# Build the Right Thing: Product Engineering (Part 1) — Kent C. Dodds, EpicProduct.engineer

## Executive summary

In the era of advanced AI agents, the value of software engineering is shifting from implementation (coding) to product judgment. The core thesis is that as development becomes commoditized, the most valuable skill is 'Product Engineering'—the ability to define and validate the 'right thing' to build. Product engineers must connect customer needs to technical constraints, focusing on defining the data model, workflow, and the smallest useful slice of a system, rather than simply executing a ticket to spec.

## Key takeaways

- The Shift in Engineering Value: As AI agents level the implementation playing field, the differentiator is no longer coding ability but the judgment required to determine if a problem deserves a software solution. The focus moves from 'Can we build it?' to 'Is it worth building?'
- Defining Product Engineering: Product engineers must take ownership of the upstream process, looking beyond the immediate request to understand the user's core problem. This involves making technical decisions regarding the data model, workflow shape, observability, constraints, failure modes, and identifying the smallest useful slice of the system.
- Validating Ideas with The Mom Test: To validate an idea, avoid asking users to evaluate the solution or diagnose the problem. Instead, ask behavioral questions like: 'When was the last time this happened?' and 'What did you do instead?' and 'What did it cost you to do that?' This reveals actual pain points and the cost of existing workarounds.

## Technical details

- System Architecture & Primitives: Engineers must understand the available technical primitives and the existing system constraints (e.g., cron job systems, database limitations). A product engineer must determine if a new feature requires expanding the system with new primitives or if existing ones can be reused, preventing the building of the wrong abstraction.
- Product Development Methodologies: The discussion covered several frameworks for idea validation and prioritization, including The Mom Test, Jobs to be solved, the Kano model, and defining the Minimum Viable Product (MVP). The goal is to quickly test the market with minimal investment.
- System Scalability and Design: When designing systems, it is critical to consider failure modes and scalability from the outset. The example of a real estate company building a high-scale system for all of Australia highlights the risk of over-engineering when a simpler, manual process would suffice.

## Practical implications

- When receiving a feature request, do not accept it simply as a ticket to implement. Instead, challenge the request by asking about the root problem, the user's current workaround, and the cost (time/money/effort) of that workaround.
- As an engineer, proactively engage with the business and user side of the product. The most valuable engineers are those who can bridge technical expertise with deep human understanding.
- When building, always aim for the 'smallest useful slice' of the system to validate the core hypothesis, rather than building a massive, over-engineered solution.

## Topics

Product Engineering, AI Agents, Software Architecture, Product Management, MVP, Build Process, User Experience (UX), The Mom Test, The Design of Everyday Things, Work OS

Source: https://www.youtube.com/watch?v=_fHTqOs5wQA