All about Visual Research (with The Home Depot, Best Buy, Tommy Hilfiger and their partner Slyce)
- Visual search is a reflex that takes hold: 30% of total searches made in the US on Google
- In a store too. The 1timecustomer activity in a store with his mobile? Take a picture of a product
- And it works: on average, retailers who offer visual search increase the conversion rate by 60% and the average order amount by 20% (Source: Slyce customers)
- The risk for a brand to wait too long? Be left behind… by Amazon’s strong ambitions in the field.
In this post you will find out:
- What is visual search?
- Why do brands and retailers need to equip themselves now?
- What are the results ?
- Use cases in Retail
- Examples from The Home Depot, Best Buy and Tommy Hilfiger.
Possess before others the new highly coveted product: a spring of rare efficiency to lead to the purchase. Tommy Hilfiger knows it well. Confused about the fashion world, the couturier was one of the first to institute the “See Now, Buy Now” during his major shows led by his muse Gigi Hadid. For 2 years, those present at its parades have been able to immediately buy the clothes worn by the models who parade, via a dedicated TommyLand Snap: Shop app. (People who watch the show from home live stream can too).
Putting on sale his ready-to-wear collection during his live shows, and not six months later as it had been done since the dawn of time, was a small revolution in the fashion sector. It is also a huge commercial success for Tommy Hilfiger.
WHAT IS VISUAL RESEARCH?
This feat is made possible by the visual search technology of Slyce, which works like this:
- A person opens the mobile app or surfs on the mobile site of a brand or store integrating Slyce
- She spots the little camera icon
- She takes a photo of the product she is interested in, in real life, on screen or in print ads
- The app shows him this product, or an equivalent product, present in the retailer’s catalog. If more than one item is in the image, Slyce can identify each separate item and can find their matches.
Example with The Home Depot.
In The Home Depot app, a watering can photographed
Slyce.it, a company well established in the United States and the Middle East, has more than 50 clients (Neiman Marcus, Macy’s, Abercrombie, Tommy Hilfiger, The Home Depot, Best Buy, Landmark Group….) Who have integrated into their Slyce’s own SDK app.
Identified by Lafayette Plug and Play, the accelerator resulting from a partnership between the Galeries Lafayette Group and the famous American Plug and Play accelerator , Slyce integrated the new promotion # batch5 in September and made the program, as pointed out Deborah Loye in Les Echos “the starting point of its deployment in Europe”.
In this type of service, speed and accuracy are two determining factors, because otherwise consumers who are not very patient and not very forgiving take their turn. Slyce clients return a result in 6-7 seconds, with an accuracy rate of 95%.
WHY EQUIP NOW?
Because tomorrow will be too late. We know the Alibaba, Amazon, eBay, Pinterest, Facebook and Google are at the forefront on the subject. Sure Pinterest, the number of monthly visual searches increased by 140% between February 2017 (250M per month) and February 2018 (600M per month).
According to Amy Vener, Retail Vertical Strategy lead at Pinterest, interviewed by Digiday, the rise of online and mobile shopping ” sacrificed the visual experience a bit, making it harder for retailers to entice consumers to make unexpected discoveries. Yet people want to seek out and discover new ideas, even if they don’t have the right words ”.
When at Amazon, he pushes his pawns. A partnership with Snapchat allows a user to identify a product through Snapchat and then be naturally directed… to Amazon.
Another partnership, with the magazine Good Housekeeping this time, in the form of a pop-up store called GH Labs in the Mall of America, which allows you to purchase the products on display by scanning the Amazon app for a QR code (a SmileCode, the small name given by Amazon to its own QR codes). Integrated payment and you’re done.
FOR WHAT RESULTS?
Does Visual Search Sell? On average, Slyce customers increase the conversion rate by 60% and the average order amount by 20%.
USE CASES IN RETAIL
- Shopping – The Example of Neiman Marcus
A common use case is purchasing. Example with Neiman Marcus: take a picture of a pair of shoes and find in Neiman Marcus’ app this pair of shoes, or half a dozen other very similar pairs (which makes it very easy to copy an outfit seen during a Fashion Week or worn by a celebrity).
Credit: Neiman Marcus
- Find a similar – If the retailer does not sell the exact item, they can still provide equivalent product results. This service is so effective at detecting and classifying images by similarity that some retailers use it to feed a Find Similar carousel on their applications and product pages.
- Complete the Look – American Eagle Outfitters challenged Slyce to associate visual search with a complete outfit, allowing customers to identify a product and “complete the look”. The person takes a picture of an item of clothing (eg a top) and gets results. By simply clicking on “Build Me An Outfit”, she will get a complete “look” with clothes and accessories that complement the top, and that can be personalized according to the customer’s CRM profile. AEO launched this service on Facebook Messenger last fall.
But the use cases do not stop there:
- Help for store salespeople – Home Depot
At Home Depot, 70-80% of photos are taken by in-store salespeople. They use it to help customers locate the products they are looking for in the store. Other use: at checkout when there is no barcode on the product.
- Creation of shopping lists at Jet.com (Walmart) andcreation of wedding lists – People create their shopping list by going around their kitchen (Jet.com) or their wedding list by going around the store (the bride and groom at Bed Bath and Beyond) and taking pictures of the products desired. The list is built even faster than by voice.
- Print to web – Retailers use this technology to activate their catalogs, flyers and printed catalogs, as well as in-store signage. The Express mode sign thus indicates that it has multiplied by three the conversion rate of its printed catalog.
- Another use case, in BtoB this time: the Buyer App. At Neiman Marcus, employees have a dedicated app. By taking a picture of a product, they immediately gain access to its sales history, full description, or similar products.
3 REAL CASES
USE CASE # 1 – The Home Depot
The experience – The scene repeated itself every day at The Home Depot. Customers wandering the aisles, looking for a recognizable salesperson with his orange apron, called to the rescue to find the screw or tool they are looking for.
In 2016, in order to help these customers, the brand integrated Slyce into its app.
When a customer takes or uploads a photo of the part they need, machine learning and computer vision technology scans every product in The Home Depot’s catalog (one million products) to return the product – or d ‘other similar products – tell them where the product is in the store, or send it back to the online page to buy it directly.
The teachings :
- Increase in loyalty to the app X by 3 for customers who use visual search
- A usage growth rate of 21% per month
- Also used by sellers : Over 70% of all visual searches for The Home Depot come from store salespeople, who use it to help a customer find and locate equipment on the shelf
USE CASE # 2 – BEST BUY
Case very similar to that of The Home Depot, with an average monthly increase in store use of 18%
USE CASE # 3 – The Tommy Hilfiger Show
Preview + Instant gratification
For 2 years, Tommy Hilfinger has been giving a boost to its shows by offering the 1timevisual recognition app designed for a parade, the TommyLand Snap: Shop. Objective: to allow its customers to immediately buy the clothes presented during the parade.
Preview + Instant gratification
For 2 years, Tommy Hilfinger has been giving a boost to its shows by offering the 1timevisual recognition app designed for a parade, the TommyLand Snap: Shop. Objective: to allow its customers to immediately buy the clothes presented during the show.
The experience is as follows:
ON SITE, DURING THE SHOW
To Web Poster : buy on an advertising poster
Store to Web: Buy in the pop up store during the Show
DURING THE SHOW OR REMOTE
Instantly buy the garment presented during the parade.
The correspondant :
Laurence Faguer is a marketer and entrepreneur “go-between” France and USA, founder of
At the request of French companies, she identifies in person innovations in Digital, Mobile and Retail in the United States, before they are known in France, then helps them to successfully transpose these successful strategies in the US
Laurence is US expert for FrenchWeb who resumes from time to time the publication of articles from his blog.