The sizing, a major challenge
Choosing the right size in the online ready-to-wear world for an end-user relies on a very complex system because it is not standardized.
Unlike an in-store experience, the customer can no longer talk to a salesperson and it is therefore more difficult to get recommendations and advice on sizes or fits.
Despite efforts to standardize this system and the advancement of new technologies, sizes vary considerably from one country to another, from one brand to another and even within the same brand on different collections.
The fact is simple: look in your wardrobe at the sizes of your different clothes, most of the time they will never be similar. The truth is that the numbers on these labels are arbitrary and there is no real standardized sizing system for clothes these days. This is a real challenge for fashion e-commerce retailers. We analyzed the different approaches proposed by the current solutions on the market that aim to answer this challenge.
Sizing, a key concept to master
Sizing is a fundamental element, both for the customer or for the company. A bad sizing will have negative consequences for the company (more returns).
On the company's side:
Returns lead to a lower sales performance for the company, especially if the customer orders two sizes for the same product due to hesitation, and returns one of the two sizes. Returns lead to uncertainty in product inflows, which would not have a positive impact on the company's accounts, due to the variability of these inflows. In addition, returns and exchanges are often at the seller's expense, which represent a significant amount of money for the company, and a significant logistical burden.
On the customer's side:
The customer which is not satisfied with the size ordered will return the product.
According to the FEVAD, a potential buyer will order on average 22 times a year; among these orders, the e-commerce return rate is 24%.
According to Body Labs, in the US, there is an average return rate of 30% for products ordered online. Among these 30%, 64% of returns are for a product that does not fit the customer. A customer who is unsure about sizing is less likely to order from a site.
Current sizing methods
1. The brand's global size guide: a frequently used tool that is far from perfect
There are different ways of proposing a sizing: the most common is the size guide, a matrix that indicates to the customer the ideal size, most of the time defined according to his measurements.
· On the company side: the size guide is inexpensive and not very difficult. You can build matrices based on statistics. The size will be defined according to weight and measurements.
· On the customer side: This method allows the customer to determine his size in a way that is conventional for him as long as size guides are common.
There are a few drawbacks:
· On the company side: due to the variability of the cuts, materials and shapes of the different models, the lack of precision is a problem as it leads to frequent returns. Indeed, it is not uncommon for garments to fit differently, even within a collection, due to the dispersion of production units and fluctuations in production methods.
· On the customer's side: size guides are rarely established by product. This increases the uncertainty of the customer who, moreover, does not always know his measurements, which makes it difficult to determine the ideal size. Furthermore, the customer is often unwilling to make the effort of reading this matrix. Indeed, we live in a world of zapping, in which users are used to seeing hundreds of pieces of information scrolling along their news feed: their attention has thus become precious and fragile.
Since the size guide does not seem to be an ideal tool for optimising the customer experience and thus sales, other solutions have been developed, based on two types of approach: precision about the users' measurements or precision about the precise measurements of the garments.
2. The statistical approach: a simplified UX but can get limits in the precision
The statistical approach makes it possible to determine the ideal sizing according to statistics. The size proposed to the client is obtained through comparisons with existing statistics and data.
Solutions such as Fitizzy, Fit Analytics, My Size ID or True Fit are part of this approach by allowing to implement on the brand's website a module creating a direct interaction with the customer, in order to help him to find his ideal size.
- On the company side: The solution is efficient and easy to implement.
- On the customer side: the service suggests to the customers the ideal size according to four main criteria: age, height, weight and body shape. Finally, for some solutions, the customer is given the ideal size according to 8 morphological profiles, which is simple and intuitive.
Some drawbacks to note:
- On the company side: the decrease in the return rate can be limited if the answers are not reliable.
- On the customer side: the information requested may seem too intrusive for the customer. They do not want to communicate their weight or age, for example. The lack of precision about their measurements linked to this type of approach is a problem. For some solutions, it may seem reductive to limit the approach to 8 morphologies since each client is unique by his body and his preferences.
3. The millimeter measurements approach: ideal in theory, but logistically complex in practice
So, the statistical approach satisfies the need for precision about garment measurements, but has its shortcomings from the users' perspective.
Thus, another angle consists in centering the precision on the customer's measurements. Companies that have tried this approach have experienced difficulties in their business development.
A first example of this approach is the Japanese brand Zozo, which developed Zozo Suit, an innovative sizing system based on a suit that we receive at home and that takes our measurements.
Another solution is the one that had been developed by Sizomatic and Meality or Veertus and Presize, a scanner that takes the customer's measurements confidentially and independently. Sizomatic did it via a booth while Veertus did it via photographs of their body taken by the customers on their smartphone.
On the Company’s side: the high quality accuracy of the measurements allows to limit the volume of returns and thus to guarantee a greater cash flow security.
On the customer's side: the ZozoSuit combination allows the customer to obtain an extreme precision in the measurements, since this technology allows the collection of measurements on a total of 400 points, where a traditional tailor would only take about 30 points. The cabin also allows for maximum optimization of the client's measurements.
Some drawbacks to note:
On the company’s side: this solution is very expensive because the technologies deployed are important. It requires a highly qualified workforce and complex logistics to implement. It is necessary to send to each customer the combination allowing them to take the measurements, then to treat them on a case by case basis, since each customer is unique. Moreover, it is not a model that can be transposed to other brands. As for the booth, it requires sufficient space in each store and can be costly.
On the customer's side: the process is long since he must wait to receive the suit to take his measurements, before choosing the garment he wants to buy. If the measurements correspond to the most common ones, a stock is provided, the product is delivered quite quickly, but if the measurements are rarer, the customer will have to wait for an additional period of time, which can be more than a month. Moreover, the customer may be reluctant to accept this type of approach.
An innovative UX: Cleed.ai
1. Cleed.ai's approach to sizing
Cleed.ai has built its innovative UX based on the global sizing guide or on the size guide by model if available.
Thus, a chatbot answers customers' questions, guides them on sizing, to suggest the ideal size for the customer, for that specific product, and not a general recommendation based on estimates, as other methods would do. The quick dynamic questionnaire collects the sizing information and then matches it.
Customers from all over the world will be able to find the right size for each model. This is how Cleed.ai allows its customers to group numerous size guides by categories / models / gender / international names, etc., and thus simplify the experience while guaranteeing a high level of accuracy, and controlled data, unlike the statistical approach.
2. Business impact of choosing Cleed.ai
- A higher conversion rate
Because customers get quality advice, conversion rates increase significantly. Our customers benefit from better sales performance, due to the right size recommendation but also recommendations based on targeted user preferences.
Indeed, Cleed.ai has integrated a visual search tool into its recommendation and search system, creating an optimal customer experience. In addition, since the chatbot now knows the customer's size, it can now recommend products that might appeal to them, but especially in their ideal size.
In addition, this method helps to improve the company's CRM, by collecting user-specific data. This allows them to know better their preferences, in order to improve the company's sales performance in the future.
- A satisfied and reassured customer
Thanks to Cleed.ai's personal shopper assistant, customers feel like they are in a boutique. They get personalized advice according to their specific needs. Today's consumers want to have a service, a product, that is unique to them. Thus, the uncertainty about sizing and the lack of personalization of the classic online experience disappear in favor of a true immersive experience with this omniscient personal shopper assistant, which allows the customer to live a unique moment and find the ideal product.
- A significant decrease in returns
Brands that have implemented the service offered by Cleed.ai have seen a significant reduction in their rate of returns due to an unsuitable size choice. This is a concrete validation of the approach we described above, which has the advantage of being as accurate as possible (ideally with a size guide by category/collection).
Thus, the approach based on the control of the precision of the measurements encountered too many logistic difficulties. The statistical method provides satisfactory results overall. But the most ideal method is to have a size guide by category/collection and to confront it with the user's measurements, via a fluid UX, like the one proposed by Cleed.ai.