On October 23, 2025, Write the Docs Bay Area hosted an in-person meetup at WRITER HQ, sponsored by WRITER and organized by Words n Logic. The theme, How Do Today’s Technical Writers Get Things Done?, inspired an evening filled with practical insights, personal stories, and cross-functional discussions that connected documentation, AI, and teamwork.
Each speaker explored a different aspect of “getting things done,” from using metrics to understand readers and measure documentation impact to applying AI as a creative partner and building scalable systems for collaboration. Together, their talks painted a picture of how documentation teams are evolving in the modern era. Today’s technical writers are not just producing content; they are building systems that connect people, processes, and information across entire organizations.
This was my third in-person meetup of 2025, following Smarter Documentation: AI, Tools, and Modern Workflows. Like the earlier events, this session brought together professionals from technical writing, developer relations, and product management who shared how documentation influences user experience and product success.
My long-term goal is to create TEDx-style events for technical writers, providing spaces where practitioners can collaborate, exchange ideas, and push the boundaries of what technical communication can be within enterprise technology.
Below are a few photos and highlights from the October 23 event that capture the spirit of the evening.
The theme of the event was “How Do Today’s Technical Writers Get Things Done?” In hindsight, I should have chosen a more engaging title, as I believe the generic name didn’t attract as much interest. Consider that a lesson learned for other organizers planning their own tech meetups.????
Each speaker explored the idea of “getting things done” from a unique perspective:
- Renée Carignan (Lead Technical Writer at Gem) demonstrated how metrics reveal the truth behind user engagement.
- Adam Michael Wood (Developer Relations at Martian) explained how AI can serve as a partner rather than a replacement for writers.
- Sarah Deaton (WRITER) shared how context engineering enables effective collaboration between humans and AI.
- Frances Liu (Co-founder of Promptless) closed the evening by outlining systems and habits that help unblock collaboration between technical writers, product managers, and subject-matter experts.
The complete slide deck with extra speaker content is embedded below for you to explore.
Here’s a brief summary of each talk along with thoughts of my top three findings.
Speaker 1 – Renee Carignan
Talk: Find the Frequency: Metrics for Better Docs and Happier Readers
Role: Lead Technical Writer @ Gem | Ex-Heap, Ex-Pendo
Summary of Renée’s Talk
Renée Carignan’s session focused on how data-driven insights can transform documentation from guesswork into an evidence-based practice. She explained that tracking metrics isn’t about vanity numbers, but that it’s about understanding user behavior and improving the reader experience. Beginning with entry-point data (like Google Search Console and AI-bot queries), she explained how to analyze pageviews, heatmaps, clicks, search terms, and feedback surveys to identify where users struggle or disengage.
Her central message: use the tools you already have, start small, and prioritize metrics that directly drive improvements to content and usability. Renée encouraged writers to “learn to love data” by setting a regular cadence for reviewing metrics, auditing their relevance, and always digging into the “why” behind user actions.
Top 3 Highlights
1. Homepage Heatmaps for UX Insights
Renée emphasized using heatmaps to visualize how users interact with the documentation homepage. By identifying which sections attract clicks and how far users scroll, writers can detect friction points and eliminate unnecessary elements.
Example: Her team discovered most users stopped at category cards and removed everything below, improving navigation clarity. This is a low-effort, high-impact method that can be done with free trials of heatmap tools and immediately improves user experience.
2. Search Data Analysis to Uncover User Intent
Analyzing internal search terms helps reveal what users are actually seeking in your docs and what they can’t find. Renée recommended tracking:
- Most common search terms over time to identify trending needs.
- Searches with no results, which often expose missing topics or confusing phrasing. Although 80 % may be noise, the remaining 20 % provide actionable insight. This is a practical, ongoing feedback loop for documentation improvement and content planning.
3. Feedback Surveys with Required Contact Info
Renée’s team at Heap improved feedback quality by requiring an email field in their doc-feedback form. This allowed them to follow up directly with users, connect them to support, and identify recurring pain points. By combining quantitative ratings with qualitative comments, teams can map satisfaction to specific articles and prioritize updates. It’s a lightweight but powerful way to close the loop between documentation and user success.
Speaker 2 – Adam Michael Wood
Talk: Pros and Cons of GenAI for Docs
Role: Tech Writer / Developer & Research Relations @ Martian | Ex-Google, O’Reilly, Facebook, Intel
Summary of Adam Michael Wood’s talk
Adam frames GenAI not as a push-button “agent” that replaces writers, but as a practical writing partner that boosts speed, reduces errors, and enforces planning when you set up good inputs and strong guardrails. He acknowledges real drawbacks (odd tone, verbosity, hallucinations in prose/code), then shows how to mitigate them with a prep->write->verify workflow: establish style and examples up front, structure prompts and scaffolds during drafting, and finish with testing/linting. The goal isn’t “set it and forget it,” but faster, cleaner docs with fewer lookups and mistakes, grounded in your team’s existing house style and codebase.
Top 3 Highlights
- Treat LLMs as partners, not agents
Don’t expect end-to-end automation; use the model to co-draft, expand outlines, and iterate. This mindset shift keeps you in control of tone, scope, and accuracy while still gaining speed. - Win in the setup: pre-generation guardrails
- Put a style guide in tooling (e.g., .cursorrules) or link to one.
- Seed with house-style exemplars (at least a solid Quickstart you wrote).
- Strengthen in-code docs (docstrings, types) so the model pulls accurate context.
- Ship with a verify loop: post-generation checks
Good inputs → better first drafts, less cleanup.
Counter verbosity and hallucinations with testing and linting (for code blocks and links), plus a human pass for scope creep and tone. Use narrow, specific prompts and code-sample scaffolds to keep outputs on-target before you even reach the review stage.
Speaker 3 – Sarah Deaton
Talk: Context Engineering for Docs Teams: Onboarding Humans and AI
Role: Developer Experience at WRITER
Summary of Sarah Deaton’s Talk
Sarah shared a practical story about scaling documentation quality and onboarding in an AI-assisted writing environment. When she first joined her team as the only dedicated technical writer, all writing standards existed “in her head.” As the team grew, she realized that much of the documentation knowledge lived only with her, making it harder for others to contribute efficiently.
This challenge inspired her to formalize the team’s style, structure, and workflows into machine-readable context files that support both humans and AI tools. She demonstrated how she built a system of documentation governance using files like STYLE-GUIDE.md for humans, .cursor/rules and .claude/CLAUDE.md for AI tools, and vale/ configurations to enforce standards during CI and content generation.
Her key message: context engineering isn’t just about writing better prompts; it’s about designing an ecosystem where both people and AI share the same knowledge, voice, and conventions.By externalizing her team’s “tribal knowledge,” Sarah created a sustainable system that accelerates onboarding, enforces consistency, and reduces friction in documentation creation.
Top 3 Highlights
1. Codify Your Team’s Context, Don’t Keep It in Your Head
Sarah’s breakthrough moment came when she realized that much of her documentation knowledge wasn’t easily shareable. Her solution was to document all the implicit standards, from tone to structure, in a STYLE-GUIDE.md file. This empowered others to write confidently, improved onboarding, and ensured consistency across the team.
2. Engineer Context for Both Humans and AI
Sarah showed how the same onboarding principles that help new teammates can also guide LLMs. She built a multi-file context system to teach AI tools how her team writes:
.cursor/rules→ defines tone and structure for AI-generated content.claude/CLAUDE.md→ provides project-specific contextvale/→ enforces consistency through linting rules
By giving AI tools structured, well-defined context, she ensured that generated docs aligned with her team’s established voice and standards.
3. Iterate, Integrate, and Automate
Sarah encouraged teams to start small and iterate:
- Base your style guide on an existing framework (Google or Microsoft).
- Integrate it into your AI tools (.cursor/rules, copilot-instructions.md, etc.).
- Generate one doc, review what didn’t align, and refine the rules.
She emphasized that context engineering is an ongoing process in which both humans and AI need structured guidance and continuous feedback to perform reliably.
Speaker 4 – Frances Liu
Talk: How to Get Unblocked as a Documentarian
Role: Founder @ Promptless | Ex-Cloudflare, OneSignal, Graft
Summary of Frances Liu’s Talk
Frances Liu shared a systems-based approach to overcoming documentation bottlenecks, grounded in lessons she has learned from some of the best technical writers she has worked with as a product manager and now as the founder of Promptless. Drawing on her experiences at Cloudflare, OneSignal, and Graft, Frances explained how great documentation teams do more than write; they engineer collaboration between product, engineering, and writing functions through smart tools, processes, and communication patterns.
She began by reflecting on her early career, when she unknowingly became a blocker to documentation by providing minimal context and late requests. Over time, working with skilled technical writers changed her perspective and led her to identify three pillars of unblockable documentation work: (1) aligning stakeholders on the value of documentation, (2) escalating effectively to get the resources and clarity needed, and (3) making it stupidly easy for subject-matter experts (SMEs) to provide context and reviews.
Frances illustrated these strategies through real examples and tool-based workflows, ranging from using session recordings and support analytics to demonstrate the impact of documentation to designing Slack workflows, GitHub PR templates, and automation rules that capture information early. Her central message was that the most effective documentation systems minimize friction, automate reminders, and make collaboration effortless, allowing writers to focus on clarity and quality instead of chasing information.
Top 3 Highlights
1. Show That Documentation Is the Product Experience
Frances emphasized that many PMs and engineers underestimate the role documentation plays in the customer journey. She shared how her own mindset shifted after seeing:
- Session recordings showing that users often start at the docs site
- Support tickets caused by outdated information
- Feature requests for capabilities that already existed but weren’t documented
By surfacing this data through analytics, heatmaps, or tagged support tickets, technical writers can prove how documentation directly shapes product perception and adoption, turning “docs” into a shared business priority.
2. Escalate Effectively: Complain with Purpose
Frances praised technical writers who know how to “escalate effectively,” asserting their needs with calm professionalism and clear reasoning. The most successful writers she’s seen:
- Break up questions and tag one responsible person at a time to avoid the “bystander effect.”
- Explain impacts and dependencies factually (“Docs can go live two weeks after UI finalization if…”).
- Reframe context requests as product feedback (e.g., “I’m trying to complete this task and can’t find the right option” instead of “I need more context”).
This approach fosters accountability and respect while helping writers get timely answers without friction.
3. Make It Effortless for SMEs to Contribute
To prevent writers from constantly chasing updates or context, Frances proposed a three-step system:
- Ambient Capture: Add small workflow hooks where information already lives (e.g., a “Docs update required?” checkbox in GitHub PRs or a #doc-requests Slack channel).
- Structured Context Requests: Use Slack forms, Jira templates, or Linear fields to prompt contributors for key details (Figma link, PRD, spec, meeting notes).
- Automated Nudges: Let software (Zapier, Vellum, n8n) handle polite follow-ups for missed reviews or next steps.
These lightweight systems reduce manual follow-up and keep documentation aligned with development in real time.
Together, these four talks painted a clear picture of how today’s effective technical writers balance data, systems, collaboration, and AI-driven efficiency. Renée Carignan showed how metrics like heatmaps, search data, and feedback surveys help writers understand reader behavior and continuously improve documentation UX. Adam Michael Wood reframed generative AI as a partner rather than a replacement, one that accelerates workflows when paired with strong style guides, structured prompts, and post-generation testing. Sarah Deaton demonstrated how context engineering empowers both humans and AI through shared standards and codified style systems that make onboarding and authoring consistent and scalable. Finally, Frances Liu highlighted how collaboration, escalation, and automation tools keep documentation teams unblocked and aligned with product development. Collectively, these sessions underscored a central truth: great documentation doesn’t just happen; it is engineered through the thoughtful integration of people, processes, and technology.
