How the Technology Works
Virtual try-on (VTO) tools use a combination of augmented reality (AR) and artificial intelligence (AI) to digitally apply makeup to your face. When you use a VTO feature on a website or app, it accesses your phone or computer's camera. Advanced algorithms
then detect key points on your face—like the corners of your lips and eyes—to create a 3D map of your features. The tool then overlays a digital layer of the product, whether it's lipstick, eyeshadow, or foundation, onto the live image of your face. The goal is to create a realistic simulation of how the product would look in real life, allowing you to experiment with different shades from anywhere.
The Challenge of Real-World Conditions
While the technology is impressive, its accuracy is heavily influenced by factors that brands can't control. The single biggest limitation is ambient lighting. A shade that appears to match in the warm light of your living room might look completely different in cool, natural daylight. Your device's camera quality and screen color calibration also play a huge role. A low-resolution camera might not capture the nuances of your skin, and different screens can display colors with slight variations, further skewing the result. These variables mean that even with sophisticated AI, what you see on screen isn't always a perfect representation of the physical product.
The Undertone Dilemma
Matching your skin's depth is only half the battle; the real challenge lies in identifying your undertone. Undertones—the subtle cool, warm, or neutral hues beneath your skin's surface—are notoriously difficult for many people to self-identify, and AI faces a similar struggle. While some advanced tools are specifically designed to detect undertones by analysing multiple points on the face, many simpler VTOs are less effective. They may overlay a color that looks plausible on the surface but clashes with your natural undertones, which is why a shade can look 'off' even if it seems like the right depth. This is a common reason for a mismatch between a virtual recommendation and the real-world product.
The Bias in the Algorithm
A significant and ongoing issue in AI development is algorithmic bias. Many AI systems have been trained on datasets that predominantly feature lighter skin tones. This lack of diversity can lead to significant inaccuracies when analysing darker skin. Studies have shown that some AI models are far less effective at recognising features and tones in darker skin, sometimes altering a user's face to appear lighter or failing to provide an accurate match. While companies are actively working to create more inclusive datasets, using systems like the Monk Skin Tone Scale to cover a wider spectrum, gaps in performance remain a critical issue.
A Tool for Discovery, Not Diagnosis
Experts and even the companies creating these tools suggest that VTO is best used for engagement and discovery rather than for making a final purchase decision, especially for foundation. It’s a fun and effective way to play with lipstick or eyeshadow colors, where a perfect match is less critical. However, when it comes to the precise science of foundation matching, these tools are more of a helpful guide than a foolproof solution. They can help you narrow down your options, but they can't replicate the experience of feeling a product's texture or seeing how it oxidizes on your skin.
Tips for Getting a Better Virtual Match
If you want to use a virtual try-on tool, you can improve your chances of getting a more accurate result. First, find a spot with bright, natural, and indirect light; facing a window is ideal. Avoid dim or strongly colored lighting. Make sure your face is clean and free of existing makeup. Use the highest-quality camera you have available, typically the rear-facing camera on a smartphone. Finally, instead of relying on just one tool, try a few different brand websites to compare recommendations. If you find a promising shade, look for online reviews and swatches from real people with a similar skin tone and undertone before you decide to buy.
