How To Win Digital Consumers Through Inflationary Times

This is a talk by our Founder & CEO at a global event on How to win digital consumers through inflationary times.

Please find below the video recording and full transcript.

Here is the full transcript

Ganesh Subramanian, Economy and Need To Win Digital Consumers

First things first. Is recession going to come or is the inflationary trend going to remain? So here is the IMF forecast where the GDP global growth rate, which was 6% in 2021, it’s already dipped to three s something and it’s going to dip further into 2023.

The dotted line is the adverse scenario which can go as low as 1%, hopefully not, but as you can see there is a recovery in 2024, coming back to the 2022 numbers. For definitely there is a dip into 2023. It is good for all of us to be prepared.

The question today is how do we manage our business and what should be our key north star directions, particularly from a product perspective because the product is the hero in any fashion and lifestyle business and for that matter, any retail business.

This is a HubSpot research, let’s look at some numbers. Consumers are asked in the US how has this news or potential recession impacted your spending habits? As you can see here, more than close to 60% of the consumers are already conscious and they’re avoiding purchasing or they’re purchasing less.

So we are talking about close to 70% of that already happening. Another question was if a recession is actually declared, how will your home budget change? This is in case it happens. Even here you can say that close to 65% of the spending is going to come down. Therefore, close to two-thirds of consumers are going to either consciously spend or spend less.

Now that’s definitely that a reality which is hitting us pretty soon and with our lead times of planning next year, significantly high. So it’s important for us to see how do we manage this?

Now, what’s the impact of this and where consumers are going to spend more and spend less? Now it’s common that discretionary spending will be go down and it’s going to impact two dimensions: one, where will the consumers spend their money; and second is the frequency of purchase?

For example, already leading retailers in the US, they have announced that their consumers have started spending more on essentials and less on discretionary category and fashion falls under discretionary categories and hence there is going to be an impact and the frequency of purchase is also going to come down.

Impact for a Brand or Retailer

Now, what this means to a brand or a retailer? One is, when the demand is going to shrink a bit, there’s going to be minimal resources because there are some areas we will not be able to cut down and hence most of the brands are thinking about right now rationalizing their assortment.

What does that mean is, basically you’re shrinking the assortment and when you shrink an assortment, you’re not going to have a high margin of error. Now this is always true, but it’s more true right now. I keep saying that whenever the water level comes down, the stones are visible and hence, we are going to reach that scenario where the resources are going to come under constraint, therefore what to do and what not to do? And how do you choose the right thing to do?

That’s going to be very, very important from a product selection perspective. And it starts from design creation into buying, and what are we going to spot from a trend perspective, and what are you going to adopt out of that? And how responsive we want to be, how proactive we want to be, how reactive we want to be?

It’s going to be important, but we are going to be going through a phase which is not going to be very certain, it’s going to go through some phases of uncertainty. So I would say it may not be as bad as COVID, but definitely making ourselves nimble and responsive is going to be important.

Now, proactive approach, then waiting for things to happen is definitely the right way to do. And if you play this very smart, not only that, we will stay afloat and actually then hire. Because there enough research that people who exploited the recessionary periods, this is after multiple recessionary trends over long periods of time, the best of companies including Google, started during recession periods. So this is the time for incubating. Incubating ideas and also cleaning up the entire system.

And those companies who do that well during this period come back as a winner once we bounce back out of the recession.

The Winning framework

Now how do we prepare to win during this period? I call this three A framework. Now, which means that we are going to be doing less, not more, hence we need to be choosing better. If you’re choosing better, we need to be more accurate.

Now, this accuracy plays out across the value chain. Today I’ll focus a lot on product assortment, but of course one needs to be accurate across what do we make and accurate also for whom do we make, and where do we put? It’s relevant across the value chain, but we’ll focus a lot more on the accuracy of assortment.

Second, not only accuracy is important because we can’t be accurate, but take too much time to get there, hence agility is going to be important. For can we be faster? Can our supply go-to-market calendar shrink during this period better than ever before?

It’s all the more because whenever you have uncertainty, it is good to crash lead times. I do know that we have lots of supply chain disruption, but I think agility even at a slightly higher cost would pay off, compared to a rigid long lead-time supply chain, for sure.

And the last one is how does one adapt? When things change frequently it’s just not the speed and also constantly correcting on changes in the real world environment. For it comes together, for if you are accurate, agile and adapt, the probability of winning is really, really high.

Of course everybody will face the challenge, but those who do this really well will come out of the crowd, and also gain a larger share of the available wallet which is shrinking.

With that, we look at what is the North Star for accuracy? Let’s go through all the three dimension one by one. The singular North Star is “consumer at the center“. Now we have been used to consumer obsession is one way of North Star, but I think the direction is can we be our customers? Can we be our consumer?

This means so deeply we understand them that we are bringing propositions which they completely love. They really want to have it and you will have the lowest amount of supply demand gap.

Now easier said than done, I think we are going to go through some ways that this can be done, may not be the only way, but definitely I’m going to share with you what could be that way, which is definitely better than the current baseline. For this is what we enable at Stylumia, where we bring in a consumer eye for your decision-making, right from the trend spotting to buying and merchandising across the value chain.

Now how does that work? So we created huge impact so far, but it’s pretty small in the overall scheme of potential which is available. Huge amount of products , go into either markdowns or into landfill in the industry. We improve the inventory turns, and full-price sell-through. Any brand would look for these days. The key question is how do we be more effective for an invested capital in products?

Need For A Fundamental Shift

This calls for a fundamental shift. We are calling that as a mindset of B2C to C2B. We do take a lot of decisions on behalf of our consumer. Now the question here is can consumer come upfront? Now what does that mean? Is there is today a way of trend spotting, understanding what’s happening, what will happen in the future.

Now when we started Stylumia six years back, we did evaluate all the current tools and the current practices. We could appreciate the amount of challenge which is there in managing and forecasting this business which is so variable in demand, and also having a long lead time and that’s the biggest challenge in the business.

Having said that, what we saw was all the importers are getting into the forecast of the future, where most of them are supply driven in the sense all our observations are what is available, right? And what is available is half incorrect.

Just to give you some numbers, where out of 150 billion garments made in the industry every year, 50 billion garments go to landfill, right? That’s a huge amount and fair amount of them don’t sell at full price and roughly one out of two do not do that.

We have a coin toss probability of success as an industry and we do waste a lot. Now the question is, can we bring consumer insights into the process on an ongoing basis to improve this probability of success? , and that’s a C2B proposition, and this is what we talked about where huge amount of products get wasted, and how can we use consumer data?

Again, easier said than done, how do we get on a dynamic basis, consumer data? I can only get what’s available. Brands don’t share their data with all of us. You can scrape any data from the websites or any digital sources online, but that’s one way. There are tools available like that.

Our only question to you is when you see all the data, what others are doing, that’s only what they’re supplying. How do you know their demand? Not knowing the demand is a huge error, and I’m going to share with you some of the accuracy differences between following supply and following demand.

Now, if you have to solve the problem of getting the relevant products and minimizing wastage, you have to solve two problems. One is the, what problem, which is what to spot, what to make, which is the product selection, creation, assortment problem.

And the second one is how much to make, which is a quantitative problem. We can make, there is no bad product if you made the right quantity. I’ll just say that again. There is no bad product if you made the right quantity.

In other words, every design has its own threshold volume that it can sell. If you make the right quantity, we’ll all have very good full-price sell-through. Hence, how much is as important a problem to solve as what. So we’ll go through the, what problem to start with, and how do we do that better augmenting the existing process.

One of the solutions is the consumer intelligence tool. We work with over a hundred brands and retailers globally and what our customers, what we provide is, we give them demand insights, of across various dimensions by geography. Of what’s working, what’s not working, not of today, and also predictive for future and what the future trend would look like.

I just wanted to share with you how that engine works. It is one-of-its-kind Demand Sensing Engine. It collects data at internet scale. It’s basically scraping data from various sources of inspiration and competition which could be contextually relevant for you.

And we gather lots of demand signals for each one of these products on an ongoing basis to come out with not just trends, but winning trends. And winning trends in a contextual way and trend for a particular geography may not be the trend for another geography. Therefore, the current ways of looking at color forecast of next year, that’s too broad brush, right? It’s not very contextual.

For example, let’s say trends for next season for the whole world, how can that be true? We don’t see that in the demand patterns. I think demand tells a very, very different story than what the supplies and all dimensions from the higher levels of the product hierarchy to the lower levels up to the sku level of demand.

So that’s a fundamental engine that brings all of the demand level insights. Now I just want to take you through how a brand or retailer has used it and the kind of impact for example, an Omnichannel brand, one of the brands that we use where they improved their full price sell through by 20%. Now, what that means, it generates, in fact it almost doubles the profitability. Otherwise, this 20% would’ve got 50% discount, which is 10% and the EBITDA of fashion brands hovering around that mark across the world. You almost double the profitability by doing this, and increase the revenue velocity by 30%.

And why does this happen? It minimizes the combinations of attributes; the recipe that we use in making the products and improve the probability. It starts with the understanding and compared to traditional static research, it’s a dynamic research. Therefore, your constant understanding of what’s happening in the consumer space, which is very, very important particularly during these uncertain times, as we know that while all of these predictions remain, we know it may not go in the same direction so that you can make some fine-tuning.

You can do this at internet scale. There are customers that we have where we enabled them across 29 countries and the brand is based in the US, and how do you understand demand of each one of these countries so that you are very relevant in Japan, You are very relevant in Netherlands, very relevant in Germany, so that you’re not having a holistic common assortment but not at the same time, not very, very geo-relevant.

Understanding these nuances is very, very important. So not only you take product decisions, pricing is going to be a super critical attribute during this period, because inflationary times one needs to also right price the product. Not getting the attributes but also right price the product. And of course in season you can take promotion decisions.

For there is going to be a huge value unlock in investing in technologies which will help you understand consumers better. The idea is only your own data has limited knowledge., getting the knowledge of what works for the market and what doesn’t work so that it’s the intelligence of the crowd which we will get into your overall decision accuracy.

Now, if I have to go through what are the key traits during these times within an organization: one is, this is through various research that those teams which are passionate about their consumers, and second is applying data and analytics in every area, and today we are talking about how do we bring in a lot of consumer intelligence into the design, merchandising and buying process?, and being there extremely flexible.

These are times to improve the flexibility. They’re already very flexible. The question is to go and raise the bar there. And using digital channels to understand reach consumers because the digital channels give us a lot more data than the rest of the channels and hence, using… In fact for one of our customers in the US market, we are able to do intelligence at the level of zip code.

So we are able to geo-local understanding of taste of consumers, whether it is New York versus Alabama versus Denver. Now, that’s the level of intelligence that one can get these days, and the question is to use that in the assortment creation process or a distribution creation process.

I just want to share with you some of the numbers of whether a demand view of forecasting is better than supply view. This is 2021. I’m going to share also something very, very recent and we are going to publish a report which will come very soon in our website.

If one had followed supply versus demand in miniskirts as a category between ’20 and ’21, you can see the difference, right? Now, you can see almost 100% difference between supply and demand. While supply declined by 41% in luxury, demand declined by 82%.

This demand data is coming from our platform using that engine and in the fast fashion market we could see again, while it improved by 77%, about 133% from a demand perspective. There’s an opportunity loss on one side, there’s an excess of inventory on the other side.

Now, follow that through another season. Similar results, for if you just go… It’s not about just having data, it’s about asking what is that platform we are using today? What does it tell us? Is it telling us supply? Is it telling us demand. And our demand engine now has been validated by global brands and retailers and we have an accuracy of 80% plus.

Of course it’s not 100% accurate, but we are driving the demand and this is an area that we constantly work on. But if you go with the current methods, your accuracy is around the 40 to 50% mark so it is almost coin toss

Now, the question is to elevate the insights that we are getting, which is getting into our entire creation process. In fact, one of our customers… Now what are the impact right now? You improve your best sellers. One of our customers, one of the fastest going casual brands had two to three best sellers out of 10 products. Now they have four to six best sellers. And absolute rate of sale. Now, when you do that, you’re going to lift the overall rate of sale of your assortment that you’re going to have lower number of poor sellers and mediocres in the entire assortment.

Now we talked about so far, how do you get our assortment accuracy? And the next one is about how do we improve the agility of assortment to customers? This is another case study where here we are now solving the, how much problem. So what we just spotted the market, understood the demand, made the assortment. Now getting the demand the prediction, and one of the biggest problem in fashion is predicting new product demand.

Now, predicting continuing product is relatively easier. Of course, it’s challenging; the seasonality and uncertainties, but new product is a bit more difficult. For using our Demand Prediction Engine where the customer was able to see a 30% lift in their prediction accuracy. What that means is equivalent of almost doubling the profit as we see, because any accuracy difference results in under stock or overstock and overstock goes through markdowns.

And while of course, there is always a skepticism of adopting any new model and we don’t promise, in fact we ask our customers to try maybe a small part of your business. Therefore land, take up 10% of your business, adopt this and we will show performance and then we will slowly crawl, walk and run.

This is the engine where imagine that you can throw a new product, it can actually give you more so predicted demand and an ROS (rate of sale) at different levels of your business hierarchy. Now it uses, we have a proprietary engine which goes beyond traditional attributes and it also understand aesthetics which can’t be articulated in language, and that’s a fundamental differentiation and hence we are able to create lift from the baseline.

Want to share with you the kind of accuracy error reduction that we have seen. For example, this is forecast accuracy at customer needs state level of one of our Omnichannel brands where the error used to be around 13% and we are under 5% error, over last many months and continues to go down. It’ll go over, it may not be 0.33, but definitely error within under 5%.

These significant lifts, in fact, this particular customer after seeing a prediction accuracy has started to now buy purely on this forecast, and sometimes the forecast from the model is much higher than their forecast, therefore they’re buying more. In fact, the business is growing.

It’s not about saving inventory, it’s also about getting more business. Sometimes our own assessment of the future could be not as optimistic.

They also adapted fast and from moving from quarterly forecast to now weekly forecast at multiple need states over hundreds of needs states on ongoing basis to do corrections across the value chain.

The key impact areas are, as we discussed, inventory reduction, revenue growth and profit. You can see anywhere between 20 to 40%. It depends on where your baseline is.

I just want to conclude by saying that to win during these times, advanced analytics is a key growth lever and move from static ways to dynamic ways of understanding consumers across alternate data sources. Of course, we are one of those alternate data sources, you could definitely use as much as possible with your partners and get all of this data together to take informed decision.

Keep doing everything with the North Star of consumer at the center. The probability that we will meet our goals will significantly increase and you will gain a lot more market share purely because consumers are going to love your products and what you do.

Thank you very much and if you really want to reach out, I’m at ganesh.subramanian@stylumia.com and wish you all a wonderful day.

Aton:

Good stuff, Ganesh. I think there’s some questions to get into for the final few minutes of this, so it would be good to get your viewpoint on that.

So one of them is about current inflationary pressures. As you might know, inflation is going up especially here in the UK, more than most. Do you think that these current inflationary pressures will have a lasting impact on consumer demands and the trends around consumer demands?

Ganesh Subramanian:

Absolutely. I think as I mentioned before, there are two areas that will create an impact. All the forecasts are saying it’s going to be definitely not a very, very short term. It’s going to be one to two years, and second, this will make consumers to start buying essential stuff more and must-have and less of nice to have.

What that would mean could be different for different customers, which means the assortment preferences like we went through COVID, lots of leisure went up, you’re going to see a lot of those assortment shifts, which is going to happen. If we have very long term forecast, make huge bets many months ahead, we might be in for some shock. Being agile, that’s why it’s very, very important. Adapting to changes is very important and using data, dynamic data to take these decisions all come together.

That’s one of the ways that you can minimize the setback.

Aton:

Yeah, I know that you showcase some of the examples from the real world use case of the consumer intelligence tool as well. That was one of the questions that had come up is do you have any real world examples? I guess some people always want to know which brands you’re working with, but often when I host these spotlights, it’s heavily protected that kind of information, isn’t it?

Ganesh Subramanian:

We do work with Fortune 100 brands and retailers across various geographies, whether it’s US, UK, Europe, and of course we started with India, fair amount of brands.

The key use cases that people use our solutions for is number one, is validating trends. How do I know whether the trend I’m choosing is in the right direction? Will it work or not? For what is the probability of winning, de-risk at that level? Second is in-market opportunities, which is to say that what are the white spaces, which is in the market that are doing well that I missed? For capture those missed opportunities is another use case and with the recession coming in, pricing is a huge use case. How do I get my pricing right, assortment pricing intelligence from a consumer demand perspective is another use case? We started with fashion, we are just going across categories right now.

I just want to share one insight on the forecast accuracy on color. We recently compared one of the color forecasts made by some leading agency, which many brands use, and we looked at all the color palette of ’21 and their prediction for ’22, and I just want to share, we are going to publish a report very soon. The accuracy of that forecast was 14.7%. Only 14.7% of the color predicted were sitting in the top deck of consumer performance. Right?

Aton:

Wait, sorry, just explain that figure to me.

Ganesh Subramanian:

We looked at all the colors forecasted by the agency. We said, “Okay, these are the 62 colors across various color groups. These are colors that you should use.” That was the recommendation. We went and studied actual consumer demand in the same season for prediction and said, “What are the winning products and what are the colors associated with them?” Converted all them into the same base, compared the colors, and then came out with this number. Then we actually looked at, from a purely demand perspective, if you have done the forecast from ’21 to ’22, we are getting 80% plus accuracy.

What I’m just saying is using data, this is only color and product doesn’t sell well or not just because color is an important attribute. So, we are going with lot of forecasts which are not consumer driven. I’m not saying that the expertise and everything is not good, but the question is how much of that is consumer data driven?

The more and more we bring consumer data into our decision-making, we improve the accuracy. I just want to elaborate that with the use case from real world. Hope that helps.

Aton:

Yeah, it really does. I mean it’s a staggering figure. How can you, I guess with sensing, you could demand sensing. That’s how you up that figure, right?

Ganesh Subramanian:

Yeah, absolutely. We know what’s the reality, we know what the forecast was, and you can match colors with the RGB and color codes and pantones. Now you can actually see how much of them really panned out.

Yeah, I think what’s very important is we are all responsible for our own forecasts, and that’s something it’s true for us. Anybody who sends out forecast should also come back and say, “This was our forecast. This is what has happened, this is how accurate we are.”

I think we really need to get transparent because we are all leading brands, because we are very responsible and very giving information to brands to take decisions, and we also need to take ownership of what inputs we provide and how accurate or inaccurate they are.

I think the intent is to also continuously improve, because if all of us become accurate, the industry becomes better. It’s very earth friendly with minimal wastage.

Aton:

Yeah, absolutely. Ganesh, thanks so much for your time today. Honestly, over two panel discussions and the presentation, you’ve really provided a lot of insight onto this. So I really appreciate it.

Ganesh Subramanian:

Hey, thanks Aton, and thanks PI Apparel for organizing this. Wonderful.

Aton:

Yeah, it’s been great. All right, well, I know it’s late there, so you want to go to sleep. So, go and have a nice evening and yeah, we’ll see you soon.