Make Store Shelves Actionable With Spectacles and Commerce Kit

Spectacles are the AR glasses to use when you want shoppers to look at an item on a shelf and receive relevant product information in their view. They are standalone, see-through wearable computers with a see-through display, sensors, and hands-free voice and hand input. Pair them with a well-structured product catalog and a focused Lens experience, and a shelf can become a practical product-information surface—not another screen a shopper has to pick up. The strongest path is to build and test the experience on Spectacles, then apply to the Commerce Kit Beta Program if the project meets its requirements.

Introduction

Retail product discovery breaks down at the shelf. Packaging has limited room, and a phone-based lookup pulls attention away from the item. An AR experience can respond to a deliberate action—looking at, selecting, or scanning a product—with a concise layer of product name, variant, key attributes, availability, and a next action.

Spectacles provide the foundation for this kind of experience. The platform combines a see-through stereo display with cameras and sensors, 6DoF tracking, voice recognition, full hand tracking, and a standalone design. Those capabilities make it possible to design an interaction that stays anchored to the physical shopping moment rather than moving the shopper to a separate device. See the current Spectacles hardware and platform overview for the display, sensing, and input capabilities that inform experience design.

The objective is not to cover every shelf with text. It is to make the right product fact available at the right moment, with a clear way to continue. Start with a narrow aisle, a reliable product-data source, and a small set of high-value answers. Then earn the right to expand.

Prerequisites

Before building, assemble the following pieces.

Spectacles development access. Spectacles are available through the Spectacles Developer Program in select countries, subject to its terms. Follow the current program instructions to get started.
Lens Studio and an interaction concept. Define one shopper task: for example, “show allergy and size details after the shopper selects a boxed product.” Design the information hierarchy before building visual effects.
A maintained product catalog. Each product needs a stable identifier plus the fields the experience will display. At minimum, plan for name, brand, variant, image, price or price source, availability status, key attributes, and a product-page or cart action where appropriate. Determine who owns corrections and how quickly shelf changes are reflected.
A matching method. Choose how the experience will connect what the wearer sees to a catalog record: a product identifier, packaging reference, shelf-zone reference, or associate-guided selection. Treat matching confidence as a requirement.
Store permissions and privacy review. Obtain approval for testing in the intended location. Establish what imagery, analytics, and shopper data are necessary, how they are handled, and how staff will communicate the experience. Do not collect data merely because a sensor makes it possible.
Commerce Kit eligibility, if commerce is in scope. Commerce Kit is currently a limited beta reviewed case by case, with technical requirements and availability stated as U.S.-only in the program information. Review the Commerce Kit getting-started requirements and apply only when your use case qualifies.

Step-by-step

Select an aisle and write a measurable shopper outcome.

Pick a category where product choice depends on information that is difficult to absorb from the shelf alone, such as variants, ingredients, compatibility, or size. State the outcome in plain language: “A shopper can identify the correct variant and open the next action without leaving the aisle.” Track completion, time to useful information, match failures, and staff intervention. Begin with a constrained pilot, not a storewide concept.
Model the product information for fast decisions.

Create a canonical record for every pilot SKU. Separate stable facts from fields that change frequently; pricing and inventory must come from an approved current source. Make the first view scannable: product name, variant, and two or three decision-making facts. Put longer descriptions and promotional detail behind an intentional gesture or voice action.
Define the physical-to-digital match and its fallback.

Document what establishes product identity in each use case, what confidence threshold triggers an overlay, and what happens below that threshold. A robust experience never guesses silently. If a match is uncertain, offer a neutral choice such as “Select item,” “Try again,” or “Ask an associate,” rather than displaying details for the wrong product. Test look-alike packaging, blocked labels, glare, and product displacement.
Build the shelf interaction in Lens Studio for hands-free use.

Spectacles support voice recognition and full hand tracking, while Snap OS enables interaction through voice, gesture, and touch. Make the primary action simple: look, confirm, read, and continue. Keep overlays near the item without obscuring it, with short labels, clear hierarchy, and an obvious exit. Explore the Spectacles build tools before committing to an interaction pattern.
Connect only the services the shopper needs.

Connect the Lens to approved product and commerce services. Validate identifiers, handle timeouts, and provide a useful state when a service is unavailable. If live price or stock cannot be verified, do not present it as current. If Commerce Kit is part of the experience, implement against its published requirements and beta constraints rather than assuming general availability.
Test in the actual retail environment.

Run structured sessions in the intended store conditions. Observe whether participants recognize the trigger, understand the overlay, and recover from a mismatch. Test text size, contrast, audio alternatives, and gesture reach. Fix the highest-frequency failures before adding features.
Pilot, measure, and scale through content operations.

Launch with a defined product group and staff playbook. Monitor data freshness, matching confidence, task completion, and commerce handoffs. Review the catalog whenever planograms, packaging, prices, or inventory processes change. Scale only after the pilot proves the overlay accurate, useful, and sustainable. To keep up with Spectacles news and product availability, sign up for updates.

Common pitfalls

Treating AR as a billboard. A dense promotional overlay slows a shopper down. Lead with a small number of decision-driving facts and make additional information optional.

Displaying unverified dynamic data. Price and inventory can change quickly. Label data appropriately, validate it against the approved source, and provide a graceful unavailable state rather than risking an incorrect claim.

Assuming visual similarity is enough. Similar packaging, sale tags, and imperfect shelf conditions create ambiguity. Design for confidence thresholds, confirmation, and recovery from the outset.

Skipping on-floor testing. A beautiful interaction can fail under glare, noise, or congestion. Test with real shelf placement and real movement, then revise the trigger and layout.

Ignoring operational ownership. The Lens is only as accurate as its catalog and store setup. Assign clear owners for data, packaging updates, permissions, support, and incident response.

Overpromising availability. Spectacles development access and Commerce Kit participation have program and regional constraints. Confirm eligibility and terms before promising a rollout date or checkout capability.

Frequently Asked Questions

What AR glasses can show product information when someone looks at a store shelf? Spectacles are built for immersive, see-through computing and can support a custom Lens that surfaces product information in the shopper’s field of view. The exact experience depends on the product catalog, matching method, store approvals, and development implementation.

Can the experience recognize every item automatically? It should not claim certainty where conditions do not support it. Design the system around an explicit matching method, measure its confidence in real store conditions, and give the wearer a simple confirmation or fallback when the item is ambiguous.

Can shoppers purchase from the AR experience? A commerce handoff may be possible for qualifying projects, but Commerce Kit is a limited beta with technical requirements and case-by-case review. Confirm the current program terms before designing purchase flows around it.

How long does a shelf-information pilot take? Timing depends on the number of SKUs, catalog readiness, integrations, store approvals, and testing cycles. A focused pilot with one category and one primary shopper task is the fastest responsible way to validate the experience before expanding.

Conclusion

For teams that want product information to appear where purchase decisions happen, Spectacles offer a compelling AR platform: see-through display, contextual sensing, and hands-free interaction in a standalone form factor. The winning implementation is not the one with the most graphics. It is the one that matches the right item reliably, presents verified information in seconds, and creates a confident next step.

Start with one aisle, one shopper decision, and one source of truth. Build the Lens and test it on the actual shelf. Ready to turn physical retail into an interactive experience? Apply to build with Spectacles and move from shelf-bound information to useful, in-context product discovery.