Does augmented reality in retail increase sales?

In 2016, Charlotte Tilbury put augmented reality mirrors into its London store that could show customers ten different makeup looks in under a minute, without applying or removing a single product. UK technology studio Holition built the experience, which was shortlisted for a Retail Week customer innovation award the following year.

Trying on makeup in a shop usually means dabbing colours on the back of a hand and hoping the shade under the shop lighting looks the same at home. The mirror removed much of that guesswork by showing the customer a complete look straight away.

John Lewis took the idea into homeware with its Virtual Sofa feature, letting customers scan their own living room and see how a sofa would look and fit before buying it. M&S went in another direction, trialling an in-store tool that guided shoppers to the shelf they needed.

None of this made augmented reality something UK shoppers could expect to find as standard across retail. Nearly a decade later, however, there is a long enough track record to ask whether any of it actually increased sales and reduced returns, or whether the most reliable outcome was a good press day.

What is augmented reality in retail?

Augmented reality in retail overlays digital products or information onto the real world, using a phone camera, tablet or in-store screen. Shoppers can see a rug on their living room floor, try a lipstick shade without opening the box, or scan a shelf to check stock, product details and reviews.

Some tools marketed as virtual try-on don't use a live camera at all. The customer uploads a single photo of themselves, and the product gets placed onto that image, so turning their head or moving doesn't change anything on screen. It can still give a rough sense of colour or style, but it won't show whether a pair of glasses actually sits right or how a shade looks from another angle.

Where UK retailers are using augmented reality

Although augmented reality receives a lot of attention, its use in UK retail is still fairly selective. A GetApp survey of more than 1,000 UK consumers found that over half had never used AR while shopping online or in a physical store. Even so, most said they would be willing to try it. That reflects where the market stands today. Many retailers have experimented with AR, but most projects have stayed within a flagship store, a single product category or a limited pilot rather than expanding across the entire business.

Try-on experiences

Specsavers has been doing this for several years with its Frame Styler tool, one of the most common uses for AR in retail, helping customers see how something will look before they buy it. First launched in stores in 2018, it scans a customer's face using an in-store tablet, recommends frames to suit their face shape and lets them see different pairs from multiple angles before trying any on. The same experience has since been made available online, so customers can use it from home.

Charlotte Tilbury took a similar approach with makeup. Its Magic Mirror, introduced at the brand's London store in 2016, could show up to ten different makeup looks in under a minute without applying or removing a single product. Developed by UK agency Holition, the experience was later shortlisted for a Retail Week customer innovation award.

M&S also explored virtual try-on for fashion. Working with technology partner Zyler, it tested an in-store experience at its Oxford Circus and Harrogate stores, allowing shoppers to see Jaeger clothing on themselves digitally before deciding whether to try it on.

Furniture at home

John Lewis introduced a Virtual Sofa feature in its app, tackling a problem furniture retailers know well: helping customers understand how something will fit into their own home. It lets customers scan their room and place one of its best-selling sofas in the space. They can rotate the sofa, compare colours against their existing décor and get a much better sense of scale before placing an order. At the time, the retailer said buying a sofa often involved as many as 13 separate decisions, and the feature was designed to make that process easier.

Argos followed a similar idea by working with creative agency REYDAR on an AR-enabled app that lets customers place furniture inside their homes before buying.

Store navigation

M&S took AR in a different direction with List&Go, a wayfinding app developed with UK startup Dent Reality that guides shoppers around the store rather than helping them choose products. It's part of a wider trial across five of the retailer's "digital test stores," and the app also pulls from M&S's digital planograms, meaning shelf locations update automatically whenever the store layout changes. After creating a shopping list, customers are shown the quickest route to each item, with shelf locations displayed over the live camera view. The experience launched first at the retailer's White City store in London and was described by M&S as a UK first.

Interactive campaigns

Tesco used AR less as a shopping tool and more as a way to make marketing campaigns and packaging more engaging, launching its Discover app in 2013 to let shoppers point their phone at selected wine bottles and unlock videos and additional product information.

Sainsbury's used AR for a different purpose, creating an Easter bunny hunt in 2019 that turned a regular supermarket visit into an interactive game for families.

JD Sports leaned into entertainment by partnering with Snapchat for an AR arcade game at its Oxford Street flagship during the Christmas shopping season. Shoppers scanned a code on the storefront to unlock and play the experience.

Does AR increase retail sales?

The answer depends on what's being measured.

Purchase confidence

Vogue Business and Snap surveyed 1,151 UK luxury shoppers in early 2023. They found that 72% considered AR shopping experiences important when buying from a brand. Snap's business depends on wider AR adoption, so this is more a measure of how customers feel than proof that AR increases sales.

Specsavers lets customers see how different frames look before visiting a store. Charlotte Tilbury's Magic Mirror lets shoppers try on multiple makeup looks without touching a single product. John Lewis lets shoppers place furniture inside their own home before deciding. The products are different. Specsavers, Charlotte Tilbury and John Lewis all answer the same question for a customer standing in front of a screen: will this actually suit me?

Tesco's Discover app reached almost 150,000 active users within its first 18 months. That suggests people found enough value to keep coming back, rather than trying it once and moving on.

Returns

Returns are easier to measure than confidence. Alter Agents surveyed 4,028 shoppers across the UK, US, France and Saudi Arabia for Snap and Publicis Media. Two-thirds said they were less likely to return a product after using an AR feature. The research was commissioned by companies with an interest in AR, so it's worth reading with that in mind. But the behaviour it describes matches a problem retailers already understand.

A customer who has seen how a sofa fits the room, or how glasses look on their face, knows more before they order than someone working from a photo alone. That's likely why furniture placement and virtual try-on were some of the first ways retailers used AR.

Where the biggest gains appear

The strongest evidence comes from research that looks at what people bought rather than what they said. A 2022 study in the Journal of Marketing analysed real transaction data from an international cosmetics retailer. It found the biggest effect where shoppers faced the most uncertainty. Sales rose most for brands that weren't well known, products that cost more, and items that only appealed to a narrower group of shoppers. The largest gains came from customers who were new to online shopping, or buying from that category for the first time.

The research wasn't carried out in the UK, but the situations it describes are familiar. It's hard to choose prescription glasses, a new makeup shade, or a sofa from a photo alone. Those are exactly the purchases where UK retailers have used AR first. It works best when it answers a question a standard product page can't.

How should retailers calculate AR ROI?

The simplest way to calculate an AR feature's return is to compare how the same products perform before and after it's introduced. Conversion rate, average basket value and return rate are usually enough to build a useful picture, as long as they're measured over similar periods. It's best to avoid sale events or seasonal peaks, otherwise the comparison starts reflecting everything except the AR feature itself.

If returns fall, conversion improves or customers spend a little more on those products, the feature can reasonably take the credit. Looking at overall sales tells you much less, because they're affected by everything from marketing campaigns to the time of year.

Returns are the easiest of the three to price, since a retailer already knows the courier cost, the restocking work and the markdown a returned item often needs before it can be sold again.

Return saving = (old return rate - new return rate) ✕ number of orders ✕ cost per return

For example, a retailer taking 1,000 orders a month, with a return rate that falls from 25% to 20% after introducing AR and a return cost of £15, would save:

(25% − 20%) × 1,000 × £15 = £750 a month

Conversion and basket size are harder to measure than returns because there's no fixed cost to work from. The increase has to be measured against the value of the extra sales those products generate.

Running the feature costs money too. M&S's List&Go only works because its store planograms are kept up to date. A room-based AR feature needs accurate 3D models, while a virtual try-on depends on tracking that continues to work well across a wide range of real customers. Those costs need to be included when working out the return.

The Journal of Marketing study discussed earlier found the biggest gains on products customers felt less certain about, including unfamiliar brands, higher-priced items and first purchases in a category. That's where AR is most likely to make a measurable difference.

What to check before starting a pilot

1. Does the product get worn, or does it sit in a room? Products worn on the face or body need face and body tracking, while furniture and anything placed in a room needs room-based AR. Settling which one applies before any development starts avoids building the wrong kind of prototype and having to start again.

2. Is there a team in-house that can build this, or does it need a specialist? AR development calls for skills most retail teams don't have on staff already, computer vision, 3D modelling, tracking calibration, so it's worth checking early whether anyone in-house has built something comparable. If not, budgeting for an agency from the outset avoids a mid-project scramble to find one.

3. Does the underlying data already exist? Which data matters depends on the use case: room-based AR needs accurate 3D models of the products involved, try-on tools need tracking that stays accurate across a wide range of real customers, and wayfinding needs store location data that stays current. These don't exist by default, so it's worth confirming they're in hand, or budgeted for, early. 

4. Has a baseline been recorded yet? Record conversion rate, basket value and return rate on the specific products getting the AR feature before launch. That baseline is what makes it possible to tell afterwards whether the feature actually made a difference.

5. Can the pilot stay small? One store and one product line gives a clean comparison, and an easy point to stop if it doesn't work, before committing to a full rollout.

The bottom line

Shoppers say they're less likely to return an item after using AR. Sales rise too, mainly on products people aren't sure about yet, an unfamiliar brand or a first purchase in a new category. Specsavers, Charlotte Tilbury and John Lewis all built their AR features around exactly that kind of purchase.

Retailers testing AR should start in the same place. Measure conversion, basket size and returns on a product line customers hesitate over, then check whether the improvement covers what the feature costs to run.

Fynd GlamAR builds virtual try-on for eyewear, makeup, jewellery and accessories, letting a customer see a product on their own face or body through a phone camera rather than a fitting room. It can also convert existing product photos into 3D models, so getting started doesn't require a full product reshoot.

Talk to us