savings effectively impacts customer perception
    • the “rule of 100” – when comparing currency versus percent displays, show whichever discount ammount is the larger number
    • regional pricing models affect fairness perception
  • Experimentation hygiene
    • AA tests catch noise and SRM early
    • split-URL redirects risk leakage issues
    • multiple control groups strengthen data integrity
    • SRM kills validity
  • The why’s of A/A Testing
    • Run A/A tests help diagnose SRM issues
    • Validating new metrics and setups
    • Identify discrepancies in analytics data
    • Understand noise in specific test areas
  • The video game industry focuses on the wrong metrics
    • loot-box rewards drive addictive play
    • metrics chosen to please executives
    • cherry-picked metrics misrepresent user satisfaction and ignore churn
    • user data is often overlooked in decision making
  • Meeting structure & semantic HTML
    • good agendas improve AI note-takers
    • headings and IDs aid website comprehension
    • HTML5 landmarks support accessibility
  • Form Design and Errors
    • Errors increase form abandonment rates
    • Design can mitigate errors through inline correction
    • Understanding user input intentions is crucial

The Quotes

Standout quips from this week

  • “Everything is okay in the bay.” — Iqbal Ali
  • “Bots suck. They’re so obviously bad.” — Shiva
  • “It’s like putting a moving sidewalk on top of a moving sidewalk.” — Matt Beischel

Book Club

Relevant reads recommended this week

  • No books this week, sorry!

Off-Topic Sidebars

Experimentation isn’t the only thing we talk about at the CRO Roundtable. There’s often a healthy dose of discussion on shared interests, personal passions, and hobbies.

  • Cycling the London canal network
    • wildlife and zoo monkeys spotted
    • stopping at cafés and eating fresh fruit
  • The video-game industry reliance on bots
    • an overabundance of bots inflate engagement, which are used to mask server issues
    • Solos mode with bots is not popular among players
    • Cheating and server issues are not adequately addressed
  • Video-game price hikes
    • setting market expectations with $80 titles
    • the risk of poor-quality games becoming overpriced
    • publishers testing higher tiers
  • Quick freelance gigs
    • two hours work is better than nothing
    • temporarily joining an agency as a hired gun

Sidebar Shareables

Amusing sidebar content shared this week

Share to: