In a world of data, how do we react to information and insights that are not only relevant but also accurate? Can the data humanity uses to dress itself truly reflect the realities of what’s out there? In a world of data, how do we react to information and insights that are not only relevant but also accurate? Can the data humanity uses to dress itself truly reflect the realities of what’s out there?
While it is true that many of the world’s fashion brands are struggling to adapt to the times, what is not true is that they are unable to find a way to do so. The reality is that the fashion industry is awash in solid data that would support many of the decisions that they are making. [But they’re not using it.]
Large-scale changes in consumer expectations and behavior towards clothing have increased consumer expectations towards fashion brands. And since modern fashion is produced in advance for future consumption, it is imperative to understand the wishes and desires of buyers in advance to satisfy them with appropriate collections and reduce unsold inventory. Innovation and digitalization offer few tools to help brands understand the needs of today’s customers. While radical digitization and innovation are necessary for consumers, brands and retailers looking to embrace these advances face many challenges. While we often only associate innovation in fashion with e-commerce, AR/VR technology and textile innovation, in this series of articles I will focus on the current disadvantages and limitations of physical (offline) shopping. I am also trying to find systematic results of this existing gap between online and offline commerce and see how we can ultimately improve the status quo and build hypotheses about the exact outcome of such changes. I would also distinguish between the data needed to create and publish collections and the data needed to find and target customers to generate (more) sales. In my opinion, the digitization of offline retail has great potential to reshape the fashion industry. In this article, I would like to focus more on the role of data in curating collections and understanding buyers’ desires and expectations in this regard. I want to look at specific data that can help with the awareness and forecasting challenges that many fashion brands face, and that can allow for better planning and subsequently greater effectiveness. What specific data is most useful to improve the efficiency of all processes in fashion? Can data influence sustainable development? If so, what data can we use? Where and how do you get it? Fashion has always been a plurality of designer voices and consumer preferences. Wants and needs are highly dependent on geography, age groups, body type and other variables – perhaps even more so than in the food industry, for example. The decision-making process involves several phases (observation of fashion trends, search for the best offer and the most suitable clothes, etc.), with trends and styles constantly changing. Due to this plethora of different and segmented processes, the role of data should be essential for fashion brands, especially in analyzing customer behavior and being able to make the right predictions and perceptions. However, the fundamental implications of Big Data and its solutions in the fashion industry go unnoticed precisely because of the complexity of the fashion industry and the difficulties in implementation. This is discussed in more detail below.
What data helps fashion brands to meet customer expectations and customize and design clothing collections that are more likely to sell?
As I said in my previous post, I truly believe that any unsold garment is inherently unsustainable, regardless of how sustainably it was produced and delivered. Much of the sustainability problem in fashion can be attributed to the inability to correctly predict what will be consumed where. In this context, I reflect on the relevance of different groups of factors and information that I believe would be extremely important for fashion brands to improve collection creation and gain strategic insights into distribution and adaptation aspects: 1. Data about the product – the customer – the relationship. It’s about why, where and who liked it, who followed me and eventually bought my garment. What is the origin of this interest? Was it spontaneous? Was the client influenced? If so, by whom? 2. Purchase of goods – data for decision-making. How and why did someone decide to buy my clothes? What was the journey and what influenced this decision? Did price or willingness to buy a particular product from my brand make a difference? 3. The customer as a consumer of KYC. This group of data segments and analyzes my brand’s consumers in more depth. What social, demographic, and behavioral patterns lead to my particular brand? I know this seems far removed from the current reality, but I am convinced that technological solutions will be necessary to enable this first necessary product in terms of traction in fashion. Essentially, it’s about connecting fashion brands and fashion lovers. I also think that we need to find a way to make this information and data fully compliant with data protection directives such as the GDPR. I also think this data should be available to fashion brands.
What data do we currently have? Impact of disproportionate fraction consumption on existing fashion and business models.
Due to the complexity of distribution channels in the industry, and even more so due to the lack of appropriate technological solutions, it is still almost impossible to adequately reflect offline consumer data (which accounts for 80% of total fashion consumption) in current forecast data, thus failing to adequately respond to customer needs. Anonymous shopping at stores in the city is still the norm and is widespread (unless one uses loyalty or membership cards). Even if the maps are used, what can this information really tell us about the issues described above? In fact, a small percentage of purchases in physical retail are tracked by customer management systems as ordinary sales transactions attributed to an identified customer. Technology is trying to solve the above problem with ingenious solutions. These include cameras equipped with artificial intelligence that can recognize the reaction of a consumer’s face when they look at certain items of clothing, or classify consumers based on their footwear and single out a particular preferred style and personal customer group to gain valuable data. Some stores using AI assistants with facial recognition are offering familiar customers personalized offers based on their previous purchases. Other proposed methods for collecting repeated consumption data include tracking purchase events through credit card transactions at offline merchants. One of the ways retailers try to attract customers is by using geolocation technology that can show how you and your phone move around the world. Other sources used primarily by retailers include social media sites or purchase data from adtech companies. While these choices may seem interesting, the first critical issue is the validity of the data – facial reactions and footwear choices cannot guarantee infallible results, as these factors may be inconsistent or influenced by a momentary whim or fleeting mood, and may not reflect the style and habits of the observed consumer. These are temporary expressions, not permanently created data attributes. This highly sensitive and personal data, and the way it is collected, is still the subject of intense debate in the EU. Another issue is the attribution of the data – in my opinion, the extracted information cannot yet be interpreted correctly, as it is only a small part of the chain of relevant behaviors and attributions in the whole context of interaction with fashion. The existence of these solutions unwittingly betrays the urgent need for offline consumer data. Due to the irregularity of fashion consumption patterns and the lack of clear and uniform technological tools for data collection and customer acquisition by offline retailers, there is a major limitation in the availability of relevant data. For fashion brands, there is almost nothing left to do but apply the available online data to the offline world and rely on the completed sales data to make new predictions for future collections. Looking at the quality of existing offline retail data readily available to brands, the hard truth is that over 90% (by my own estimation) of existing physical retail customer data is not usable due to the quality of information described above. By way of explanation: Firstly, as already shown, the possibilities of autonomous data are very limited. The small available portion contains less relevant information related only to a specific shopping event or to the customer’s behavior in the store, without addressing a variety of user behaviors and the behavior of mobile consumers. But in reality, the range of behavioral patterns is much broader than what is available from tracking a single purchase in a store. The collection and attribution of this data is possible online, at least in theory, if all transactions take place on the same e-commerce platform. Another question is how brands can access this data. We’ll come back to this later. As for customers leaving the store empty handed, no one knows why people go into a store and don’t buy anything. What about this data? In fact, only one in five shoppers make a purchase when they visit a store; according to a recent statistic, about 96% of shoppers go home empty-handed. This means that we (probably) have a small amount of relevant data about one customer who made a purchase (see above), but nothing about the other four who left the store empty-handed. In summary, I estimate that approximately 99% of the relevant data is not in fact available or accessible to anyone in the clothing retail sector. Suppose we could get data on every person who entered the store, went out and made a purchase in the store? What to do with this emptiness? It is much more relevant to analyze why a customer was initially interested in a particular product or brand but later did not buy it, as well as to analyze the expectations when they enter the store and the factors that play a role in the decision-making process. What if we could extract this data in real time and use it to solve the problem of limited and complex product searches and help customers find what they are looking for? As McKinsey has shown, behaviour is becoming increasingly unpredictable; people may be influenced by a brand online but end up buying something similar offline, or vice versa – they look for something offline and end up buying online because of the variety of offers. Linking online and offline retail data in terms of customer and product behaviour appears to be important for assessing data quality. How do brands get their data today? What is available and how is it used? As mentioned earlier, brands have little or no qualitative data on offline retailing. The only available and reliable data still comes from tracking consumer behavior/purchases/advertising consumption on the Internet. However, problems of interpretation also arise: on the one hand, the diversification of online retailing and the availability of data for brands (due to the abundance of online shops) makes it difficult to attribute the available information to customer behavioural patterns or tangible garments. Fashion brands that produce collections, distribute them and seek to maintain contact with customers naturally rely on numerous distribution channels (depending on the distribution strategy). There is a wealth of literature on the challenges brands face in staying profitable and launching online, especially with one of the widely used e-omnichannel solutions. Here you can read how brands are using the new e-commerce sales channels to increase their sales. However, these e-commerce platforms serve their own purpose by using customer contact and technology to obtain unique data at the source. It’s no secret that one of the main strengths of all electronics giants is the amount of data they store about their customers and the consumption of their products. This also creates a big dilemma in terms of collecting and using data to make predictions and perceptions, which is important for fashion brands. In turn, by applying certain mathematical methods, we reduce the amount of data readily available to brands that they need to make informed decisions about their future collections. The readily available data is then fed into algorithms that provide basic information to the brand. Again, however, the type of information available is important. Data from fashion brands’ online stores provide very limited information about shopper behaviour, as they are limited to their own in-store collection, while the broader data available to the online multi-brand giants is not directly accessible to the brands. What a dilemma! A logical consequence of this is that fashion brands often turn to external data providers to help them extract relevant data from various sources using artificial intelligence, pattern recognition and the application of algorithms needed to draw conclusions about upcoming trends and gather insights about competitors. Some fashion brands hire data analysts to get relevant data. Giants like Nike even buy data companies to better understand their customers. But even the best data scientists and algorithms cannot solve the problem of lack of data in offline cases and in some cases its unavailability (online cases). There is a huge gap in data availability and accessibility that technology can and should fill. Brands need to better understand their customers immediately and engage with them directly. It would be ideal to extend this linkage with various technological capabilities to get a 360 degree view of behavioral patterns in fashion. Does the lack of intelligence have to do with fashion problems? After discussing the different ways consumers interact with fashion, the answer is: Yes, the lack of intelligence is partly related to fashion and plays an active role in maintaining the current state of the fashion system. This shortage directly affects production, distribution and logistics decisions, resulting in overproduction and unsold stock. None of these factors should be overlooked, as they have a direct impact on the profitability and even the reputation of the brand. With the right data, brands can produce what is most likely to be consumed, reduce dead stock and reduce their carbon footprint, increasing sustainability and profitability. By producing, sourcing and selling smarter products, brands can automatically increase sustainability. Specific data will further reduce fashion’s environmental footprint while increasing profits. The data can even indicate where the brand should deliver its products, i.e. the most suitable location for a particular product; this eliminates the need for unnecessary and expensive logistics and reverse logistics. The role of unique and comprehensive data sets (including consumer habits and product attitudes) is critical, especially in these challenging times and beyond, when efficiency alone will determine who succeeds. However, my personal conclusion is that current business models in the fashion industry will never prevent brands from obtaining a large amount of reliable and relevant data on product-related consumer behavior. While we swim in an ocean of personal data, there is a fundamental problem of accessing it and turning it into valuable analytics for fashion brands. Is it really a matter of data? No one wants to promise that data will solve all fashion problems. Data collection itself depends on the technology used and business practices and models. But I’d like to start a conversation about it. Fashion is a consumer industry and should be able to collect data while respecting consumer privacy. Data availability will always depend on the business model and tool that connects fashion brands with customers, as well as any third-party intervention in the business. We need a unified, highly intelligent technology solution that can synchronize the entire data roll to capture different behavioral patterns at different levels of the consumer journey and deliver them directly to the fashion brand. In addition, this solution should be able to provide the same data and customer connections for offline retail. Even if every brand decided to overcome its challenges and systematically digitize its inventory and processes, it would not provide the desired connection with customers or the right amount of disparate data to accurately observe and predict consumer behavior and future trends. But whether it’s putting together a collection, rethinking strategy or making decisions about the customer experience, all of these actions need to be based on more specific, relevant and reliable data. This will solve everything, at least in the near future. So is it possible to develop something revolutionary, a product that meets the many needs and demands for access to reliable fashion data? First publication on SCROBLE.com Anna Salewski Founder and CEO, creator of @SCROBLE The founder and CEO of SCROBLE is a passionate dreamer of innovation, business models and disruptive technologies, especially those that improve social impact and quality of life.