By Sana Remekie, CEO, Conscia.AI

Picture this: Your team has spent months creating comprehensive technical documentation in your content management system. It's well-structured, detailed, and theoretically helpful. But when someone asks ChatGPT or your company's AI assistant a specific question about your product, it either can't find the right information or gives a generic answer that misses the nuances of your particular setup.

Sound familiar? You're not alone. Most technical documentation sits isolated in content management systems, invisible to the AI agents that could make it incredibly useful.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

What If Your Docs Could Talk to AI?

Imagine if every piece of your technical documentation could be instantly discovered and understood by AI agents. Not just keyword searching, but truly intelligent retrieval that understands context, relationships between topics, and can provide exactly the right information at exactly the right moment.

That's exactly what this guide will show you how to build—a system that transforms your static documentation into an AI-native knowledge assistant.

The Big Picture: From Static Docs to Smart Answers

Here's what we're building: a pipeline that takes your structured documentation (like DITA content from Heretto CCMS) and makes it discoverable by AI agents through natural language queries. When someone asks "How do I install the CLI?", the system doesn't just return a list of potentially relevant pages—it understands the context, finds related prerequisites and troubleshooting guides, and delivers a comprehensive, structured answer.

Why This Approach Works

The secret sauce is treating your documentation as what it really is: structured knowledge with relationships and meaning. When you author content in formats like DITA XML, you're not just creating documents—you're encoding semantics. The system knows the difference between a task, a concept, and reference material. It understands prerequisites and can follow relationship links between topics.

This structure is what makes AI retrieval so much more powerful than traditional search.

The Architecture: Five Key Layers

Let's break down how this system works without getting too deep into the technical weeds:

1. Content Authoring (Your Writers Keep Writing)

Your content team continues working in their familiar environment—CCMS, creating DITA topics and maps. Nothing changes for them, which is crucial for adoption.

2. Smart Data Pipeline (The Magic Happens Here)

When writers publish new content, the system automatically:

• Exports the structured content from your CCMS

• Transforms it into a knowledge graph that understands relationships between topics

• Creates searchable indexes that support both keyword and semantic search

• Sets up real-time synchronization so updates appear almost immediately

3. Hybrid Search Engine (Best of All Worlds)

Instead of relying on just one search method, the system combines:

• Keyword search for precise term matching

• Vector search for semantic understanding

• Graph relationships for contextual connections

This "hybrid RAG" (Retrieval-Augmented Generation) approach means when someone asks about troubleshooting, the system can find not just documents containing those words, but related concepts, prerequisites, and follow-up tasks.

4. Intelligent Ranking (Business Logic Meets AI)

The system doesn't just return results—it ranks them intelligently using:

• Search relevance scores

• Business rules (like prioritizing task-oriented content)

• Relationship density (topics with more connections often contain richer information)

• Optional AI re-ranking for the most relevant results

5. AI-Native API (The Interface)

Everything is exposed through a Model Context Protocol (MCP) server that AI agents can easily integrate with. Whether it's ChatGPT, your custom assistant, or a voice bot, they can all ask natural language questions and get structured, actionable answers.

The Implementation Flow: How It All Works

Here's what happens when someone asks your AI assistant a documentation question:

1. Query Processing: The AI agent sends a natural language query to your documentation system

2. Hybrid Search: The system simultaneously searches for keyword matches and semantic similarity

3. Graph Enrichment: For each potential result, the system looks up related topics, prerequisites, and troubleshooting information

4. Intelligent Scoring: Results are ranked using both relevance and business rules

5. Structured Response: The AI agent receives enriched, contextual information to craft a comprehensive answer

All of this happens in under 200 milliseconds—fast enough for real-time conversations.

Real-World Benefits You'll See

For Your Users

• Instant, accurate answers to technical questions instead of hunting through documentation

• Contextual information that includes prerequisites, related topics, and troubleshooting steps

• Consistent experience across different AI interfaces (ChatGPT, company chatbots, etc.)

For Your Content Team

• No workflow changes—they keep writing in their familiar tools

• Automatic synchronization means their updates are immediately available to AI agents

• Analytics and insights about which content is most useful and where gaps exist

For Your Technical Organization

• Reduced support burden as users can self-serve more effectively

• Improved onboarding for new team members who can ask natural language questions

• Cross-system integration that connects documentation to products, support tickets, and other business data

Getting Started: The Building Blocks

To implement this system, you'll need several key components working together:

Content Export: Automated extraction from your current documentation system (like Heretto's Deploy API for regular JSON exports)

Data Processing: Jobs that transform your content into a searchable knowledge graph with proper relationships and metadata

Search Infrastructure: A hybrid search engine (like OpenSearch) that can handle both keyword and vector searches simultaneously

Orchestration Engine: A system that coordinates the search, graph lookup, and ranking processes

Agent API Gateway: An MCP server that exposes your documentation capabilities to AI agents

Key Success Metrics to Track

When you implement this system, here are the metrics that matter:

• Discovery Accuracy: AI agents should find relevant documentation with over 90% accuracy

• Response Speed: Search responses delivered in under 200ms for good user experience

• Synchronization Time: New content should be available to AI agents within 5 minutes of publishing

• Query Satisfaction: Users should find complete answers without needing to dig through multiple documents

Common Pitfalls to Avoid

Over-Engineering: Start with basic hybrid search before adding complex AI re-ranking. You can always enhance later.

Ignoring Content Structure: The power of this approach comes from structured content. If your documentation isn't well-structured, fix that first.

Neglecting Performance: AI agents need fast responses. Optimize for speed from day one.

Forgetting the Human Element: Even with great AI integration, your content still needs to be well-written and logically organized.

The Bottom Line

Your technical documentation is a goldmine of structured knowledge that AI agents can unlock—if you build the right bridges. By implementing a system that combines hybrid search, knowledge graphs, and intelligent ranking, you transform static documentation into a powerful AI-native knowledge assistant.

The result? Your users get instant, accurate answers to complex technical questions, your content team sees their work having greater impact, and your organization reduces support burden while improving user experience.

Most importantly, this isn't a replacement for good documentation—it's a multiplier that makes your existing investment in content dramatically more valuable.

Ready to make your documentation AI-native? The pieces are all available, and the benefits are real. The question isn't whether this approach works—it's how quickly you can implement it to start seeing the benefits. 🤠

Sana Remekie is CEO and co-founder of Conscia.ai, a Toronto-based enterprise software-as-a-service company.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

This post was inspired by a thoughtful LinkedIn post from content operations expert Rahel Anne Bailie. In it, she identifies five common situations where content professionals are expected to sacrifice their own efficiency to make life easier for others — usually in ways that increase overhead, introduce unnecessary complexity, and lead to long-term content chaos.

I’ve taken each of her five points and written a short reflection about each aimed at technical writers, sharing what it often feels like when these scenarios show up in the real world.

Imagine being handed a butter knife and asked to slice through a frozen ham. That’s what it feels like when you’re expected to write enterprise-grade content in tools meant for grocery lists and birthday invitations.

“But it’s free,” they say, while you scream internally, massaging the carpal tunnel you got from trying to force a table to behave in Microsoft Word.

Sure, you can do it. You can also cut your lawn with nail scissors. But should you?

Let’s stop pretending that “free” tools are actually free. They cost us time, sanity, and clean content. Tools should serve the process, not sabotage it.

Read more

If you’ve ever watched hard-won product copy get mangled in translation, you’ll appreciate the recorded webinar “How Semantically Rich Global Content Drives Profitability and Scale.” It’s free, on-demand here.

Who’s talking?

• Scott Abel – Content Strategy Evangelist at Heretto and founder of The Content Wrangler. Scott distills decades of structured-content know-how into clear actions technical writers can use today.

• Ron Myers – Director of NA SMB Omnichannel Digital Transformation at Lenovo. Ron brings 25 years of leading digital-first content programs that tie directly to revenue.

Who’s footing the bill?

The webinar is sponsored by Acclaro—a global translation and localization partner with offices on four continents. Acclaro combines expert linguists and AI-enabled workflows to help brands ship consistent, market-ready content in 125 + languages.

Why hit play?

• See how semantic markup shrinks translation memory costs.

• Learn practical steps to make reuse and personalization painless.

• Hear real metrics that connect structured content to profit.

Set aside 40 minutes, grab a coffee, and discover how better structure today means faster, cheaper global delivery tomorrow. Your future multilingual readers—and your budget—will thank you. 🤠

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

My friends call me a digital music collage artist—which is a fancy way of saying I repurpose and rearrange bits of audio into reworked remixes of popular dance tracks. I take stems—those individual pieces of a song like vocals, bass lines, and drum loops—and weave them into something fresh. Something that moves you, maybe even makes you sweat a little (or a lot) on the dance floor.

See also: The Audio Wrangler channel on SoundCloud

That instinct to break things down and build them back up again doesn’t stop with music. It shows up in my art, too.

I make art using discarded components from old, broken, costume jewelry. Not to fix them. Not to make new jewelry. But to make something else entirely—art born from scraps of jewelry. And, not surprisingly, it’s the same instinct that drives my work in technical documentation.

Read more

Let’s be honest: most content improvement projects begin not with a bang but with a passive-aggressive comment in a Zoom chat.

We see the chaos. We see the ancient templates that haven’t been updated since Windows Vista. The publishing process that requires a prayer, a goat sacrifice, and five rounds of manual PDF tweaking. And we think: this is a disaster.

And yet, we smile. We nod. We click “resolve comment” and move on. Because advocating for change? That sounds exhausting. And vaguely like project management.

Here’s the truth we need to face: we’re the best people to make the business case for fixing this mess.

We know the problem. We live it. And we can connect the dots between “this doc is a hot mess” and “this is costing us real money.”

We Know Exactly Where the Content Skeletons Are Buried

Let’s not pretend otherwise. We are aware of which content exists in twelve different versions across five systems. We know which reviewers never respond unless bribed with tasty snacks. We know which parts of our publishing process involve someone manually replacing curly quotes—every single time.

And while it’s tempting to keep it to ourselves (because it makes for great group chat gossip), we are in the best position to shine a light on what’s broken. More importantly, we can explain what fixing it could save—in time, money, and human dignity.

We’re Bilingual (and One of Those Languages Is XML)

We speak human. We speak machine. Sometimes, both are in the same sentence.

We can explain to developers why their UI labels don’t make sense and then turn around and explain to execs why their chatbot is hallucinating because the source content is garbage.

This ability to translate chaos into coherent strategy? That’s a superpower. And it’s precisely what’s needed to get people to care about content problems that are otherwise invisible until something explodes (usually late in the day on a Friday).

We Can Make the Invisible Gloriously Visible

Good content is like good Wi-Fi. You only notice it when it’s not working.

So, let’s make the consequences of bad content impossible to ignore:

• Repetitive support tickets about the same missing steps

• Global customers waiting weeks for translated updates

• Engineers explaining the same thing in Slack ‚ again and again

Then, let’s show what happens when content is supported:

• Tickets drop

• Launches speed up

• Reviewers stop weeping

You don’t need a PhD in data science. Just tally up the hours you spend doing the same thing over and over. That’s your ROI, right there.

We Don’t Need a Title to Lead—Just a Spine

We don’t have a “Senior Vice President of Content Optimization Strategy and Global Enlightenment” in our email signatures.

But we have something better: credibility.

People trust us. They come to us to make sense of nonsense. That means we already have a seat at the table—even if it’s one of the folding chairs off to the side.

So, let’s use our credibility to pitch ideas that improve things. Let’s get bold enough to say: “Hey, if we fix this, maybe we won’t have to spend our Fridays re-exporting 14 help topics because the logo changed by one pixel.”

We Can Start Small and Still Win Big

Here’s how to begin your journey as a business case cheerleader without breaking out in hives:

✅ Identify the recurring pain. What do we complain about on Slack the most? That’s your project.

✅ Find your allies. Who else is sick of it? Support? Localization? Sales? Pull them in. Make it a movement.

✅ Tie it to something shiny. Is the company launching AI features? Expanding internationally? Link your content improvements to those significant, budget-worthy initiatives.

✅ Talk dollars and hours. “Fixing this will save us 250 hours and cut our translation costs by 30%.” No one ignores that sentence.

Let’s Fight For Better Content

If we don’t fight for better content, who will?

Carl from Accounting? Please.

We are the ones with the receipts. The ones who know how bad it is and how good it could be.

So, let’s stop whispering about our content woes and start waving the flag. Let’s turn our frustration into action. Let’s write the kind of business case that makes a VP say, “Wait—we’re not already doing this?”

Because we’re not just writers.

We’re advocates.

We’re translators.

We’re professional dot-connectors.

And honestly? We’re kind of tired of doing the same workaround for the 100th time. 🤠

You know that reel where a chipper robot voice narrates a Reddit breakup over Subway Surfers footage? The one you swipe past while convincing yourself this is “research”? That’s not harmless background noise—it’s rehearsal footage for a political movement that prefers spectacle to substance and uniformity to thought.

RattleSnake Studio says it out loud in their video “AI Slop, Technofascism, and You.”

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Watch this video on YouTube and share it with others.

AI Slop, Technofascism, and You

Wait—what exactly is “AI slop”?

AI slop is mass‑produced, low‑effort, AI‑generated junk—text, video, images—optimized for engagement metrics, not truth or usefulness. Think of it as content feedlot runoff: plenty of volume, zero nutrition.

Tech reporters and search pros are already using the term to describe the flood of machine-made filler clogging social and search results.

Fascism Runs on Aesthetics And AI Delivers Aesthetic-On-Demand

Walter Benjamin warned that fascism “aestheticizes politics”—it gives people goosebumps instead of rights. It turns power into a pageant while leaving the power structure intact.

Historically, fascist regimes curated a narrow visual vocabulary—heroic bodies, pastoral nostalgia, mythic pasts—and censored or criminalized artists who colored outside the lines. The Nazis literally staged a “Degenerate Art” show to ridicule work that didn’t fit the template, while centralizing “acceptable” imagery through a propaganda ministry.

Enter generative AI: a machine built to remix templates and crank out nostalgic pastiches at scale.

• Uniformity? Check.

• Backward-looking mythos? Easy.

• Speed over nuance? Built in.

No need to wrangle opinionated artists with pesky ethics when you can just prompt your way to a thousand glossy Rockwell knockoffs, extra fingers included.

But I Write Release Notes, Not Manifestos, So, Why Should I Care?

Because documentation is how organizations tell the truth about what their products do, what users can and can’t do, what’s broken, and what’s at stake. When low-effort AI slurry seeps into doc sets, you don’t just annoy readers—you erode trust, blur authorship, and normalize myth-making. Your style guide becomes a speed bump, not a guardrail.

Authoritarian movements don’t just target “capital A” Artists. They go after anyone who controls narratives, defines terms, or keeps receipts. That’s you.

1. Provenance or it didn’t happen

Bake source, authorship, and review metadata into your content model. Who wrote it, who reviewed it, what model (if any) helped, when it was last fact-checked—make that machine-readable and auditable.

2. Institute an “AI Slop Check”

Before you ship anything generated or assisted by AI, ask:

• Does this add anything new, or just remix clichés?

• Can every factual claim be traced to a verifiable source?

• Can I point to a human who stands behind this? If the answer is “uhhh,” you’re looking at slop. Fix it or kill it.

3. Document reality, not just requirements

The RattleSnake video urges creatives to connect the personal and the political, to show real human stakes. In docs, that means capturing actual user scenarios, edge cases, and consequences—not just the happy path marketing wants. Propaganda loves a vacuum; fill it with verifiable detail.

4. Myth filters in your workflow

Structured authoring plus governance = fewer unsupported claims. Require citations or subject matter expert sign-off for every “we’re the first,” “industry-leading,” or “everyone knows” sentence. Schedule periodic audits for marketing creep.

5. Teach your team to spot the tells

Host short, repeatable trainings on:

• Common AI artifacts (those weird hands, uncanny phrasings, generic “inspirational” tone)

• How deepfakes and synthetic text spread

• Your internal rules for when AI is allowed and how it’s disclosed

6. Keep the hope (and humor) alive

Fascism’s endgame is compliance through despair. Authentic voice, honest status pages, and humane UX writing keep readers oriented—and remind them humans are still in the loop. That matters more than you think.

A Quick (And Reusable) Pre‑Publish Checklist

• Human in the loop? Named reviewer approved it

• Source trail visible? Every fact has a link you’d defend in public

• Original value? Solves a user problem, not just fills a Jira ticket

• Model disclosure? Which tools were used, and how

• Inclusive & accurate? No in‑group mythologizing, no nostalgia gloss

• Versioned? Easy rollback if an “auto-improve” feature degrades meaning

Copy, paste, stick it in your CMS workflow. Tattoo it on your product owner if necessary.

The Takeaway

AI isn’t evil. And, slop isn’t inevitable. But fascism loves a shortcut, and generative tech hands it the keys: infinite spectacle, zero accountability. Our counter is boringly heroic: receipts, rigor, reality—told by humans, for humans.

Keep making real human documentation. Keep telling true stories. Keep that pilot light on. 🤠

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

I hope you'll join me in Brussels at Hotel Le Plaza Brussels, February 2-3, 2026, for DITA EUROPE 2026 — the conference for technical writers and information developers looking to build advanced information management and delivery capabilities. If your technical documentation team produces content using the Darwin Information Typing Architecture (DITA) — or you’re considering doing so ‚ this is the conference for you.

Call for presenters

The call for presenters is open now. If you've got a case study, best practices, or lessons learned to share, the conference organizers want to hear from you.

👉🏼 Submit your proposal (Sept 1, 2025 is deadline).

And, if you decide to attend, you can save 10% off the cost of registration using discount code — TCW — at checkout.

DITA Europe is brought to you by the Center for Information Development Management, (CIDM) — a membership organization that connects content professionals with one another to share insights on trends, best practices, and innovations in the industry. CIDM offers networking opportunities, hosts conferences, leads roundtable discussions, and publishes newsletters to support continuous learning. CIDM is managed by Comtech Services, a consulting firm specializing in helping organizations design, create, and deliver user-focused information products using modern standards and technologies.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

If you’re a technical writer, you probably have a love-hate relationship with your content management system (CMS). On a good day, it’s your best friend who helps you organize, publish, and maintain documentation. On a bad day, it’s like trying to communicate with an old fax machine that only speaks Klingon.

Let's face it — documentation doesn't exist in a vacuum. It needs to connect with the rest of your company's tools to stay relevant and valuable. If you've ever wished your CMS could play nice with those tools—congratulations — you're already thinking about API-driven architecture!

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

What Is '“API-Driven,” Anyway?

API-driven means that a system isn't a walled garden—it's an open, flexible platform that communicates with other software through Application Programming Interfaces (APIs).

Imagine your CMS as a restaurant. In a traditional setup, customers (users) can only get what's on the menu (built-in features). However, An API-driven restaurant allows customers to send in their ingredients and have the chef cook something custom.

For technical documentation teams, an API-driven Component Content Management System (CCMS) lets you connect your content to the rest of your tech stack. That means your documentation doesn't have to live in isolation—it can interact with support ticketing systems, version control repositories, customer relationship management (CRM) platforms, and more.

Related: API-driven architecture, simplifying software development

The Benefits of an API-Driven CCMS

Integration with Other Tools (Because No One Likes Copy-Pasting)

Manually moving content between platforms can quickly become a nightmare. With an API-driven CCMS, you can automatically push documentation updates to customer service tools like Zendesk or Salesforce, ensuring that support agents always have the latest information—no more frantic emails begging someone to update a knowledge base article.

Automation for the Win

If your documentation team follows a Docs-as-Code workflow, an API-driven CCMS can integrate with version control systems like Git or GitHub. When a developer updates the codebase, your docs can automatically reflect those changes. No longer will you need to chase down engineers to ask, "Wait, did we release that feature?"

Dynamic Content Delivery

Static documentation is so 2010. With API-driven content, you can serve up personalized, real-time documentation that updates based on user data. Imagine a chatbot that pulls live documentation from your CCMS, or a mobile app that displays different help articles based on the user's permissions.

With dynamically delivered self-service content, customers can fix their problems anytime. As tech evolves, this isn’t a bonus—it’s the new normal, reshaping what “customer support” means.

Related: Is Disorganized Content Making Your Chatbot Look Bad?

Scalability Without the Headaches

As your documentation needs grow, an API-driven system makes it easy to integrate with new platforms without a complete overhaul. Want to pull in analytics from Looker Studio or push updates to a Learning Management System (LMS)? An API-driven CCMS can make it happen with minimal effort.

Security and Compliance

If your company deals with sensitive information, APIs help ensure your documentation follows security best practices. You can integrate Single Sign-On (SSO) for authentication or set up API-based audit logs to track changes and access history.

Real-World API Integrations That Make Your Life Easier

Here are just a few of the services that you might want to connect to your API-driven CCMS:

• Zendesk – Keep your support articles in sync with your documentation.

• Salesforce – Serve up relevant documentation directly in CRM workflows.

• GitHub/GitLab/Bitbucket – Automate documentation updates when code changes.

• Slack – Get notifications when content is updated or reviewed.

• Confluence – Sync content between your CCMS and internal knowledge bases.

• Google Analytics – Track how users interact with your documentation.

• Chatbots – Provide real-time answers based on live documentation.

Related: How to Nail Single-Source Publishing

Wrapping It Up (Without a TL;DR Cop-Out)

If your documentation team is spending more time managing content than actually writing it, it's time to consider an API-driven CCMS like Heretto. By integrating with your company's existing tools, you can streamline workflows, automate updates, and ensure your content reaches the right people at the right time.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Right‑to‑Repair laws keep expanding—Oregon’s tough electronics law took effect on Jan 1, 2025 and legislators in more than 40 States are considering similar requirements. These rules make technical documentation a compliance hotspot.

To see how prepared our community is, I’ve put together a short, anonymous Right‑to‑Repair Readiness Poll (five questions, under two minutes).

• Take the poll

See also:

• How Will The 'Right To Repair' Revolution Impact Technical Writers?

• The Environmental Impact of E-Waste

• Data Requirements For Digital Product Passports, and

• Ecodesign for Sustainable Products Regulation

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

All digital product passport (DPP) information requirements under the Ecodesign for Sustainable Products Regulation (ESPR) must be purposeful and relevant. Each data element should have a well-defined scope and deliver a tangible benefit for users at different stages of the product's life cycle.

The ESPR outlines that a DPP must include the following types of data, as well as any additional "voluntary data" that supports digitized and streamlined information-sharing processes.

Product Identification

Includes product name, model number, serial number, and other distinguishing features. A unique identifier is essential for connecting the physical product with its associated digital information.

Materials

It covers all materials used throughout the product's life, including their origins, the availability of critical raw materials, and material flow data.

Product Design

Designers and engineers transform materials into the final product configuration, including details on material flow and other relevant aspects.

Technical Specifications

Includes performance characteristics and technical parameters such as power consumption, size, weight, and other relevant metrics.

Product Lifecycle

Defines the sequential stages of a product's existence, from raw material extraction to end-of-life treatment.

Installation and Maintenance

Provides comprehensive guidance on how to install, use, maintain, and repair the product—including third-party software where applicable.

Material Composition

Details the inclusion of recycled and renewable materials within the product.

Microplastics

Information on the potential release of microplastics or nanoplastics during manufacturing, use, or disposal.

Environmental Impact

Reports on the product's carbon footprint, material footprint, and other relevant environmental effects.

Repair and Replacement

Offers data on spare parts, compatibility of components, and modular design, along with non-destructive disassembly procedures for repairs and replacements.

Reuse and Recycling

Explains how users can reuse, remanufacture, or recycle materials to promote circularity and reduce waste.

Product Maintenance History

Includes records of service events, repairs, and upgrades, as well as maintenance schedules.

Warranty Information

Outlines the terms and conditions of the warranty, including what is covered and for how long.

Energy Recovery

Covers methods for generating energy from the product through incineration, including heat recovery from combustible waste.

Substances of Concern

• Identification: Lists the substance name, identification codes (IUPAC, EC number, CAS number), location in the product, and concentration levels.

• Usage and Handling: Offers safe handling instructions, disassembly procedures, and protocols for managing hazardous substances.

• Lifecycle Tracking: Includes procedures for monitoring substances of concern across the product's lifecycle.

• End-of-Life Management: Provides guidance on disassembly, recycling, and safe disposal of hazardous materials, including reuse, refurbishment, and remanufacturing options.

If you’re a tech writer and you’ve never heard of the EU’s new Digital Product Passport (DPP) requirement, you’re not alone. It hasn’t exactly made front-page news on Technical Communication Monthly (if that were a thing—it’s not). But make no mistake, this new regulation, introduced by lawmakers in 2022, it took effect in 2024, with major rule‑setting actions (working plan, standards) are expected by December 2025. Full implementation is anticipated by 2030.

Why Does This Matter To Tech Writers?

The DPP requirements will change how we create, share, and manage product information. If you think these requirements don't apply to you because you live or work in the United States—or anywhere not covered in metric measurements—you might want to sit down. This regulation isn’t just some Eurocratic experiment in circular economics. It’s a global wake-up call—and technical communicators everywhere need to listen.

So… What Is a Digital Product Passport?

Think of a DPP as a digital file cabinet stapled to your product (digital, of course), full of standardized, machine-readable facts about what’s in it, how it was made, how to fix it, and how to dispose of it without setting the Earth on fire.

DPPs are part of the EU’s push for a circular economy—a polite way of saying “we’d like to stop drowning in e-waste and polyester.” These passports will be required for a range of products, starting with batteries and electronics and expanding to textiles, furniture, and more.

Why Should You Care?

Because if your company sells products into the EU—or plans to—you’re going to need to generate this data. And guess who’s already sitting on a pile of that information, or at least knows where it lives?

That’s right. You. The technical writer.

The person who already documents material content, safety specs, repair steps, and take-back instructions—only now, you’ll need to do it in a structured, standardized format designed for machines and humans. (Don’t worry, robots still can’t proofread.)

Even if your product never crosses the Atlantic, chances are your company’s regulatory, sustainability, or marketing teams are paying attention. Because once the EU sets the bar, other countries tend to follow. (see also: GDPR panic of 2018.)

What’s In a DPP, Anyway?

Read more

In the past, we wrote for people—real, living, blinking humans who clicked, skimmed, and sometimes said thanks (usually when something broke). But now? Now we're writing for a split audience: the human on one side and a tireless army of robots on the other. These bots don't sleep, they don't eat, and they don't appreciate your clever pun in paragraph three.

So what's a technical writer to do when your job description quietly expands to include "AI whisperer" and "SEO sorcerer"? This post is your survival guide.

The New Audience: Humans, Crawlers, and Language Models

Three groups now read your content: the human who needs help fast, the search engine crawler that indexes your page, and the large language model (LLM) that spits out your words in someone else's chatbot session. All three are trying to understand what your content is about. If it's not clear, they'll get confused, make things up, or move on. Think of them like houseguests: keep things tidy, label the snacks, and don't make them hunt for the bathroom.

The Role of Clear Headings

Headings aren't decoration—they're navigation. Clear, consistent, logical headings help both humans and machines know what to expect. They act as signals: "This section explains setup." "This part solves a specific problem."

Don't get artsy or vague. "Things to Know" won't cut it. "How to Configure the Widget Permissions Panel Without Crying"—now that's helpful. Give readers and bots a map, not a maze.

Writing That Actually Works

Use active voice. Keep sentences short. Deliver one idea at a time. You're not just writing for people—you're writing for machines that will rephrase, summarize, and remix your content in ways you didn't intend. If your sentence sounds like it came from a committee with a thesaurus addiction, expect it to land weird when paraphrased.

Think of your writing as source code. If it's sloppy, the build will fail.

Metadata, While Invisible, It's Also Mighty

Metadata is what helps machines understand what your content really is. Your page title, meta description, image alt text, and tags whisper clues to crawlers and models alike. Don't settle for auto-filled nonsense like "Page 1."

That's the digital equivalent of labeling your spice rack with "stuff."

Use structured data when you can. Schema markup makes it easier for machines to categorize and serve your content correctly. Robots love structure the way librarians love the Dewey Decimal System.

Write for Voice, Not Just Eyes

Keywords are still helpful, but you're seasoning, not stuffing. Avoid awkward repetition and focus on sounding human. Many users now ask questions out loud, so write like you're answering them in conversation.

Use natural phrasing. Cut the jargon unless you're sure it's part of your audience's vocabulary. Speak to both the voice assistant and the person shouting at it.

Old Content Still Matters

Just because you wrote it years ago doesn't mean it's retired. LLMs don't care about freshness—they'll pull from that dusty 2019 how-to if it's still floating around. If it's outdated, it may provide incorrect information to someone who trusts it.

Review your older content regularly. Fix links. Update details. Clarify anything that might age poorly. A little maintenance goes a long way when the audience includes time-traveling AI.

Tools to Keep You Sane

Use tools that help you bridge the human-robot gap. Grammar checkers to catch drift. SEO plugins to highlight opportunities. AI preview tools that show what a chatbot might pull from your content.

And don't forget to test. Google your content. Ask ChatGPT to summarize it. If what it grabs doesn't make sense, revisit what you've written.

Sometimes, the AI quotes you like scripture. Make sure it's not quoting a rough draft.

You're Not Just a Writer—You're a Translator for the Machine Age

You're bridging the gap between human need and machine logic. Between the question someone asks at 2 AM and the answer that floats back in a chatbot's voice.

It's not easy. But neither is writing documentation for software that ships while you're still editing the release notes.

Write with care. Structure like a librarian. And when in doubt, imagine you're explaining it to both your future self and an extremely literal alien with a search engine for a brain.

Got tips? Mistakes? Robot horror stories? Share them in the comments. Misery, after all, loves metadata.

I once tried mapping out my career journey on a whiteboard. It started strong—college, first job, some light existential dread—but by the third marker color, it devolved into something resembling a subway map drawn by a raccoon with a Red Bull problem.

That, my friends, is how most customer journey maps end up: overly optimistic at first, then quickly unraveling into chaos and assumptions.

And yet, if you're a technical writer, someone probably waved a journey map in front of you at some point and said, "This is where your content fits in." It's like summing up your entire contribution to the customer experience in a tidy box labeled "Post-Purchase Onboarding PDF."

So, let's discuss how customer journey mapping can go wrong, especially when tech writers are not invited to the meeting until it's too late.

1. The Fantasy of the Linear Journey

You've seen that clean diagram in which a customer starts at awareness, gently floats through the consideration phase, gracefully landing at loyalty like a well-trained swan.

Reality check: people don't shop like they're on a conveyor belt at IKEA.

They loop, backtrack, get distracted by shiny objects, and rage-quit halfway through. If you don't account for detours, you plan your customer journey like a road trip using a 1984 Rand McNally atlas—without GPS or a clue.

2. It's Built on Vibes, Not Data

Some journey maps are less "customer insights" and more "vague recollections from a marketing brainstorm fueled by muffins."

If your map reflects what you think customers do instead of what they actually do, congratulations—you've created a piece of customer fiction.

Technical writers, you deserve better. Ask for real data: search queries, support tickets, behavioral heatmaps. Bonus points if someone on the team has spoken to a customer during the past decade.

3. Everyone Is One Dimensional

"Meet Sarah. She's a project manager who loves yoga and needs scalable documentation tools."

Cute. But also, no. 🙈

People don't exist as frozen archetypes. They're multitasking (okay, not really — but many believe they are), context-switching, sometimes grumpy beings whose needs change hourly. One minute, Sarah's researching your product; the next, she's trying to figure out why your documentation refers to three different UI labels for the same feature.

Don't write for cardboard personas. Write for people who are busy, tired, and want to know where the hell the settings panel moved to.

4. No Writers in the Room

Well-intentioned teams create a lot of journey maps in workshop sessions filled with sticky notes, lattes, and absolutely no content people.

That's a mistake. A huge one.

Writers (and, by extension, customer support staff) know what users ask, where they get stuck, and how often they type "how to turn it off and back on again." Their absence from the mapping table means the voice of reason—and readability—is missing.

And guess who gets blamed when the help content doesn't magically solve everything?

5. It Ends at the Sale

Oh good, we've sold them the thing! Let's pack up and move on. Surely they'll never need help again.

Except… they will.

And if you don't map what happens after the sale—onboarding, troubleshooting, renewal—you're just designing a haunted house with no exits.

Don't let support content be an afterthought. That's your moment to shine.

6. It's Treated Like a Piece of Art

Some journey maps get framed. Literally. In hallways.

And then they're never updated again.

A journey map should be a living, breathing, occasionally grumpy document that changes when your product does. If yours still references that chatbot you retired in 2022, it's time for a refresh.

7. Nobody Measures If It Works

Is the journey getting better? Are users actually succeeding? Or are they rage-clicking through your knowledge base like it owes them money?

If you're not tracking outcomes—or worse, can't track them because the content lives in twelve different tools with no analytics—your map is more aspiration than guidance.

Final Thought (Before You Add Another Sticky Note)

If customer journey maps are going to be useful, they have to be honest. Messy. Real.

Most importantly, they have to include content that actually helps customers succeed. That's where you come in, dear technical writer.

Remember the journey is complex, but your contribution doesn't have to be an afterthought. You're not just writing documentation. You're plotting the breadcrumbs that help customers find their way out of the dark forest. Preferably without screaming. 🤠

Metrics That Matter: ROI Of Documentation In The AI Era

🗓️ June 11, 2025 — 8am PT
👉🏼 Register

Join Sofiya Minnath ZA, Senior Manager of Growth Engineering at fabric.inc, as she delves into innovative metrics that go beyond traditional KPIs to demonstrate the real value of documentation.

Building a Business Case for DITA and a CCMS for Technical Documentation

🗓️ June 12, 2025 — 8am PT
👉🏼 Register

Join us for a chat with Melanie Denise Davis, Founder and CEO of Dragonfly Diva Docs. Discover how your tech docs team can communicate the value of shifting from unstructured content creation to leveraging the power of semantically-rich structured content managed in a component content management system (CCMS).

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Rewrite the Rules: GenAI and The Future of Content Strategy

🗓️ June 12, 2025 — 10am PT
👉🏼 Register

Amanda Patterson cuts through the AI hype and gives you the real talk on making GenAI a seamless, strategic part of your content ecosystem. You’ll learn where AI fits, what risks you actually need to worry about, and how to convince leadership that this isn’t just another tech trend—it’s the future.

Graph-Driven Retrieval Augmented Generative AI Powered By DITA

🗓️ June 18, 2025 — 10am PT
👉🏼 Register

Can generative AI benefit from structured content? What role play knowledge graphs in this? And what does all of that have to do with DITA? After they discovered the shortcomings of vector-based RAG there is a buzz in the industry around knowledge graph-driven retrieval-augmented generation (Graph RAG). This session will examine whether and how DITA fits into that equation. Michael Iantosca, Senior Director of Content Platforms and Knowledge Engineering at Avalara will demonstrate a full-scale implementation of a graph-driven RAG based on intelligent structured content.

[Book Release] Docs-as-Tests: A Strategy for Resilient Docs

🗓️ June 19, 2025 — 8am PT
👉🏼 Register

Join us for a chat with Manny Silva, head of documentation at SkyFlow, and author of the newly released book, “Docs as Tests: A Strategy for Resilient Technical Documentation.” Manny will introduce us to the Docs-as-Tests approach—an innovative strategy that treats documentation as a first-class citizen in the software development lifecycle. Stick around for live Q&A—bring your questions and curiosity. One lucky viewer will win a free copy of Manny’s new book.

How GenAI Helps Developers Craft Better Internal Documentation

🗓️ June 25, 2025 — 8am PT
👉🏼 Register

Thiemo von Gillhaußen explores how a hybrid approach—combining generative AI, intelligent linguistic checking, and retrieval augmented generation (RAG)—can help contributors like software devs and engineers, write better content, faster.

Discover how Congree enables you to govern generative AI output using corporate style guides and terminology preferences—ensuring content quality without overwhelming contributors. Learn how using a local large language model (LLLM) to automatically align text with company standards while simplifying the writing and editing process for contributors.

The Power of Semantic Content Knowledge Graphs

🗓️ June 26, 2025 — 8am PT
👉🏼 Register

Join us for a chat with Arle Lommel, senior analyst at CSA Research and expert in language technology and semantic interoperability. Together, they’ll explore the growing importance of semantic content knowledge graphs and their role in improving content findability, automation, and personalization.

Lower the Barrier of Entry: Empowering Non-writer Contributions to Docs

🗓️ July 10, 2025 — 10am PT
👉🏼 Register

A presentation from Manny Silva, head of documentation at SkyFlow, during which he will demonstrate how modern tooling can create a supportive environment that empowers non-writers to confidently contribute to documentation.

The Sky Is Falling But Your Content Is Fine

🗓️ July 23, 2025 — 10am PT
👉🏼 Register

Jack Molisani explores how structured content will future-proof your content operations no matter what tech trends come along. Learn how to prepare content once and publish everywhere, from toasters to chatbots to jumbotrons and beyond.

What's Wrong With AI-Generated Technical Documentation?

🗓️ July 24, 2025 — 8am PT
👉🏼 Register

Tom Johnson, creator of I’d Rather Be Writing, and Fabrizio Ferri-Benedetti, technical communication expert — will unpack the challenges and pitfalls of relying on AI-generated docs. Drawing inspiration from the thought-provoking piece written by Ferri-Benedetti entitled, What’s Wrong with AI-Generated Docs?, they’ll discuss the shortcomings, practical issues, and unexpected consequences of letting AI take the wheel in tech comm. Expect an honest discussion about accuracy, consistency, and whether generative AI is truly ready to replace human writers.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

It’s time we talk about customer experience. No, not the "smile-and-nod" kind you give your barista when they remember your name. I'm talking about the full-blown, Broadway-production-level experience companies try to orchestrate from first click to final sigh of satisfaction.

The understudy carrying the whole show? It's your content! Yes, your dusty ol' documentation, your FAQs, your user guides, and your troubleshooting trees that look like a caffeinated spider drew them.

A recent literature review by Muhammad Waqas and friends found that the customer experience (CX) is a sprawling, multidimensional beast. It's cognitive! It's emotional! It's social! It's got mood swings!

But while firms spend money on influencers and create branded experiences that sparkle like sequins on a Vegas showgirl, they continue to treat content like the guy who parks the cars.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The Role of Content (Or, Why Your PDF Might Matter More Than Your Logo)

Let's say you've bought a smart home device that promises to connect your toaster to the cloud. You're excited until you open the box, and the instructions look like they were translated by a toaster. That's not a fantastic product experience—it's a customer experience catastrophe.

Your sleek brand image didn't just take a hit—it got derailed by a help document that cheerfully opened with:

"To initiate, caress the power button until enlightenment occurs."

That's not user guidance, it's a riddle in a meditation app.

That's because content is the experience. It's the thing customers interact with when they need answers. It's their silent partner in crime during a midnight troubleshooting session. It's the difference between "I got this!" and "I'm returning this."

From Stimulus to Sob Story

The researchers behind the review unpacked over two decades of CX literature and came to a rather humbling realization: customers aren't passive recipients of brand fairy dust. They co-create meaning.

And what helps them do that?

Content.

Great content supports customer meaning-making. It lets people make sense of your product in their own messy, beautiful way. It helps them feel empowered and understood, not like they're being punished for wanting to use your product without a PhD in interface design.

So What Should CX Leaders Do?

First, stop treating content like a neglected appendix—tacked on at the end, outdated, and quietly sabotaging your carefully crafted customer journey. Invest in it, elevate it, and give it a seat at the strategy table—preferably one without a sticky note that says "intern."

• Make content findable, usable, and enjoyable. Not "compliance-checked" enjoyable. Enjoyable.

• Connect your technical writers, marketers, and CX folks so they can create a consistent, empathetic experience.

• Let your documentation reflect your brand values—not just in tone, but in structure, accessibility, and clarity.

• Track how your content affects satisfaction, loyalty, and that mysterious "didn't call support" metric no one wants to discuss.

Getting Documentation Right: Your Secret Weapon

If we're serious about delivering exceptional customer experiences, we must realize that content is not background scenery. It's the dialogue, the choreography, the plot twist. It's when your customer decides whether they're in a story worth sticking with, or one they can't wait to leave behind.

So go ahead. Put your documentation center stage. Give your product information a standing ovation. And maybe, just maybe, your customers will give you one, too.

Cue the curtain. 🤠

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

If you work in content marketing, you're probably exhausted. You're trying to create engaging content that builds trust, proves your brand knows what it's doing, gets people to stick around, and—if all goes well—eventually convinces someone to buy something.

All while battling the algorithm, the inbox, and the legal department.

But here's a little secret: Your technical documentation team—the folks who write the help articles no one claims to read? They're doing half your job already. And they're doing it well.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Let's Talk About Your Job

You want to:

• Attract people

• Keep them interested

• Make them trust you

• Get them to take action (buy, sign up, ask for a demo, tell a friend, whatever)

• Make your brand more visible

• Generate leads

• Keep customers happy

• Show up in search results without mortgaging your soul to Google

Now, let's talk about technical documentation. It does all of that, too.

Documentation Is Content Marketing For People Who Don't Want To Be Marketed To

Have you ever Googled a product question and landed on a help article that just got you?

It answered your question. It didn't try to upsell you. It was factual, well-organized, and surprisingly calming. That's documentation.

And guess what? That content often gets more traffic than your blog posts. Why? Because it's what people search for.

Think about it:

• "How do I connect X to Y?"

• "What's the difference between plan A and plan B?"

• "Can I cancel without crying?"

These are real questions asked by real humans with real intent. Your docs team is already writing answers. You, meanwhile, are trying to figure out how to turn a white paper into a TikTok.

Docs Build Trust The Slow, Boring Way (The Way That Works)

Documentation doesn't hype. It helps. And helping is what builds trust.

Marketing often tries to prove expertise by discussing innovation, leadership, and synergy. Docs prove expertise by telling you exactly which button to push and what happens if you push the wrong one. That's trust, baby.

Want to be seen as a thought leader? Link to the doc. Better yet, be the doc.

Self-Service = Self-Sell

Here's the part where it gets juicy: good documentation sells—but in disguise.

When someone finds your documentation, they're already halfway down the funnel. They're not wondering if they need a solution. They're wondering if your solution will work for them. And your help content is the difference between "add to cart" and "back to search."

SEO Without The Sadness

Docs rank. They do. Long-tail queries. Voice search. "I'm about to rage-quit" moments. Documentation is there, whispering calm instructions while your beautifully branded landing page flails in a sea of bounce rates.

Your tech docs team uses structured content. Metadata. Schema. All the things search engine optimization (SEO) pros love, are already baked in. You don't have to teach them—just copy their homework (with permission, please).

Retention, Loyalty, And The Magical Land Of "I Figured It Out Myself"

Want to improve customer retention? Make people feel smart. When they solve their own problems using your documentation, they feel powerful, capable, and like they've leveled up.

That feeling is sticky. It's also the foundation of customer loyalty.

You could send them a nurture email or give them a search bar that actually works. It's your choice.

So… Now What?

Talk to your documentation team. Bring coffee. Ask to see what content gets the most traffic. Ask the team what customers complain about. Ask them if they'll help you turn that dry but brilliant troubleshooting guide into something you can use upstream.

They've already built the library. You just need to put it in the front window.

Taking Advantage of Content Marketing’s Sensible Cousin

Content marketing doesn't have to do everything alone. Technical documentation is content marketing's shy, sensible cousin—who writes the manual, remembers a laundry list of style guide rules, and never brags about being right (even when they usually are).

It's time you collaborated.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

You know that feeling when you’ve just nailed a project, and someone walks in with a shiny new technology and says, “Hey, we’re doing this now!” Yeah, me too. It’s like cooking an elaborate meal only to find out your dinner guests have gone vegan. Suddenly, all your hard work feels less like a triumph and more like an impressive mistake.

Welcome to the Model Context Protocol (MCP) – the new “vegan dinner guest” of technical writing. Except this time, it’s worth your attention.

The MCP Reality Check

First, a little context (pun intended). The Model Context Protocol, or MCP, is not just some shiny new jargon. It’s an open standard developed by Anthropic. Essentially, it works to help AI models access and use data from different sources, like your carefully crafted technical documentation, in a way that doesn’t make them look like confused toddlers trying to operate a can opener.

Think of MCP as the universal translator between two Star Wars droids who don’t speak the same language. It helps AI tools (and droids) make sense of data from various sources (like R2-D2’s beeps and whistles) without sounding like a garbled mess.

Why Bother Learning About MCP?

Look, I get it. We’re all tired of being told we must learn yet another technology to stay relevant. It’s exhausting. You’ve just figured out how to make Markdown not look like the aftermath of a typewriter fight, and now this?

Here’s the thing: MCP is helpful.

It makes your content accessible to AI agents, which means your documentation can become part of a broader, interconnected ecosystem. Imagine your technical content being capable of smoothly integrating with AI-powered support systems, or your troubleshooting guide delivering value in real time by a virtual assistant that understands what it’s saying. That’s what MCP can do.

So, What Do You Need to Do?

Start by understanding how MCP works.

MCP works like a client-server model, where MCP clients (basically AI tools with big ambitions) chat with MCP servers (the cozy little spots where your data hangs out) using a common language. You’ve got to structure your content so both humans and machines can make sense of it – kind of like explaining calculus to both a fifth grader and a quantum physicist without losing either one in the process.

If you’re already using DITA XML, you’re not starting from scratch. You just need to think about how to make your structured content accessible to AI applications through MCP. It’s like taking your well-organized closet and labeling it so your slightly robotic roommate knows where to find the winter boots without mistaking them for garden tools.

Second, make friends with your developers. I know, they’re the ones who start conversations with, “So, I tried this cool new API…” and end with, “Oops, that was production data.” But you’ll need their help to make sure your content is MCP-friendly.

They’ll love that you care about interoperability, and you’ll love that they can handle the parts with code.

Finally, don’t let the fear of sounding like a tech bro stop you from getting involved. Sure, the name “Model Context Protocol” makes it sound like a secret society where people talk about their new AI overlords. But it’s just about making your content play nicely with modern technology.

DITA XML Writers: You’re Already Way Ahead

If you’re a technical writer producing content in DITA XML using a Component Content Management System (CCMS), give yourself a pat on the back. You’ve already done most of the heavy lifting to make your content interoperable. DITA XML’s structured format makes your documentation organized, modular, and easy for machines to process. You’ve essentially created a library of well-labeled boxes rather than one gigantic, unmarked suitcase of content.

The Model Context Protocol doesn’t change how you structure your content; it just adds a way for AI tools to find and use it efficiently. Think of it this way: You’ve built an IKEA bookshelf (complete with painfully clear DITA instructions), and MCP is the moving crew that knows how to take it apart and put it back together without accidentally turning it into a wobbly pile of wood.

Your Future as an MCP-Savvy Tech Writer

Learning MCP won’t just make you more valuable to your current gig; it will future-proof your skills. Imagine being the person in the room who knows how to make content interoperable between AI agents and websites.

You’ll be the tech writing equivalent of the person who can fix the office coffee machine – indispensable and (kind of) magical.

So, take a breath. MCP isn’t here to make your job harder. It’s here to make your content more useful. And that, my friend, is worth investing a little time in – if only to keep your documentation from looking like a toddler’s first art project when AI gets hold of it.

And if nothing else, when your boss asks, “Do you know about MCP?” you can confidently say, “Absolutely. It’s like the universal translator for AI and content. Piece of cake.”

You’ll look smart. You’ll sound like you know what you’re talking about. And who knows? You might make your documentation the gold standard of MCP readiness.

Think of the bragging rights when you tell your coworkers, “Yeah, I made my content AI-ready. No big deal.”

Now, go out there and make your content a little smarter. Your future self – and probably a few overworked AI agents – will thank you.

Imagine you're deep in the trenches of a new product release, translating cryptic developer notes into something a human can read. You're not just writing—you're decoding, interpreting, and transforming a tangle of jargon into explicit, understandable content, as if rescuing it from a keyboard having a nervous breakdown.

And then, the existential question hits: Could a robot do this?

According to the latest report by the Society for Human Resource Management (SHRM), business, finance, and HR jobs are the most likely to face the automation axe. But technical writers? We're not even on the list of highly at-risk jobs. (😮‍💨 Cue a cautious sigh of relief.)

Now, before you get too cozy, let's talk about why. AI tools can efficiently generate introductory text and automate formatting tasks, but they struggle to handle the nuance of technical writing. That's because technical writing involves more than slapping words on a page.

It requires its practitioners to wrangle complex concepts into something accurate and digestible, while anticipating the confusion of users who may not know an API from their elbow.

AI may help with routine tasks—generating tables, auto-summarizing, or suggesting simpler phrasing—but it just doesn't cut it when it comes to making sense of a feature held together by duct tape and hope or translating developer shorthand into something that won't make users cry.

So, while automation might take a bite out of some industries, technical writing—at least involving critical thinking, creativity, and plain old human intuition—still needs people. We're not just typists. We're sense-makers, storytellers, and those who can spot when a spec is about to send the entire project sideways.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

A special kind of rage is reserved for that moment when you need to speak to a human being, preferably one not controlled by an algorithm. The website has buried the phone number like the last bag of Halloween candy hidden from your spouse.

You scroll, click, scroll some more, and finally type something unhinged like, "WHY CAN'T I F*CKIN' SPEAK TO A HUMAN?" into the search bar. If you're lucky, this meltdown might summon a chatbot that understands English well enough to direct you to a telephone support number—the unholy grail of customer experience.

The Digital Gatekeepers

Let's address the obvious: Companies don't hide their contact information to foster a sense of adventure. They do it because they believe that if they make it hard enough, you'll either give up or figure it out alone, leaving them to sip their metaphorical piña coladas. At the same time, you wrestle with questions that are allegedly asked frequently and written by someone who has little understanding of your needs.

When the Human Touch Matters

The problem with this hide-and-seek game is that sometimes you must speak to someone. You know, someone who can hear the panic in your voice when you're trying to get a refund for the mystery charge that appeared on your bank statement. Sure, there's live chat, but let's be honest: if the conversation starts with "Hi! I'm Bot-73! How can I help you today?" you're already at a disadvantage.

Hiding Isn't Helping

Here's the kicker: Hiding contact options doesn't decrease support requests. It just annoys customers. It's like putting a maze between your front door and living room and expecting your guests to be thrilled with the obstacle course. People still call — except they are now angrier than they were initially.

You thought they were frustrated? Wait until they spend 40 minutes on hold only to hear they're in the wrong queue.

The Myth of the "Efficient" Support Model

There's a corporate logic to this madness, of course. Some companies think that forcing you to navigate a labyrinth of self-service options promotes "efficiency." What they're promoting is a spike in your blood pressure. Meanwhile, they boast about "streamlining" support, ignoring the screaming hordes begging for a human voice on the other end of the line.

Why Accessibility Matters

Imagine this: You're trying to report a fraudulent charge or cancel a subscription, and the only way to reach the company is by sending a carrier pigeon. Not much better than navigating an unmarked website. Real accessibility means offering contact options that respect people's varying needs and technological skills. Not everyone wants to type their life story into a chatbox or wait for an email response that arrives next Tuesday.

Customer Loyalty Is Fragile

Suppose you have tried to contact your cell phone provider and ended up in a queue that feels like a lifetime sentence. In that case, you know that hiding support options isn't just inconvenient—it's a betrayal. Customers remember. And when a competitor makes it easy to call them, they might take their business (and their simmering resentment) elsewhere.

The Gist: Make It Easy

Building lasting customer relationships is impossible when your organization treats customer support like a game of Whac-A-Mole.

Make it easy to call, chat, or email. Don't make us channel our inner detective just to find help. Nothing says "we value your business" quite like giving people the dignity of human connection—without the digital scavenger hunt.

The Content Wrangler is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

If you've ever worked in a content management system and thought, "Surely, there must be a better way," you're not alone. Michael Iantosca—technical documentation's equivalent of a weathered lighthouse keeper yelling at ships to avoid the rocks—thinks so, too.

In his blog post, "A Graph-Based Universal Component Content Management System," Iantosca proposes a better way to manage content: combine the strengths of component content systems with the intelligence of knowledge graphs.

It’s a practical fix for a very real problem. Here's the gist, minus the jargon and caffeine-fueled despair.

Related webinar: Graph-Driven RAG AI Powered by DITA

Old Software Systems, New Headaches

Most content management systems are fine if you treat content like Lego blocks—snap one to another and pretend that context doesn’t matter. Software developers built these systems to manage chunks of content (called components) but didn't design them to understand how those chunks relate to each other. They store content well—but they fail to understand how those chunks relate to each other.

Think of them like a warehouse that stores all your boxes but forgets where each one came from or how they fit together.

Worse still, the content gets flattened—like a soufflé in a marching band—losing the nuance, context, and relationships that make it useful. That lack of context and connection becomes problematic if you care about accuracy, personalization, and keeping your job.

That lack of context and connection becomes a problem if you care about accuracy, personalization, or delivering the right information to the right audience. These systems can store your content just fine. But they can’t reason with it, and they certainly can’t help AI do that either.

Knowledge Graphs (And Why You Should Care)

Iantosca introduces the idea of using knowledge graphs—a structure that captures not just content, but the rich web of relationships between content, metadata, audience needs, product versions, and more.

A knowledge graph doesn’t just store data. It understands the data. It maps relationships. It provides context. And that context makes it possible to deliver more accurate, personalized, and AI-ready content experiences.

Instead of asking writers to manually stitch together relevance, a graph-based system knows that if a topic is part of a larger process, and that process only applies to users in Canada on version 3.2, the system should deliver only that. No hacks. No guesswork.

What A “Universal” System Might Look Like

Iantosca’s ideal system treats your content the same whether you write it in DITA, HTML, Markdown, or scrawl it on a bar napkin (okay, maybe not the napkin 😆).

It doesn’t care about format—it cares about meaning. It uses ontologies (semantic rules and structures) to help define relationships and guide how the content connects across your entire ecosystem.

It enables:

• Format-agnostic content management

• Semantic search and smart reuse

• Personalized delivery at scale

• AI integration without hallucination

Instead of flattening content into disconnected blobs, this system turns your content into a dynamic, living network. It understands the why behind the what.

From Vision To Reality

This vision of a graph-based, format-agnostic content system isn’t just pie in the sky. At Avalara, Iantosca's team is already putting it into practice. They’re using DITA to structure content, an enterprise CCMS to manage it, and a knowledge graph to map relationships across it all.

The result? Content that’s easier to find, tailor, maintain, and serve—whether it’s for human readers or AI systems trying to answer customer questions without guessing.

Why It Matters —> Now

Technical content is no longer just documentation—it’s infrastructure. And if we expect that infrastructure to support everything from customer self-service to intelligent chatbots, we can’t afford systems that treat content like static blobs.

Iantosca’s argument is clear: we need a new foundation. One that understands how content works, how it connects, and how it serves real users in real contexts.

And we need to build it now—before the old systems start causing new problems we can’t patch with duct tape.

Read Iantosca’s paper. 🤠