Introduction
Augmented reality (AR) adds digital elements to a view of the physical world, while virtual reality (VR) presents an immersive digital environment. In marketing, either can help demonstrate a product or experience, but usefulness depends on the customer need, device support and execution. Neither technology guarantees engagement or sales. This guide explains how to plan a limited pilot and measure whether it is worthwhile.
The Process
Augmented Reality (AR) in Marketing
AR overlays digital information onto the real world, providing interactive and immersive experiences. In marketing, AR can be used in various ways:
- Product Visualization: AR allows customers to visualize products in their own environment before making a purchase. For example, furniture retailers use AR to show how a piece of furniture would look in a customer’s home.
- Interactive Advertising: AR can make advertisements interactive, engaging customers in a unique and memorable way. For example, movie posters can come to life when viewed through an AR app.
Virtual Reality (VR) in Marketing
VR provides a fully immersive experience, transporting users to a different reality. In marketing, VR can be used to:
- Virtual Storefronts: Retailers can create virtual storefronts, allowing customers to browse and shop in a virtual environment.
- Product Demonstrations: Companies can use VR to provide immersive product demonstrations, allowing customers to experience a product virtually.
How to plan a useful AR or VR campaign
Start with a customer question that a normal photograph or video does not answer well. A furniture shop might test whether shoppers can judge a chair’s scale in their room. This is a proposed use case, not evidence that AR increases sales. VR may suit a guided demonstration when customers can access the required headset; it is not a necessary addition to every campaign.
1. Define the decision and baseline. Choose one product range and one desired action, such as an enquiry or purchase. Record the existing conversion rate, device mix, loading time and common support questions before building the experience.
2. Prepare accurate assets. Obtain permission to use the models, textures and branding. Check dimensions, colours, variants and scale against the physical product. Budget for modelling, optimisation, integration, testing, hosting and ongoing updates. Obtain a scoped quote rather than assuming a universal project price. A visual preview should not promise an exact colour, physical fit or product performance.
3. Choose a supported delivery route. As checked on 31 October 2026, Google’s Scene Viewer supports viewing web-hosted 3D models in AR on supported Android devices and provides fallback options. It is an implementation example, not a universal device guarantee. Confirm the current platform requirements and test on the actual phones, browsers or headsets used by your audience.
4. Keep a usable alternative. Show conventional images, dimensions, text descriptions and a clear purchase or enquiry route when immersive viewing is unavailable or declined. Test keyboard navigation, screen-reader access, readable controls and captions where relevant. For headset demonstrations, offer a seated or non-headset alternative and let participants stop immediately if uncomfortable.
5. Review privacy and release quality. Explain why camera access is requested before asking for it. Minimise collection and retention; do not record room imagery or movement data by default. Review vendor data handling and applicable consent requirements. Test denied permissions, slow connections, missing assets, scale errors and checkout continuity before release.
Measure impact without assuming a sales lift
Track how many eligible visitors load the experience successfully, start it, complete the intended interaction and continue to the product action. Log failures by device. Engagement time alone does not establish commercial value: a long session may indicate confusion.
Where feasible, randomly offer comparable visitors the standard page or the page with the experience. Define the primary outcome and test duration in advance, retain unsuccessful loads in the relevant analysis, and report uncertainty. Comparing only people who choose AR with people who do not is vulnerable to selection bias. Changes in offers, traffic or seasonality can also affect results.
Assess incremental contribution after production, software, support and maintenance costs. Include returns and customer complaints when evaluating product visualisation. A pilot that reduces uncertainty for shoppers may still fail to pay for itself; retain, improve or stop it according to the measured outcome.
A small pilot before a large rollout
Illustrative plan: test one chair model on its existing product page, keep the original photography and dimensions, provide a non-AR fallback, and assign one owner to check scale, loading and feedback. Set the budget and evaluation criteria before development. This is a planning example, not a published campaign result or a promise of higher sales.
