The State of Product Development 🔍 — with Doug Peete

2026-05-29 · Source: Refactoring · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, extended

Summary

The Refactoring Podcast episode 63, featuring Doug Peete, Chief Product Officer at Atono, discusses key findings from a recent industry report on product development involving over 350 teams. The report reveals significant challenges: more than 60% of teams routinely discover missing tasks mid-cycle, and only one in four engineering teams receives clear success and acceptance criteria. A striking finding is that less than 10% of teams currently utilize AI for product requirements and specifications, a statistic Doug Peete found surprising. The discussion highlights the fragility of planning, the erosion of story quality, and the prevalence of critical knowledge residing in individuals' heads rather than documented sources. Peete advocates for upfront diligence in specs, cross-functional collaboration, and using AI, like Claude, for peer-reviewing stories and generating mockups to improve product development efficiency and clarity.

Key takeaway

For Product Managers aiming to enhance spec quality and team efficiency, you should actively integrate AI into your requirement-gathering process. Utilize tools like LLMs for early peer review of stories and to generate interactive mockups, which can significantly reduce mid-cycle discoveries and clarify acceptance criteria. Additionally, advocate for dedicated team time to build and refine shared AI context files, ensuring your entire team benefits from improved tooling and consistent product knowledge.

Key insights

Despite AI's potential, less than 10% of teams use it for product requirements, highlighting a significant adoption gap.

Principles

• Prioritize diligence in product specifications.
• Foster cross-functional team collaboration early.
• Dedicate time for "tool sharpening" and process improvement.

Method

Implement AI-assisted peer review for product stories and specs, conduct lean weekly design reviews, and utilize asynchronous "shoulder surfs" to gather early feedback and codify design decisions.

In practice

• Use LLMs (e.g., Claude) as a "thinking partner" for spec refinement.
• Generate HTML mockups with AI to visualize transitions.
• Capture product context and design decisions centrally for AI agents.

Topics

• Product Development
• Product Requirements
• AI Adoption
• Spec-Driven Development
• Knowledge Management
• LLM Applications

Best for: Product Manager, Executive, AI Product Manager, Director of AI/ML, Software Engineer

Related on AIssential

• Tech bros discovered coding isn't the hard part — AI has made building products easier, shifting the primary challenge to distribution and user engagement.
• How I'm Using AI Today — A structured, principled approach to AI agent management mitigates LLM context limitations and enhances productivity.
• Domain expertise still wanted: the latest trends in AI-assisted knowledge for developers — Developers increasingly use AI for learning and work, but a significant trust gap necessitates validation with traditional resources.
• How Uber’s product teams built a PRD reviewer that catches gaps before they reach leadership — Uber's AI-powered PRD reviewer enhances product document quality by providing comprehensive context for human judgment.
• From Napkin Sketch to Roadmap: 10x Your Product Workflow — Translating product ideas into clear PRDs requires structured thinking to avoid ambiguity and ensure shared understanding.
• Building production-ready LLM-powered applications @ Scale By the Bay — Transitioning LLM prototypes to production requires a pragmatic, product-focused approach centered on continuous evaluation and user outcomes.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Refactoring.