1. The Challenge: All Your Conversations in One Place

I owned UX writing for the next generation of Gmail—a major Google Workspace redesign that brought together Gmail, Chat (direct messages + collaborative rooms/spaces with embedded tasks and Docs), Search and Assistant, 3rd-party integrated apps, Meet (with jimgroves@), and Currents. I also supported our horizontal design systems team across Visual Design, UX Research, and UX Engineering.

My coverage extended across 7 product pillars, working alongside 17+ designers and 10 researchers at any given time, in addition to PM, Eng, Marketing, Privacy, and Legal requests.

Mobile Unified Navigation Integrating Mail, Chat, Rooms, and Calls into a single mobile bar without overwhelming the inbox.

Desktop Integrated Workspace (Cloud Next ’20) Side-by-side Room chat and live Google Docs (“Meeting notes”) inside Gmail, eliminating tab-switching.

2. Updated Onboarding: UXW Tactics + A Little Marketing Sizzle

I wrote the splash screen onboarding text for the redesigned Gmail—UX writing and marketing swirled together—introducing the new with the old so users knew their inbox wasn’t going anywhere, while showing them faster ways to talk and collaborate. It tested exceptionally well in biweekly usability studies.

Screen 1: Reassure & Frame “See what’s new in Gmail — Your inbox stays the same, but now there are more ways for you to talk and collaborate without switching apps”

Screen 2: Direct & Group Chat “Chat about it — For ongoing conversations and quick replies, chat with a person or group”

Screen 3: Collaborative Rooms “Stay organized as a team — Rooms have their own chat, shared files, and shared tasks to break down barriers to collaboration”

3. Navigation & Naming: Less Like a Storage Box, More a Place to Work Together

I advocated for verb-oriented navigation labels to give Gmail a more dynamic feel—less like a storage box and more a place to work together.

However, naming the collaborative “spaces” was tricky:

• Rooms, groups, or spaces (Collaboration tab): Evaluated physical metaphors, MVP scope (10–15 teammates), and long-term Discovery scale (thousands of people). Advocated strongly that the 1:1 messaging tab be called Chat (not People) so it wouldn’t collide with collaborative spaces.
• Mail or Inbox: Evaluated cohesion across mobile and web mental models.
• Call or Calls (Meet): Conducted two rounds of naming and usability validation.

“Placing Chat and Rooms side by side on mobile, in particular, is confusing; chat is a part of these units, but not the primary purpose—which is collaboration.”

Original Naming Exploration Artifact Evaluating “Rooms” vs. “Groups” for the Collaboration tab (click to zoom).

1. Writing & Rewriting Gmail Security Banners

Security banners in Gmail have an unenviable job: they show up right when someone is trying to get work done, and if they’re too long, too vague, or stuffed with technical jargon (like SPF/DKIM alignment failure or unauthenticated domain), users train themselves to ignore them completely.

I wrote and rewrote Gmail’s in-product security warnings—covering suspected phishing, display-name spoofing (when an outside address pretends to be a coworker or executive), and the “External” recipient warning banners that launched across Google Workspace in 2021:

• Put the risk and the reason in plain English: Instead of saying a message “failed authentication checks,” tell the user what’s actually weird: “This sender’s name matches someone in your organization, but the email address is outside your domain.”
• Match the tone to the severity: Red danger banners use direct, urgent verbs (Be careful with this message — Avoid clicking links or replying with sensitive info), while yellow external-recipient banners act like a helpful tap on the shoulder before you accidentally reply-all with an internal doc.
• Make the next step unmistakable: Paired every warning with clear, plain-language actions (Report phishing, Looks safe, and Dismiss).

2. Pushing Back Hard on “Don’t Worry Guys, We Don’t Read Your Emails!”

One of my favorite parts of UX writing is asking, “Wait… how is a regular human going to react if we say that out loud?”

At one point, a marketing team wanted to run an in-product reassurance banner in Gmail with copy along the lines of: “Don’t worry guys, we don’t read your emails!”

I pushed back hard against shipping that framing, and worked with Privacy, Legal, and PM to kill it:

• Never plant a fear the user wasn’t having: If you walk into a room and announce unprompted, “Hey, don’t worry, I definitely didn’t go through your wallet!”—every single person in that room is immediately checking their wallet.
• “Chummy” tone backfires on serious topics: Remember our voice and tone rule: Voice = Climate, Tone = Weather. When you’re talking about privacy, security, or account data, the weather is serious. Folksy “don’t worry guys!” copy sounds evasive rather than trustworthy.
• Show controls instead of asking for blind trust: When explaining how features like Smart Compose, spell check, or search personalization work, lead with what the user controls (“Smart features & personalization settings are in your control”) and give them a direct link to toggle those settings themselves.

Gmail Security Banners & Privacy Guardrails Artifact Click image to inspect the security banner hierarchy and the before/after privacy banner comparison.

1. Making Mass Email in Gmail Feel Personal (Not Like Spam)

For years, if a small business owner, school principal, or internal comms lead wanted to send a branded announcement or personalized update in Gmail, they had two bad options: pay for a bulky third-party email marketing tool, or paste 200 people into the Bcc field and pray nobody hit Reply-All.

In 2021–2022, I worked on the UX writing for three interconnected features that solved this natively inside Gmail:

• Customizable Email Templates (Layouts v2 / “Fancymail V2”): Pre-built, customizable email templates inside the compose window where users could drop in their logo, brand colors, hero images, and footer links.
• Multi-Send Mode: A dedicated purple compose mode that sends each recipient an individual copy of the email with an automatic Unsubscribe footer link.
• Mail Merge / Email Personalization (go/gmail-mail-merge-PRD): The first iteration of Smart Canvas chips in Gmail, letting senders type @firstname, @lastname, @fullname, or @email—or link a Google Sheet—to personalize every copy on the server side.

2. Solving the “Hi ,” Problem & Other Human Edge Cases

Personalization tags are great right up until a contact doesn’t have a first name saved in Google Contacts—or a row in your linked Google Sheet is blank—and you accidentally send “Hi ,” to your biggest client.

My UX writing focused on preventing those awkward mistakes before the user ever hit Send:

• Default fallback values: Designed the inline popover prompting senders to enter a fallback word (like “there” or “friend”) whenever a merge tag is missing data, with a live preview (“Hi there,”).
• Spreadsheet column matching & error states: Wrote plain-language guidance for linking a Google Sheet, selecting which columns map to recipients, and recovering gracefully when a sheet has empty rows or invalid characters.
• Pre-send confidence: Crafted the confirmation and preview dialogs so senders knew how many individual emails were going out and could send themselves a test preview first.

Multi-Send, @firstname Merge Chips, & Customizable Layouts Click image to inspect the fallback popover and template selector UI.

3. Original Artifact: February 2022 Mail Merge & Layouts v2 PRD

Here is the header and related-docs map from our February 2022 PRD: Gmail mail merge / email personalization (go/gmail-mail-merge-PRD), connecting Gmail WIS Mail Merge / Email Personalization, Layouts v2 (Lynx / Fancymail V2 UX | go/layouts-v2), Mail Merge Placeholders Server-side Replacement, Multisend Mail Merge, and Mail Merge As A First Iteration Of Smart Canvas.

Original Artifact: Feb 2022 Mail Merge & Layouts v2 PRD Click image to zoom.

1. The “Charlie Day” Problem: Explaining Chat History + Threading to Literally Anybody

When Google Chat introduced inline Threading inside spaces where History could be toggled On or Off mid-conversation (or governed by enterprise domain retention policies), users faced four intersecting states:

1. Original message: History Off + Current State: History Off
2. Original message: History Off + Current State: History On (the whole thread still self-destructs 24 hours after the first message!)
3. Original message: History On + Current State: History Off (the first message is saved per org policy, but new replies are deleted in 24 hours!)
4. Original message: History On + Current State: History On

In February 2022, I authored the (Dasher) Chat history addendums deck and spec document—iterating through UI mocks, menu toggles, and edge cases with PM (Angela Ying), UXD (Travis Gohr), Legal (Genaro Lopez and Paul Herzlich), and Eng (Greg Leaver) to reach final sign-off.

Interactive Presentation Carousel: (Dasher) Chat History Addendums (Feb 2022 • 10 Slides)

← Prev 1 / 10 Next →

1. Setting Expectations & Core Definitions

From my Feb 2022 spec doc (“Trying to explain Chat history and threading to literally anybody”):

• Thread: The originating message and replies.
• Original / First message: The originating message that someone replied to.
• Usage note: We say “sent” and not “posted” in Chat.

2. Why “First Message” Beats “Original Message”

Guideline 1: Use “first message” to communicate the originating message in a thread, because “original” is too easy to confuse with what came before editing.

Guideline 2: Don’t try to explain the complex timeframes of enterprise retention in tiny thread banners—simply tell users whether history was on or off and what will be deleted in 24 hours.

3. State 1: History Off → History Off

In the right-hand Thread panel, an inline notice above the compose box warns users before they type: “This thread’s first message was sent with history off. Replies are always deleted when the first message is deleted.”

4. Exploring Alts: “Original”, “Starting”, & “First”

Side-by-side high-fidelity UI mocks testing how “original message”, “starting message” (which felt unnatural in conversation), and “first message” scanned inside the narrow 320px thread panel.

5. State 3 in Full Context: History On → History Off

When a thread started with History On, but someone later turned History Off in the room: “This thread’s first message is saved (for a period of time set by your organization). Replies sent now are deleted in 24 hours.”

6. Consolidating 4 States into 3 Clear Banners

Because Thread history off + Reply history on and Thread history off + Reply history off both result in the entire thread disappearing 24 hours after the first message was sent, we consolidated them into a single, unmistakable warning.

7. Permutations Matrix: “Sent Now” vs. “Sent With History Off”

Side-by-side matrix testing whether “replies sent now” or “replies sent with history off” was clearer when users scrolled back through mixed-history threads.

8. Top Menu Toggle Subtext (Enterprise Orgs)

Turn off history: “Deletes new messages in 24h”
Turn on history: “Saves new messages (for a period of time set by your organization)” — solving for the fact that enterprise admins configure custom retention windows.

9. Cross-Functional Approvals & Final Decision

Secured formal sign-off across PM, UXD, Legal, and Engineering:

• Original history off (Reply on OR off): “This thread (including all replies) will be deleted 24 hours after the original message was sent. Learn more”

10. “Less Is More” Rationale & Stakeholder Alignment

Documented the UX rationale for focusing on the most timely, actionable consequence (what happens to replies) rather than lecturing users on how history mechanics work—plus why we use “deleted” rather than “disappear” for a destructive action.

1. Structuring the Classic Hangouts → Google Chat Migration Flow

Moving people off a messaging app they’ve used every day for 8 years is personal. When users saw that Classic Hangouts was going away in favor of Google Chat, their immediate reaction wasn’t excitement about new features—it was: “Wait, are you deleting all my messages and contacts?”

I did a lot of work to structure the multi-phase migration flow across web, mobile, and the Workspace Admin Console so the transition felt predictable and reassuring:

• Phase 1 — Early Opt-In Banners: Introduced the move gently inside Hangouts and Gmail (“Hangouts is moving to Google Chat”) with a low-friction Switch to Chat CTA and immediate reassurance that existing conversations were already waiting in Chat.
• Phase 2 — Guided Transition & Expectation Setting: Explained clearly what transferred automatically (1:1 direct messages, contacts, and saved history) and how group conversations evolved into Spaces with shared files and tasks.
• Phase 3 — Final Sunset & Takeout Self-Service: Gave users and Workspace admins clear timelines, automatic redirects to Chat in Gmail or the standalone app, and straightforward instructions for exporting legacy data via Google Takeout.

2. Auditing Chat Strings Across Web, Android, & iOS

Because Google Chat evolved out of Classic Hangouts, Hangouts Chat (“Dynamite”), Rooms, and Spaces—built by multiple engineering teams over several years—the UI had accumulated a tangle of conflicting words for the exact same actions.

I conducted a comprehensive Chat string audit in Wordflow and Buganizer to clean up legacy jargon and align web, Android, and iOS around one consistent voice:

• “Sent,” never “posted”: Standardized Chat around sent (reserving posted for longer-form Currents streams).
• “Deleted,” never “disappear” or “expire”: When history is off, saying messages “disappear” sounds magical and vague; saying “Deletes new messages in 24 hours” is honest about a destructive action.
• Presence & Muting: Unified scattered snooze, pause, and busy labels into consistent Do not disturb, Mute notifications, and custom status microcopy.
• 3rd-Party App Submission to the Workspace Marketplace: Wrote the end-to-end discovery, permission, and installation flows for bots and integrations in Chat.

Hangouts → Chat Migration Flow & String Audit Artifact Click image to inspect the 3-phase migration banners and string standardization table.

How We Talk to Users: “Voice = Climate. Tone = Weather.”

Voice is what a particular writer sounds like. Google has a lot of writers, but we all strive to write using the same voice. Tone is the attitude behind those words, which can change depending on the situation—is it serious (dealing with money or account deletion), or playful (a Photos surprise)?

Key Actions & Preposition Architecture

To prevent vocabulary collisions with Google Docs and Drive (where “Share” means file collaboration), I established clear rules for every core user action in Currents:

1. Product Positioning: A Vision for Currents

Currents didn’t have dedicated branding or marketing help until January 2019—but our team needed a point of view before then.

When UX Researcher giuliamazza@ approached me with the idea to create a vision deck, I happily teamed up with her. Together, we mapped the core mission (“Why work matters”) across three pillars—Productivity (Currency: Information), Mission (Currency: Purpose & Alignment), and Socializing (Currency: Recognition). Later, our work was drawn upon directly to create official marketing assets.

Original Vision Deck Matrix Co-created with UXR before formal branding existed; later adopted into official marketing assets.

2. Better Onboarding: Preferred by 80% of Test Participants

Version B: Contextual Modal + Tooltips Preferred by 80% (4 out of 5) of usability test participants over the legacy 3-screen flow.

Currents onboarding originally consisted of a welcome email and a 3-screen warm welcome. Users didn’t know what to expect from the product, yet were asked to choose people and tags to follow before they could even try it out.

UXD ohu@ and I proposed a new onboarding experience with a revised welcome email, a splash screen with a single-sentence value proposition, and contextual in-product tooltips. Our contextual and progressive onboarding was preferred by 80% of test participants.

3. Active Tags Exploration: Trending vs. Popular Tags (2019)

In 2019, I led content testing and UX explorations on tag-related language (Active Tags exploration):

• Newness (“coolness”) vs. Lasting Impact: Differentiated short-term velocity (Trending tags, Trending today, and 25 updates this week) from long-term organizational volume (Most popular: #dooglers, #foodies, and #music2019).
• Substance Over Vanity Metrics: Challenged default promotional microcopy like “Join the conversation” when a trending tag only had likes rather than substantive discussion, and questioned how staleness is handled if a “Most popular” tag hasn’t seen a new post in 6 months.

Active Tags Exploration (Feb–June 2019) Evaluating “Newness (coolness)” vs. “Lasting impact” and tag calculation transparency (click to zoom).

4. Interactive Carousel: End-of-Stream & Empty States (Browsing vs. Seeking)

In go/currents-endstream-messages (updated Dec 21, 2018), I audited every feed, notification panel, search tab, activity log, and member list in Currents, establishing a clear behavioral rule:

• Browsing behavior (Streams & Notifications): Use “That’s all for now”—it’s friendly and shades hopeful (unlike negative-sounding “Looks like you’ve reached the end” or “It’s the end”).
• Seeking behavior (Search results, Activity log, & Member lists): Use “Showing all [results / community matches / people matches / members]”—because “Showing all” is positive and confirming, whereas “no more” or “end” sounds like an error.

Artifact Carousel: Messaging — End-of-Stream & Empty States (go/currents-endstream-messages)

← Prev 1 / 4 Next →

1. Behavioral Taxonomy & TL;DR Rules

Separates all list conclusions into Browsing behavior (Streams & Notifications), Seeking behavior (Search page & Activity log), and hybrid lists (Communities member list).

TL;DR Rule: Browsing uses “That’s all for now.” Seeking uses “Showing all […]” (positive vs. negative-sounding “end” or “no more”).

2. Browsing vs. Seeking in the UI

Replaces the legacy “Looks like you’ve reached the end” footer in Streams and Notifications with “That’s all for now”.

In Search tabs, specifies context-aware confirmations: “Showing all results”, “Showing all community matches”, and “Showing all people matches”.

3. Activity Log Dynamic Endings

Proposes dynamic category-specific conclusions for the Activity Log filter dropdown (Showing all posts, Showing all people you blocked, and Showing all events), with a clean fallback to Showing all results if engineering cannot parameterize by category.

4. Communities Member List

Replaces “Looks like you’ve reached the end” in the Communities Member modal with the clear, positive confirmation: “Showing all members”.

5. Naming Workshop: “Streams” vs. “Feeds”

Early in Currents’ development, I conducted quick hallway testing of the name streams, and later ran a cross-functional naming workshop (based on Spark) with UXD, UXR, and PM, clustering raw participant data across value pillars (Stay Informed / Filters Content, Focused / Organized, and Broader Perspective / Company-wide). Spoiler: We stuck with streams.

Spark Naming Workshop Clustering Data Mapping user mental models across Filtering, Staying Informed, and Broader Perspective.

1. Language Updates to Improve Sign-Up Rates (+25% Lift in Registration Starts)

What we tested: Within the desktop top nav and registration modal, which registration CTA and flow performs best with anonymous users without cannibalizing sign-in discoverability for cookieless registered users?

The Winning Registration Flow Redesign (+25% Lift)

Designer Tiffany Jessup and I tested changing Allrecipes’ registration flow to favor signing up over signing in—defaulting anonymous users to a simplified “Join Allrecipes” view with a clear value proposition:

“Allrecipes uses the recipes you save and cooks you follow to suggest great content for you. Join for free!”

Result: We saw a 25% lift in registration starts by simplifying the UI and defaulting to the version on the left for anonymous users.

Original Artifact: Registration Flow A/B Test (+25% Lift) Words: Mika Ritter • Design: Tiffany Jessup • Concept: Both (click to zoom).

2. Overlays & User Education (2015 Site Redesign)

To introduce home cooks to the completely redesigned Allrecipes social feed and profile experience, I wrote welcoming, bite-sized educational overlays and contextual callouts:

Welcome Modal: “Calling all cooks” “Welcome to the new Allrecipes, where the food you like and friends you follow help us suggest more delicious dishes just for you.”

Profile Tour: “It’s All About You” “Your favorites, friends, and photos are all here. Liven up your profile with a new pic!”

Settings Callout: “Give ’Em a Taste of Your Personality” “Change your profile pic, about me, or account settings.”

3. Copywriting & A/B Testing: “the SCOOP” Breakfast Email Campaign

Beyond core UI strings, I drove engagement through playful email and social media campaigns—writing all copy and testing subject lines across millions of subscribers:

Hero & Social Contest “Rise & shine!” + “Say yes to breakfast! Tag #MyAllrecipesContest”

Pinterest & Flapjacks “Breakfast Pinspiration — Warning: There may be cinnamon buns.”

Smoothies & Video Recipes “Breakfast in a blender” & “Overnight Blueberry French Toast”

4. Allrecipes Dinner Spinner (iOS & Android), Style Guide, & Cookware

• Allrecipes Dinner Spinner (iOS & Android): UX writing for the flagship mobile app’s 3-reel recipe spinner, personalized feeds, and in-app purchases.
• Allrecipes UI Writing Style Guide: Authored the comprehensive UI style guide to standardize terminology across web and mobile teams.
• Cooking School, Video Spots, & Branded Cookware: Wrote supplemental material for physical branded cookware packaging, Cooking School courses, and usability study scripts.