10 Strategy · Multi-device · Cross-team · Localization

Multi-device
content strategy

Meta's wearables portfolio was expanding. Each product had its own content patterns, its own terminology, its own way of referring to devices. As users started owning multiple devices, the experience was going to fracture. I built the strategy to prevent that — before it was on anyone's roadmap.

1st
Multi-device strategy in AR wearables org
2
Product teams aligned
Org-wide
Product Marketing standards alignment
Meta-wide
Rollout planned

Not on anyone's roadmap.
But on mine.

Meta's wearables portfolio was expanding: smart glasses, a wrist wearable, and a shared companion app that needed to support both. Each product had its own design team, its own content patterns, and its own way of referring to devices, features, and settings. As users started owning multiple devices, the content experience was going to fracture.

There was no multi-device content strategy. Nobody had written one. It wasn't on anyone's roadmap. I had no dedicated PM support, no engineering resources, and no explicit mandate. I had to prove the value while building the thing.

The constraint
This work spanned multiple product teams that didn't share a reporting structure. The smart glasses team and the wrist wearable team had different priorities, different timelines, and different PMs. Localization added another layer — inconsistent source copy across devices meant the same concept could be translated differently depending on which product it appeared in.

Build alignment
before the crisis.

I proactively identified the problem and started building alignment before it became a crisis. I facilitated cross-team communication between smart glasses content designers, wrist wearable content designers, engineering, and localization to establish shared standards.

I created a unified strategy for how we refer to devices across the ecosystem — when to use the generic name, the branded name, or a custom/personalized name. I mapped where these references appeared (out-of-box experience, pairing, settings, notifications) and documented the rules for each context.

I worked with localization managers to identify where inconsistencies were already causing problems and built guidance to prevent future drift. I also introduced dynamic copy logic for personalized device names and mapped edge cases across out-of-box experience, pairing, and settings for i18n, privacy, and delight.

Two surface types.
One coherent system.

The core of the strategy was a two-surface model: global surfaces (screens shared across device types) and device-specific surfaces (screens created for a single device). The distinction determined when to use dynamic strings, when to use device-specific language, and when to use a generic fallback.

The overarching standard: always use device-specific language whenever possible. Because the phone is technically also a "device," vague references create confusion — especially in constellation and bundled device scenarios where multiple devices are discussed in the same step.

Framework Multi-device content framework: Global vs. device-specific surfaces Internal strategy doc · 2024
Global surfaces
Screens shared across more than one device type
  • Use dynamic strings to specify device type
  • Use "device/s" when device isn't yet known
Examples: Welcome screen, permissions pre-prompts, pairing, activation, OTA
Device-specific surfaces
Screens created for a single device type
  • Use "glasses," "watch," or "band" as device types
  • Don't use brand name unless differentiating models
Examples: Pairing troubleshooting, charging instructions, device-specific tutorials
Example dynamic string: "When setting up for the first time, {your glasses / watch / band} should automatically appear in this app."

SKU-forward naming.
Consistent across five devices.

As the device portfolio expanded to include Ray-Ban, Oakley, and wrist form factors, naming in device cards became a source of inconsistency. I led the cross-functional alignment on SKU-forward naming — a system where the product SKU name leads (e.g., Wayfarers XXXX, HSTN XXXX) rather than the brand name, with a 18-character maximum to prevent truncation across key surfaces.

The 18-character limit wasn't arbitrary. It was the maximum that avoided truncation across device cards, settings headers, toasts, and push notifications simultaneously. Some surfaces could handle more, but pushing the limit would reduce clarity or break layouts elsewhere.

Standard Device naming alignment Cross-functional decision · 2024
Device Format Logo
Ray-Ban glasses Wayfarers XXXX Meta | Ray-Ban
Oakley glasses HSTN XXXX Oakley | Meta
Display glasses Display XXXX Meta only
Wrist band Band XXXX Meta only
Watch Meta Watch Meta only
18-character max across all device names to prevent truncation in toasts and push notifications
Device cards showing SKU-forward naming across Ray-Ban Wayfarers, Oakley HSTN, Oakley Vanguard, Display glasses, and Band
Device cards in the companion app showing SKU-forward naming across five device types. Internal, Meta. Names shown as placeholders.

Dynamic strings.
A localization problem.

The elegant solution was dynamic strings — a single string template that could swap in the right device name at runtime. The less elegant reality: it didn't work. Not yet.

The variables would cause translation issues. Plural and singular forms, gendered language, and articles preceding device types translate differently across locales. A string like "Unable to find your {glasses|watch|band}" would require a split for grammatical accuracy in French, Spanish, and dozens of other languages. The API couldn't support it.

So we built a localizability check: 25 sample strings with variance in parameters, sent to every supported locale for feedback. The goal was to understand which version was most scalable and uncover any other flags before we committed to an engineering approach.

French example
Impossible de trouver [vos] lunettes.
Impossible de trouver [votre] montre.
The article changes based on grammatical gender — making a single dynamic string unworkable without locale-specific splits.

A name that's yours.
Not a serial code.

The SKU-forward naming system solved the default state — how devices appear before a user has done anything. But it didn't solve the harder problem: users who own multiple devices of the same style can't tell them apart. "Wayfarer 0011" and "Wayfarer 002F" are technically distinct. They're not humanly distinct.

The solution was personalized auto-naming: users would be able to assign a custom name (up to 18 characters) to each device during out-of-box experience and choose your own adventure onboarding. The custom name would then replace the default name across every surface where the device is referenced — action banners, home cards, notifications, settings cards, and within the Bluetooth OS settings. Future goal: the custom name persists in the companion app even after unpairing or logging out.

One distinction mattered enough to write down explicitly: the custom name and the device category are not interchangeable. "Ashlee's Wayfarers" is a custom name. "Glasses" is a device category. They serve different purposes in the content system — and conflating them would break the localization logic we'd already established for device-type references.

Standard Custom naming: how it works Internal strategy doc · 2024
Default state
Style + serial number format. Examples: Wayfarer 0011, Headliner 002F, Display 00A3. Aligns with native Bluetooth list name for cross-surface consistency.
Custom state
User-assigned name, up to 18 characters. Set during out-of-box experience or choose your own adventure onboarding. Replaces default name across all surfaces.
Where the custom name appears
Action banners · Home cards · Notifications · Settings cards · Bluetooth OS settings
Key distinction Custom name ≠ device category Content standard
Custom name
"Ashlee's Wayfarers"
User-assigned. Personal. Replaces default in UI. Not used in localized device-type references.
Device category
"glasses" / "watch" / "band"
System-defined. Used in instructional copy, error messages, and any string that must localize correctly. Not interchangeable with custom names.
Conflating these two would break the localization logic established for device-type references — which is why the distinction was written into the standard explicitly.

The first multi-device
content strategy in the org.

Strategy
Device reference guidelines

Covering generic, branded, and custom naming conventions — with rules for every surface where device references appear: out-of-box experience, pairing, settings, notifications.

Engineering
Dynamic copy logic

Logic for personalized device names using profile or account names — so the experience adapts to the user's context without requiring manual content updates.

Localization
Consistency across markets

Alignment documentation to ensure consistency across markets — preventing the same concept from being translated differently depending on which product it appeared in.

"This lays a proper foundation for building robust, scalable experiences. Doing this properly today reduces efforts down the road — one reason being less engineering tech debt."

— Nishil Shah, Engineer

Reduced tech debt.
Improved dogfooding.

The strategy was adopted across the wearables org and aligned with product marketing standards. It resolved existing content bugs in the companion app that were confusing dogfooders. The naming conventions improved onboarding efficiency and set a precedent for system-wide consistency.

Engineering acknowledged the long-term value directly — calling it a "proper foundation for building robust, scalable experiences" that reduces engineering tech debt down the road. The strategy was piloted in Wearables with a planned Meta-wide rollout.

"Despite not being a part of the H2 roadmap, Ashlee recognized the importance and prioritized efforts to develop guidance on multi-device content support. Her contributions enabled us to resolve the 'glasses' content bugs to improve the dogfooding experience. Longer term, this lays a proper foundation for building robust, scalable experiences. Doing this properly today reduces efforts down the road — one reason being less engineering tech debt."

— Nishil Shah, Engineer