Mervin Manufacturing | AI Transformation Index

Overview

Mervin Manufacturing designs and hand-builds snowboards, skis, surfboards, and skateboards at its domestic factory in Sequim, Washington, selling premium branded products (Lib Tech, GNU, Roxy, Bent Metal) through wholesale and direct channels.

Executive Summary

Mervin Manufacturing is a physical manufacturing business whose core economics are structurally insulated from LLM-era AI disruption — the customer outcome (a high-performance board you ride) cannot be digitally substituted, and the production process requires physical infrastructure, materials science expertise, and craft labor that AI cannot replicate. The AI opportunity is correspondingly limited: with fewer than 10% of activities being cognitive in nature and a 45-person workforce built around factory production, there is no meaningful cognitive labor pool for AI to compress. For Altamont, the strategic AI story here is not transformation — it is that this asset is unusually durable against the disruption forces reshaping other portfolio companies, making it a low-AI-risk hold rather than an AI-upside play.

Durability Rationale

Mervin Manufacturing is structurally well-insulated from AI disruption because its core economic output is irreducibly physical — you cannot AI-generate a snowboard that performs on snow. The combination of domestic physical manufacturing infrastructure, proprietary composite materials processes, strong brand authenticity in a culturally discerning premium niche, and the absence of any intermediary layer to disintermediate makes this business largely orthogonal to the LLM disruption thesis. The primary durability risks are not AI-related at all: they are demand-side (participation rates in snow sports, climate change affecting ski seasons) and competitive (offshore manufacturers improving quality while undercutting on price), neither of which is an AI story.

Opportunity Rationale

AI's potential for Mervin is genuinely limited because the business is a physical manufacturing operation, not a cognitive labor enterprise. The company's value creation happens on the factory floor through skilled manual processes that LLMs and AI agents cannot touch. The cognitive work that does exist — design, marketing, forecasting — is a thin layer over a physical production function, and even if AI improved that layer dramatically, the impact on unit economics would be marginal given the cost structure. This is not a business where a PE sponsor bringing AI transformation expertise would unlock step-change value; the constraint on growth is factory capacity and market demand, not cognitive labor inefficiency.

Key Risks