facebook Created with Sketch. instagram Created with Sketch. linkedin Created with Sketch. twitter Created with Sketch.

What Makes Super Fashion Forecasting Possible?

| 4 min read

Philip Tetlock, a Canadian American political science writer, spent two decades measuring people’s ability to predict world events. As per Tetlock, a small set of superforecasters were able to predict better than even professors from top universities (experts) by a margin of 70%. A dig into this research tells us what makes super fashion forecasting possible.

It is interesting to note that these superforecasters did not come from top universities or any special backgrounds.

What Differentiates Superforecasters?

Before we get into how to make super fashion forecasting possible, we need to understand what makes the superforecasters super.

The superforecasters were not smarter than others. They also did not have more knowledge or experience than everyone else. In most cases, they were amateurs. Inspite of this, they outperformed even the CIA (Central Intelligence Agency). They beat the CIA by 30%.

The secret is,

What made superforecasters so great at being right was that they were great at being wrong

Julia Galef

Being Great At Being Wrong

What does it mean to be great at being wrong.

The superforecasters changed their mind all the time. They do not sit with their view that they are right. They change their mind and make subtle revisions as they learn new information.

People who update their forecasts in tiny steps are superforecasters compared to those who fire and forget.

Tim Minto, the highest-scoring super forecaster changed his mind at least 12 times on a single forecast and sometimes 40-50 times. He is constantly course-correcting like a ship captain.

Every dot in the graph represents a time Minto revised his forecast over the course of 3 months.

They are Bayesian thinkers, who constantly update their probabilities based on new information.

Fashion Forecasting Status Quo

Fashion forecasting of today is mostly fire and forget. Try searching google fashion forecast for 2022 or the like, you will see some reports showing up. This is how static reports and most of them are based on subjective inferences. Following such trends is costing the fashion industry heavily. With over 50 billion garments of wastage every year due to ill-informed trend forecasting, fashion forecasting, and demand forecasting, we have a serious problem on hand.

Do not go by the claims of what people say “World’s #1 Fashion Forecasting etc…”. Pedigree does not matter in forecasting as Tetlock discovered. We rather need a child’s mindset of constant learning.

If you are referring to such static reports, time to pause and re-invent a new way.

What Is the New Way For Super Fashion Forecasting

The key is to have an “UPDATE” mindset. The update happens with every new information. This needs information that is a reflection of True Demand for the update. This leads to the question of how to get True Demand information for fashion. This is not public information.

We at Stylumia always wanted to enable the ability of super forecaster to every brand and retailer across the world. We built a unique one-of-its-kind Demand Sensing (detailed article on Demand Science® here) engine which does the job of dynamic update of fashion demand information across the world. Such a dynamic update eliminates the confusion and noise from the subjective information.

If you are considering or using tools that give you data, it is important for you to look at whether the data represent supply or demand. Supply information in fashion has an over 50% error rate. You will have the same probability of error when adopting insights from such platforms. Data alone is not good if noise is not removed (check a detailed article on fashion noise here).

Stylumia’s Consumer (Period) Intelligence Tool C.IT, gives you a dynamic trend like a Super forecaster.

dynamic demand trend from stylumia consumer intelligence tool using demand science
Dynamic Demand-Trend From Stylumia C.IT

It is important to have the forecast calibrated with outside-in consumer intelligence and inside-out intelligence from omnichannel information. While C.IT helps you get the outside-in view, Stylumia APOLLO helps you discover True Demand from your omnichannel data. Sales information is a means to find True Demand and by itself is not the demand. This is a nontrivial problem.

The evidence is in our clients see their forecast accuracy improving upto 30-50% from the current baselines and staying relevant even through the covid period.

Stylumia APOLLO comes with a unique taste model which predicts the demand of products given the constraint of an assortment. For example demand for a blue sweater without any blue in the assortment is very different than when you have other hues of blue in the assortment. This is a simplistic view of the complexity to be handled to predict the dynamic demand of products across the omnichannel.

In Conclusion,

It is important to build capabilities of super fashion forecasting within your brand organization and capture the maximum share of consumer demand. With free cash flow of paramount importance in building long-term value, you need to choose the right tools which will enable you to have the capabilities of a super fashion forecaster, trend forecaster, and demand planner.

These are times to move on from the old to the new even if it’s uncomfortable for the right reasons. Let the endowment effect not impact your decisions.

This is also the time for the fashion industry to re-consider the tools on which we are placing our future bets. Question the fundamentals.

If you would like to have a free demo of our solutions, please feel free to reach out here.

Subscribe to our insights

Subscribe now to receive our thought leading insights right into your inbox

Related Blogs

Brat vs Demure: The Ultimate Trend Face-off

| 4 min readIn the fast-paced world of fashion, trends come and go, but two are currently stealing the spotlight: the brat and the anti-brat, or as Jools Lebron likes to call it, the “demure.” Although these trends have reached saturation on social media, they’re still alive and kicking in the fashion scene. For brands and retailers, they […]

on September 24, 2024

Case Study: AI Strategies to Create More Hero Products

| 4 min readExecutive Summary In an innovative move away from traditional market expansion strategies, a leading women’s wear brand sought to escalate its growth not by increasing product selection but by amplifying its ‘hero products’. These products, known for outsailing others by 10 to 20 times, represent significant business potential. Facing the limitations of conventional market insight […]

on June 9, 2024

Adaptive Retail Growth Strategy Framework

| 6 min readUnderstanding the Adaptive Retail Landscape In today’s volatile marketplace, consumer preferences and needs evolve at an unprecedented pace, necessitating a responsive and proactive approach from retailers and brands alike. To not only survive but thrive, it is crucial for businesses to not just understand but consistently exceed consumer expectations. This dynamic adaptation is the cornerstone […]

on June 9, 2024