Content Strategy & Systems Design

Mac McDonough

I'm a content strategist working in language systems — information architecture, content operations, and the structures that make complex products navigable. Writing is one part of that practice. At ProQuest I built UX documentation, internal knowledge bases, and content systems for cross-functional teams — including a SharePoint site, an Airtable database for accessibility workflows, and structured documentation that made complex internal processes legible to non-technical colleagues. These samples demonstrate how I think about language, structure, and audience across different content contexts — all oriented toward SaaS and tech products.

01
Database Design

Accessibility Workflow: Airtable Database

Context: Real database project at ProQuest — building a tracking system for accessibility work that spanned legal, licensing, and platform teams, replacing a spreadsheet and email setup that couldn't show shared status
Before

Making content accessible required three separate kinds of work on the same item: legal documentation, licensing and permissions clearance, and the platform work to actually implement the accessible format. Each ran on its own timeline. Status lived in a spreadsheet plus whatever email threads the spreadsheet couldn't capture.

No one could see another team's progress without asking directly. Teams duplicated effort on some items because they didn't know work was already underway. Other items stalled because no one knew whose turn it was to act next.

After

Three linked trackers — legal, licensing, platform — tied back to the same content item. Each task moved through the same stages: requested, in progress, review, done. Notes and question fields carried clarification and back-and-forth on individual records, so context stayed attached to the work instead of living in a separate email thread.

Multiple views let the team group and filter records by publisher, task type, or whatever else mattered for the work in front of them. If one publisher had several records that amounted to the same task, someone could see that, claim all of them, and do the work in one pass instead of picking it up piecemeal. I built the views and taught the team how to build their own.

The database went to the UX team for review before launch — its views, fields, and navigation held to the same accessibility standard the database existed to track.

What changed: The problem wasn't a missing spreadsheet column. It was that three teams were doing interdependent work with no shared view of each other's progress. A spreadsheet can hold a status. It can't link that status to two other teams' parallel work on the same item, or let someone regroup the same data by a different axis depending on what they need that day. The fix needed relationships between records and views flexible enough for people to organize their own work — not just a better spreadsheet.
View full case study: Accessibility Workflow Database →
02
Information Architecture

Rights & Bibliographic Governance: SharePoint Restructure

Context: Real information architecture project at ProQuest — rebuilding a cross-team SharePoint structure that had no taxonomy, just years of accumulated folders, to make rights and licensing records findable across departments
Before

Rights and bibliographic records lived across several drives. They were organized by whoever had touched them last. No team had ever designed a taxonomy. Structure accreted by habit, not decision.

Alliance managers were, about as often as not, tilling the ground for historical records: rights granted, royalties paid, licensing terms across different forms — PDF, platform-only, and so on. A request went from manager to team lead to team member to a search across multiple shared drives, none organized the same way. A routine request took about a week. Older or more obscure records took longer. Anyone on vacation added a full stop to the chain.

After

A single structure, organized around what requesters actually needed to know: rights status, licensing period, content type. Sensitive records stayed access-gated. Routing for those requests became defined instead of improvised — requests went to a named manager, team lead, or senior member. A request that took a week now took a few days.

What changed: The problem was never competing systems. It was the absence of one. Rights and bibliographic governance touched Alliance, Platform, Legal, and Publisher Relations. Each handoff exposed the same failure: information organized for no one in particular is information no one outside the original team can find. The fix wasn't reconciling formats. It was authoring a taxonomy for the first time, oriented to the question a requester was actually asking rather than to the sequence in which files had been saved.
View full case study: SharePoint Restructure →
03
Content Audit

Ubuntu Official Documentation Audit

Context: Real content audit of Ubuntu's official documentation — examining content currency, plain language, and information architecture across the upgrade docs and landing page
Ubuntu Official Documentation — Upgrade Docs & Landing Page

This audit examines the Ubuntu documentation landing page and upgrade documentation — a high-traffic entry point for users making consequential decisions about their systems. Five findings across three categories: content currency, plain language clarity, and information architecture.

  • Broken PDF links on the documentation landing page — immediate trust erosion at the entry point
  • Last-edited timestamps revealing years of neglect, including a 2019 date on upgrade-critical content
  • Upgrade instructions referencing end-of-life versions as current examples — actively misleading for users on current releases
  • Writing that assumes a technical baseline mismatched to the likely audience
  • Page naming that reflects internal taxonomy rather than user intent
View full audit: Ubuntu Official Documentation →
04
UX Copy Analysis

Thumbtack for Pros: Talking About Users vs. Talking to Them

Context: Speculative analysis of Thumbtack's pro-facing marketing page — identifying where the copy performs understanding of its users without demonstrating it, with a before/after rewrite of the core benefit section
Current copy
We limit competition for each job among pros.
Easily request and receive payments from customers.
We match you with jobs that exactly fit your preferences.
You get conversations with real customers. Not just views.
Proposed rewrite
Fair terms for you, the customer, and the platform. No surprises.
Leads for the kind of work you want. Consistently.
Quote, message, and close jobs from one place.
You get conversations with real customers. Not just views.
The diagnosis: The page talks about what pros get. It doesn't talk about what pros need to hear to sign up and stick around. Each proposed line answers a question a pro is asking: will the terms be fair, will I get real work, can I figure this thing out, can I get paid.
View full analysis: Thumbtack for Pros →
05
UX Copy Rewrite

LibreOffice First-Launch Experience

Context: Real first-launch onboarding dialog from LibreOffice — analyzing how copy and interface structure fail new users before they've written a single word, and rewriting for clarity and low friction
Before
Welcome to LibreOffice!

Options: Standard Toolbar / Tabbed / Single Toolbar / Sidebar / Tabbed Compact / Groupedbar Compact / Contextual Single

Preview: a three-item menu bar. Buttons: Apply / Next.
After
How do you want your toolbar to look?

Pick what feels right — you can change this any time in View → User Interface.

Classic / Tabbed / Minimal / Not sure (we'll use Classic for now)
What changed: Seven jargon-labeled options became four plain-language choices. Labels name what each option does for the user, not what it's called in the codebase. A reversibility signal ("you can change this any time") converts a commitment into a preference. A "not sure" option removes pressure entirely.
View full rewrite: LibreOffice Onboarding →
06
Research Brief

The Blank Slate Problem: Why Feature-Rich Tools Lose Users Before They Begin

Context: UX research brief analyzing Obsidian's first-launch experience and the broader pattern of onboarding failure in complex productivity software — drawing on direct observation and behavioral research
Research Brief — Obsidian Onboarding
Primary subject
Obsidian (v1.x)
Scope
First-launch UX + behavioral research
Output
Five actionable recommendations

The Problem

New users don't know what they want from a tool until they've had a chance to figure it out. Onboarding that front-loads configuration assumes the opposite — and loses users at the exact moment they most need to succeed. Obsidian is a powerful, genuinely useful note-taking tool. It's also a near-perfect case study in what happens when a product is built by and for people who already know what they want — and the onboarding reflects that.

Key findings

  • "Where do you want to save your vault?" — The first question Obsidian asks uses unexplained internal jargon before the user has any context for it. "Vault" means something specific here — and something different in other tools the user may have encountered.
  • The blank interface. After the vault question, the user lands in an empty workspace with no prompt and no suggested first action. Customization requires knowing what you want, which requires knowing what's possible — neither of which the new user has been given.
  • The graph view. One of Obsidian's most compelling features is meaningless until the user has a body of notes to connect. Showing it to a new user generates interest the product can't yet deliver on.
  • The folder problem. Basic organizational tasks require more time to figure out than the tasks themselves justify. The interface isn't broken — it's not explained.

Recommendations

  • Replace "vault" with plain language at first launch — introduce the concept after the user has a working environment
  • Give the blank slate a starting point — a single suggested first action closes the gap between landing and doing
  • Defer the graph view until the user has enough notes for it to mean something
  • Name the learning curve honestly — frame it as an investment, not a tax
  • Design the first-launch experience for the user who just installed it, not the one who's been using it for six months
Welcome the user you have. Not the one you wish had shown up.
View full brief: The Blank Slate Problem →
07
Style Guide

Scrivener: Voice, Structure & Content

Context: A content style guide prepared as a consultant review for Literature & Latte's Scrivener product site — addressing voice consistency, audience positioning, structural problems, and recommended rewrites across nine sections
Current — heading as label
Piece it Together

Switch instantly between editing your manuscript one section at a time and together as a whole...
Recommended — heading as proposition
Piece it Together: See the Forest, the Trees, and Everything in Between

Switch instantly between editing your writing one section at a time and together as a whole...
Core proposition: Every word should be load-bearing. If it can be removed without losing meaning, remove it. The guide covers nine areas — voice consistency, terminology, audience positioning, metaphor, headings, the SEO section, structure, and typography — each with findings and specific recommendations.
View full style guide: Scrivener →