Why do automation tools get adopted or abandoned in the enterprise?

A YC-backed startup needed to evolve its product and positioning in a crowded workflow automation category. Competitive category mapping and in-depth ethnographic research revealed a category-level finding: 75%+ of automation attempts fail in enterprise because they force users to think like programmers instead of experts in their own domains.

This insight anchored a brand architecture, behaviorally-grounded user personas, a brutalist visual design identity, and a launch-ready operating system spanning assets like landing page, templates, and outbound strategy.

AutoComputer had a functional product: a notebook-style interface with 90+ app integrations but needed to sharpen its brand, positioning, and go-to-market. The automation category had calcified into two paradigms: visual builders that forced users to think in triggers and conditionals, and chat interfaces that lacked the structure for anything repeatable.

Deep Notebook sat between the two—a natural language notebook with structured sequencing to automate repeatable tasks. The challenge: to identify the highest value users and position a new entrant in a crowded category.

I ran two parallel workstreams over six weeks.

First, a competitive audit of 30+ automation solutions mapped against interface paradigm, workflow complexity, and market positioning to identify where the category had converged and identify potential whitespaces. This analysis revealed a whitespace: automating high-complexity workflows with consumer-grade simplicity. Positioning here required solving technical, UX, and trust challenges simultaneously.

Second, primary user research to pressure-test the product and surface new hypotheses. I designed a structured research sprint, sampling respondents across dimensions of workflow complexity and adoption posture. Interviews were semi-structured: 45–50 minutes of open-ended conversation surfacing the lived experience of knowledge work, followed by a card-sorting exercise to identify jobs-to-be-done and a think-aloud session observing how users conceptualized natural-language workflows in real time. The screener for this sprint was designed to double as a standalone survey—generating quantitative validation for early market hypotheses from 100+ respondents.

Our research revealed a category-level insight: automation efforts failed in enterprise because they forced users to think like programmers (in triggers, conditionals, and logic trees) rather than like domain experts with contextual judgment. Users resisted products whose logic they couldn't inspect or override.

This reframed the strategic challenge. I anchored the brand in "cognitive sovereignty"—the idea that tools should enable and elevate expert judgment rather than replace it with configuration logic. From that frame, qualitative synthesis produced four behavioral personas defined by cultural patterns of work, personal anxieties, and jobs-to-be-done, rather than by functional role descriptions.

The strategic frame translated into a complete launch system: a brutalist design language, a fully implemented landing page, 100+ workflow templates, an onboarding sequence, and a staged channel strategy.