Can virtual try-on increase conversion in eCommerce?
it can, but not automatically or uniformly. Most published studies and cases point in the same direction (better conversion and fewer returns among users who use the try-on), but almost all of them come from suppliers and the figures vary widely by category, sector and implementation. The most reliable approach is to measure the effect in your own store with a pilot.
What the evidence says, with caution
If you look for data on virtual try-on and conversion, you will find very striking figures (double-digit increases in conversion, notable reductions in returns). They should be read with judgement:
- Many come from suppliers of the technology, who have a commercial interest.
- They are attributed to studies that often cannot be traced to the original source.
- They compare users who use the try-on with those who do not, but the groups are different: someone who uses the try-on is usually more inclined to buy (selection bias).
- Results vary by category (the effect is usually greater for glasses or accessories than for basic products) and by the quality of the implementation.
The honest conclusion is that the direction of the effect is positive in many cases, but the magnitude cannot be taken for granted.
Why it can improve conversion
- It reduces uncertainty: the customer sees how it looks on them, or how the product looks.
- It increases engagement: more time with the product.
- It makes the decision easier when there are many variants.
- It builds confidence for purchases that would normally be made in store.
- It produces shareable content: screenshots that are sent to friends and generate feedback.
Why it can reduce returns
If the customer has a more realistic expectation of how the product looks, they are less likely to return it because "it was not as I thought". The effect is clearer for aesthetic reasons than for size reasons.
When it is more likely to work
| Factor | Greater potential | Lower potential |
|---|---|---|
| Category | Glasses, jewellery, watches, make-up, hair colour | Basic or highly commoditised products |
| Order value | Medium-high | Very low |
| Purchase doubts | High (aesthetics, fit) | Low |
| Technology quality | High | Poor (delays, failures, unrealistic look) |
| Visibility | Prominent button | Hidden |
| Mobile experience | Smooth | Heavy or with errors |
How to measure it well
The most robust way to know whether it works in your store is an A/B test:
- Split the traffic into two random groups: one with the try-on and one without it.
- Measure conversion, average order value and returns.
- Run the test long enough to have reliable data (paying attention to seasonality).
- Analyse by category and by customer type.
If you cannot run an A/B test, compare similar periods and analyse cohorts, being cautious with the conclusions. More in how to measure virtual try-on performance.
What else to look at
- Usage rate: what percentage of visitors use it.
- Time on page and depth.
- Return rate by reason.
- Cost per incremental order against the cost of the project.
- Satisfaction and comments.
How to increase the chances of success
- Start with high-potential categories.
- Look after visual quality and performance.
- Put the button in a visible place on the product page.
- Integrate purchase (add to basket from the try-on).
- Promote it in campaigns, email and social media.
- Train customer service so that they recommend it.
Key ideas
- There are signs of a positive effect, but published figures should be treated with caution.
- The result depends on the category and on the quality of the implementation.
- Avoid selection bias: use an A/B test.
- Measure conversion, returns and order value.
- Start with a pilot in high-potential categories.
Frequently asked questions
What conversion increase can I expect?
A figure cannot be promised. It depends on your category and on the implementation; the pilot will tell you.
How long do you need to measure?
Long enough to collect meaningful data and cover purchase cycles; several weeks is usual.
Can it make conversion worse?
If the experience is slow, fails or looks unrealistic, it can do harm. That is why quality is key.
Does it work for small stores?
Yes, if there is enough traffic to measure and a suitable category.
More guides on this topic
How to add AR product visualisation to an online store
A guide to adding AR product visualisation to an online store: 3D models, formats, integration on the product page, platforms, performance, analytics and best practice.
How to create an interactive retail campaign with QR codes and augmented reality
A guide to creating an interactive retail campaign with QR codes and augmented reality: concept, touchpoints, mechanic, reward, legal points, launch and measurement.
How to measure virtual try-on performance
Metrics for measuring a virtual try-on: usage rate, time, conversion, order value, returns and satisfaction. A/B test, attribution and how to avoid bias.