Spatial Computing and Marketing: What Is Actually Worth Doing

When Apple announced its headset I wrote that spatial computing could be the next major shift in marketing.

The device shipped, the reviews were mixed, the price was high, and the install base stayed small. The shift did not arrive on the schedule anybody predicted.

That does not make the underlying question uninteresting. It changes the answer to what you should actually do about it.

What Spatial Computing Actually Is

Computing that understands the physical space around you and places digital content in it.

That covers headsets and it also covers phone-based augmented reality, which has vastly more users and gets far less attention.

Separating those two matters, because the marketing opportunity is almost entirely in the second one.

Why the Headset Moment Did Not Land

Price. A device costing several thousand dollars does not build a consumer install base, and marketing requires an install base.

No repeat use case. Reviewers consistently described an impressive demo followed by declining use. Without a daily reason to put it on, the device becomes occasional.

Comfort ceiling. Same physics problem as every headset. Weight and heat limit sessions to a couple of hours.

Content economics. Producing high quality spatial content is expensive and the audience is tiny. That is a bad ratio and nobody solved it.

Where Spatial Marketing Actually Works Today

Phone-based try-on and placement. This is the real, working, boring version.

Furniture placed in your room through a phone camera. Glasses on your face. Paint on your wall. These reduce return rates measurably and they run on devices everybody already owns.

If you sell something physical where fit or appearance is the purchase barrier, this is worth building. The technology is mature, the tooling is accessible, and the ROI is measurable.

Location-based AR for retail and events. Narrow but effective when the physical location is already a destination.

Product visualization in B2B sales. Showing a machine at scale in a buyer's actual facility, from a tablet. This shortens industrial sales cycles and almost nobody does it.

What Is Not Worth Doing

Building a branded headset app. The audience does not exist and the production cost is high.

Virtual storefronts. Tried repeatedly across multiple platforms and consistently produce visits with no commerce.

Anything that requires the customer to buy hardware to engage with your marketing. That has never worked in any medium.

The One Thing I Would Build

If you sell a physical product with a visual or fit component, build phone-based AR placement or try-on.

It works on hardware people have, it addresses a genuine purchase objection, and the measurement is clean. You can attribute reduced returns directly.

Everything else in this category is a bet on an install base that may or may not arrive.

How to Think About the Next Version

When the hardware improves and the price drops, the marketing opportunity will arrive through whatever becomes the everyday device, which may or may not be a headset.

The preparation that transfers is having accurate 3D assets of your products. That asset is useful today for AR placement and it is the input to whatever comes next.

That is a genuinely low regret investment. Building an experience for a specific platform is not.

The broader lesson about evaluating emerging platforms is in betting on early tech, and the retrospective on the last cycle is in what happened to the metaverse.

For the technical side, Apple's developer documentation on spatial experiences and the equivalent WebXR standards work are the places to watch for when this actually becomes accessible.

What Should You Build Now to Be Ready for Spatial Computing?

Accurate 3D models of your products, and nothing else. That asset is useful today for phone-based try-on and placement, it shortens B2B sales conversations, and it is the input to whatever the next device turns out to be. Everything else is a bet on a specific platform.

The reason this is the low regret choice is that the models outlive the platform. A headset app built for one ecosystem is worthless if that ecosystem loses, which is the normal outcome. A good 3D model of your product works in a web viewer, a phone AR experience, a configurator, and a future device nobody has announced.

What good means here: dimensionally accurate, properly textured, and optimized for real-time rendering rather than exported straight from CAD. That last point matters, because untouched engineering files are usually far too heavy to use anywhere customer-facing.

The second thing worth doing costs nothing: keep your product data structured and current. Dimensions, materials, finishes, variants. Every spatial experience is driven by that data and most companies have it scattered across spreadsheets and PDFs.

What I would not build: a branded headset application, a virtual storefront, or anything requiring a customer to buy hardware. All three have been tried repeatedly across multiple platform cycles and none has produced a durable commercial result for an ordinary business.

The preparation that transfers across platform cycles is always the underlying asset rather than the experience built on top of it. Accurate models and structured product data survive whichever device wins. A branded application built for one ecosystem does not, and the normal outcome for any given ecosystem is that it loses.

Have you built anything in AR that produced a measurable result? The try-on examples I have seen are the only ones with real numbers.