Here’s My Thesis For Starting A Profitable Fashion / Apparel / Fashion Tech / Fashion CPG Company With My Experience As A CEO / CMO / COO / Chief Data Officer And Founder Of One of The Worlds’ Largest Privately Held Fashion Brands
My name is Steven Evangelo Michas, and in 2012 I Founded one of todays’ largest privately held fashion companies. In 2012, I Founded 90 Degree by Reflex after my own due diligence in the Fashion industry. I started a Private Equity company, and I identified a partner company who was in the textiles/manufacturing business, and I took an equity operating, and ownership role in the Manufacturing business. With the Manufacturing piece in place, I founded 90 Degree by Reflex, and I grew the Manufacturing business from $15m USD per year to over $350m USD per year with full P&L experience, and I grew the eCommerce division of 90 Degree by Reflex from $0 to start to over $35m per year USD, with 2 years. Within 7 years, the entire business was generating over $350m USD per year, and I held full P&L responsibility for the entire business, and I generated all of the deals using Paid Digital Media Ads, and I use my Sales experience to gain Big Box Retail partnerships with Walmart, Nordstrom, Marshalls’, Equinox, and other leading brands.
In total, as a C-level Executive, and Member of many different Board of Directors, I have generated over 4 Billion USD in direct revenue, and I have generated a few Billion USD in Downstream Revenue over the past decade.
While I was a C-level Executive, I generated more than 2.4 Billion Unique Visitors to various websites, and I generated in excess of 540,000 Reviews across multiple platforms.
The goal of my Thesis is to create a Fashion business with the same proven Methodology I used to take 90 Degree by Reflex to $350m USD per year with full P&L experience.
My personal experience is in complete Digital Transformation, Growth Hacking, eCommerce Transformation, PE, VC, M&A, Digital Marketing, and Data Science Engineering. With these skills, I use Digital Paid Media Ads, and Data Science algorithms I made to acquire sales within a standardized, targeted budget; then I scale revenue by increasing already profitable Paid Media spend while also using AI, and Machine Learning algorithms, to make newer profitable Paid Media campaigns.
I have also Founded every type of business Department as a CEO, CMO, COO, and Chief Data Office, with full P&L ranging from a few million, to over $250m USD per year. I consider myself beyond Expert knowledge in every area of business, and I have Founded and scaled each department in many different companies, in many different niches, multiple times with over 17 years of experience.
I’ve also acquired, and merged, many different businesses over the years, and I’ve used my deep M&A experience to find further Profitability in existing business lines.
The end goal is to build a business around the following KPI: average cost per new sale (CPS, or CPNS), using Digital Paid Media Ads, and this is calculated as follows:
Average Order Amount – Cost of Goods – Total Operating Cost – Cost of Paid Media = Net Profit
Simplified, the Calculation is:
Average Order Amount – Total Cost of Goods/Operating – Cost of Paid Media = Net Profit
Since the emphasis is on building a strong Ecosystem business model, the next important calculation focuses on Profit Per Customer and then measuring how much Profit Per Customer is generated each Quarter from repeat sales.
Once a customer is generated, substantially less of the Paid Media budget can be used to gain repeat sales from existing customers, which generates consistent revenue each Quarter. Additionally, some existing Customers will also re-order without any Paid Media budget spend necessary to display ads to them, resulting in immense boosts in Quarterly Sales.
The calculation for determining Profit Per Customer is:
Number of Transactions * Total Order Amount of Each Transaction – Cost of All Goods – Operational Costs – Paid Media Cost = Net Profit
Then, reporting parameters are configured to display Net Profit for a specific range, such as Quarter 1, Quarter 2, etc.
Reporting also is more complicated and accounts for more factors, such as if Paid Media was used to trigger a sale, if a sale happened without Paid Media, etc, to gain understanding of how much total Paid Media budget should be allocated to New versus Existing Customers.
This calculation, and how it is calculated across multiple product lines, and channels, is the competitive advantage behind the entire business model, and this Thesis.
I’ve used this Thesis, and this calculation, and my operational execution experience as a CEO in over a dozen different industries, to generate over $4 Billion USD in profits, with full P&L responsibility.
As CEO, I had full responsibility over all other departments, and due to the power of my calculations, and execution of it, I’ve simultaneously held the positions of CEO/CMO/COO/Chief Data Officer, and was overseeing all organizational departments, while maintaining the highest standards of excellence, organizational culture, and EQ.
My Thesis is focused on creating an Ecosystem driven business, and my Thesis will focus on the following advantages, and natural competitive advantages, which are designed to evolve over the next 7 years:
1. Paid Media / Application of The Above Calculation / A+B Testing / Funnel Development
2. eCommerce / Reviews / Front Facing Reviews To Leverage Big Box Partnerships / SEO
3. Data Science / Reporting / Marketing Technology
4. Fashion Tech / AR / VR / 7 Year Showroom, and eShowRoom / Conversion Rate
5. Manufacturing / AR Driven Robot Manufacturing / On Demand Custom Prints
6. Manufacturing Requirements To Make The Suggested Clothing
7. AI & Machine Learning Paid Media Algorithms
Section 1:
The power of the calculation is that it is both A) Rapidly Scalable B) Applicable to many different products at once C) Duplicable to new, and existing products D) Deployable across many different channels, such as Google Ads, Amazon Ads, Email Marketing, etc E) Reporting friendly, and software can be built to calculate the profit behind every product on a per product, and per Paid Media campaign basis F) Continually improvable with cheap developments in AI & Machine Learning, and with existing customer data G) Easy to implement, and can use this data to calculate other key KPIs for acquiring sales with Paid Media
Example 1:
Here are real world examples from my previous Fashion businesses.
After looking through reports, I determined the Average Order Amount for a certain type of Mens’ T-Shirts from the eCommerce website to be approximately $54 USD.
With this information, I was able to identify the existing profitable Paid Media campaigns I created, and then I increased the daily budget, which results in more sales, since more traffic was available.
Example 2:
After I increased the budget of the Paid Media campaigns in Example 1, I made new Paid Media campaigns for the Mens’ T-Shirts with the Average Order Volume of $54 USD.
When I create campaigns, I can adjust my own campaigns, or program my AI & Machine Learning algorithms to acquire new sales within a specific dollar amount.
Example 3:
When I create a new ad, if I know the Average Order Amount of my eCommerce website, or a specific Product, then I can train my algorithms how to spend money per campaign to acquire a sale.
I can use things such as stop triggers, start triggers, and different strategies to control when, where, and how I bid on specific Paid Media Ads.
I can also include any cost of goods, operational costs, or additional costs into how much I spend to acquire a new sale.
Ultimately, I can also run Paid Media without AI & Machine Learning algorithms, and the goal would be to always acquire a new customer by spending LESS THAN a certain amount per sale.
Example 4:
A real world, specific example would be:
If a Fashion website had an Average Order Amount of $49, an Average Cost of Goods of $12 per Sale, and Average Operational Costs of $2 per sale, then we can conclude:
49 – 12 – 2 = $35
There is a Net Profit of $35 per new sale sitewide, and so we can spend anywhere from $0 – $35 to acquire a new customer.
Example 5:
Each product varies per cost of a new customer, and using my reporting, I’ve figured out ways to determine the quality of a potential new customer so that I can determine if it’s worth spending more or less per new customer based on their INTENT to reorder.
In some instances, it can be worth more to spend more to acquire certain customers, since these customers are more likely to reorder from the eCommerce website in the future.
Nevertheless, it is possible to plan for a specific Net Profitability per customer by setting hard limits on the amount of Paid Media spend used to acquire a new customer.
Example 6: As I will demonstrate later on, the Average Cost to Acquire a Customer will drastically decrease as each product gains more reviews, so it is possible to actually end up with higher Net Profits over time, since the conversion rate of the website will increase.
When the conversion rate of the website increases due to more reviews on specific product pages, then the Cost of Acquiring A New Sale will improve.
This is best visualized, and is described as follows:
If a specific Product has a Conversion Rate of 2%, and this Product has 40 reviews, then when the Product has 300 Reviews, I expect the Conversion rate to increase over time as the number of Reviews increases.
If all factors remain the same, and the Reviews increase from 40, to 300, the Conversion rate of this product can increase to 4%. This demonstrates a Realized Gain of 2% Conversion based on previous sales on this Product from Reviews that will always be there.
This is vital, and shows how Total Revenue grows over time, and benefits from the Reviews of previous Customers. This demonstrates how previous Ad Spend on Paid Media results in increased Revenue, and my Software calculates, and projects, all of these values.
Section 2:
One of the largest natural competitive advantages of this Thesis is large number of eCommerce reviews that are generated as a result of the Paid Digital Media Ads.
As the reviews are generated, they can be leveraged to built Trusted Partnerships with Big Box Retail outlets and other Distributors, including Online and Physical Retail.
The way this works is:
Example 1:
1. Paid Digital Media Ads are run using strict parameters to ensure sales are acquired within a specific budget
2. As Sales are gained on eCommerce channels, the Number of Reviews on each Product Page increase. As Reviews increase, I take the Product Pages with large number of reviews and bring them to Big Box Retail outlets such as Walmart, Target, etc, and make the case that these products are already successful, rank well in Organic Search, and are already in demand.
3. In my experience, Big Box Retail Retail outlets are more inclined to work with already successful brands, and the largest number of Reviews generated provide Proof of Concept, and show Trust is already established, and Sales are constant; this is a positive attribute since Big Box Retail outlets are more willing to sell already successful brands, and products, with Proof of Concept.
Example 2:
The other natural competitive advantage here is successful products and Product Pages typically rank well in Search Engines and Organic Search.
I am an Expert in SEO, and I have done SEO for global brands and I can implement, and execute, SEO methods to rank successful Product Pages with many Reviews on the top of Page 1 in Google, Bing, and Yahoo,
Typically, Big Box Retailers will use Analysts to evaluate the SEO practices of potential Partner Brands. If the SEO Analysts from these Big Box Retailers think the SEO is natural, and not “Spammy”, then they will recommend the Brand for Resale, and Reseller Partnerships.
I can implement natural, strong, Organic SEO to rank Product Pages at the Top of Search.
Many Brands never get a Partnership with other Big Box Retailers, and they don’t usually understand it is because Big Box Retailers employ SEO Analysts to check on how their potential Partner brands not only use SEO upon first inspection, but during the Lifetime of the Partnership.
Example 3:
As positive, natural, real, organic Reviews are built for Products, I will supplement this with long-time, positive, natural, and real Organic SEO efforts that will pass SEO Analyst inspection from Partner Resellers and Big Box Retailers.
Section 3:
This strategy, naturally, requires a robust, and substantial Data Science, and Marketing Technology division to implement the algorithms used to run, and support, all of the Paid Media Ads.
The Data Science, and Marketing Technology Division also must be built to provide superior Business Intelligence information, and reporting, to each employee to empower them to learn to make autonomous decisions.
The types of reporting I can create are excellent, and will take a lot of the guesswork out of jobs, and each employee can be empowered to make decisions supporting the goals, and culture, of the organization.
I have personally created, and oversaw development, of many Websites, Data Science Projects, Databases, Servers, AI & Machine Learning Projects, Technical Projects, Security Projects, Network Security Measures, and Development of Software, Algorithms, and all things IT.
The Data Science Division would focus on 2 main parts:
Example 1:
The first part of the Data Science Division is to develop algorithms and reporting designed to run, and manage, Paid Media Ads.
Of course, it is possible to begin using Paid Media to generate sales without all of the Data Science Division projects finished.
Example 2:
The second job of the Data Science Division is to built the infrastructure necessary to perform comprehensive split testing, and A/B funnel testing.
The Data Science job will also build many different funnels, and develop algorithms designed to show and serve these funnels and split tests to customers at key, and dynamic, parts of the customer lifecycle journey.
All of the funnel A/B testing will then be turned into Business Intelligence, and then the information will be used to improve Paid Media, UI/UX, Reporting, and Customer Experience.
Example 3:
The third part of the Data Science Division is to develop extensive Reporting based on all business activities in each department.
Reporting would be cross-divisional and existing, and new, software will be used to generate reports for every employee based on their individual KPIs.
The purpose of the reports is to give parameters that explain how to stay within budget; this allows for creative thinking, and autonomy, though keeps decision making, “tidy” within key parameters.
Example 3:
The Algorithms and Data work necessary will include basic Excel, MYSQL, Tableau, AI, Machine Learning, Regression, Neural Networks, LSTM Neural Networks, Linear Regression, and Standard Deviation.
Example 4:
All these algorithms, and all software will be built into a Visual user interface.
Section 4:
One of the most important KPI metrics on any website is the Sitewide Conversion Rate.
The higher a Conversion Rate is, the most Revenue is generated, with less Budget Spend.
Fashion is a highly visual, and artistic focused niche whereby customers expect superior and excellent Visual presentation of the products they intend to buy.
The 7 year goal of the organization will be to begin with showing Products on an eCommerce site on a model, with a static, stationary picture, and then to gradually begin using Augmented Reality, and Virtual Reality, to present products in a 360 Degree environment customers can use to interact with the product, such as by zooming in, and rotating a product. Video will eventually be used to show models walking, moving, and talking about the product on the eCommerce website Product Page.
Example 1:
Increasing the quality of Product presentation will directly increase product conversion rate. For example:
If 100 Visitors go to a website with an Average Order Amount of $73 USD, then:
– If the Sitewide Conversion Rate is 2%, the Gross Profit is $146 USD
– If the Sitewide Conversion Rate is 5%, the Gross Profit is $365 USD
All things considered, the site with 5% Conversion rate would have made an extra $219 since their Conversion Rate is higher.
Conversion Rate is usually increased due to Quality Content such as Pictures, Descriptions, and an Intuitive User Interface. Other attributes increase Conversion rate, such as: Brand strength, Number of Reviews, Ranking Top of Page in Search Results, Customer Recommendations, UI/UX, and Page Layout.
It is not possible to buy a higher Conversion Rate, and, instead, much testing must be done to achieve a higher Conversion Rate.
The purpose of the Data Science division is to use deep funnels to determine how to achieve the best Conversion Rate from changing different variables across the website.
The Data Science Division is deeply rooted with Web Design, UI, and UX, since they need to develop an Agile method of Predicting who to show different versions of a page to. In many instances, the Prediction Algorithms must determine what it thinks it the best version of an existing page is to show a Visitor. The Prediction also is responsible for creating New pages, and then rotating these New, and Older pages, to test and learn what Pages perform the best across many different segments and channels. Accomplishing this is a modern feat of
Example 2:
Fashion is highly sensitive to Technology, and, as a result, Fashion can benefit dramatically from Technological advancements, such as Augmented Reality, and Virtual Reality.
As the Founder of 90 Degree by Reflex, I achieved a high Conversion Rate when I used more Virtual Reality, and Augmented Reality on my Product Pages to display products.
The 7 year Fashion Technology plan for this Fashion Business is as follows, and the goal is to increase User Experience, while increasing Conversion Rate:
– List Products on the eCommerce website with high quality, single shot photos featuring a model wearing the clothing. Single shots of the clothing without a model should be used
– Use Augmented Reality to feature models, and clothing, in highly interactable environment. AR, and VR, allow for Customers to zoom in, rotate, and see Products in a more personal way. AR Technology is also able to perform on-demand customizable functions such as changing product color, and displaying product on different models.
– AR is great for displaying what I call Run Way shots on the eCommerce website. Customers can see models walk down a platform, and turn in different ways to show case product.
– Virtual Reality can also display models and product in a Run Way environment, only with Virtual Reality, it looks as though the Customer is participating in a Fashion Show by watching models walk down a Run Way. This happens in real time with VR
Very few Fashion companies incorporate the use of AR, or VR, and this is a natural competitive advantage, and is forecasted to be an advantage for at least the next decade.
The goal is to combine the Cost of Acquiring a Sale calculation, Profit Per Customer calculation, and the A + B testing of the Data Science Division, with the focus on increasing Conversion Rate from the use of Augmented Reality, and Virtual Reality.
Section 5:
Manufacturing of Fashion products requires, at least, access to a wide range of machines that can sew, and stitch.
It is not necessary to own these machines, since Production Time can be bought, and Run Time can be bought. This ensures lower barriers to entry.
However, some of the more in-demand, and hence more profitable products, require specialized machines that can sew beyond the basic stitches made by basic machinery, or older machinery.
Getting access to these machines is necessary, and it is possible to do so with enough lead time.
Example 1:
I already know the stitching requirements for every garment, and I have understanding as to where these Manufacturers are located
The long term strategy of the brand would be to slowly transition to using our own AI based sewing robots.
These sewing robots are pre-programmable to perform every type of stitch, so it reduces sourcing conflicts, and challenges, drastically.
In addition to using AI robots for stitching, we can build out an extensive On Demand business that is able to accept custom orders with no minimum requirements.
This is important since we can use Paid Digital Media Ads to run Ads to customers to allow them to design their own Custom garments, such as shirts, pants, shorts.
With a Custom Demand business, we can also accept large batch orders, and custom CAD, and 3D Printing Orders from commercial clients, and we can offer shorter lead times for production.
Section 6:
The business is ripe to transition to AI robots, and we can drastically reduce the need for human manufacturing labor.
It is possible to transition to a nearly 100% human labor free production line with the introduction of AI robots.
We can configure the entire production line to work with AI robots, from start, to finish.
Example 1:
It is possible to calculate the cost per garment at various stages of the manufacturing process. With 50% of the production line automated by AI robots, the cost per garment can decrease by as much as 15% – 30%.
I would be able to Found and create the Engineering team necessary to install the AI robots, or we can Partner with existing providers in the early stages of growth.
Example 2:
We could achieve 50% AI automation within 2-3 years, and save as much as 15% – 30% per garment due to AI robots.
Section 7:
The entire strategy I propose is Founded around my Digital Paid Media Ad strategy, and I have personally developed this Methodology over the last 17 years.
As mentioned earlier, this strategy is driven to use Paid Digital Media Ads to rapidly scale growth using Paid Ads. The Paid Ads are bought, and they drive targeted, research traffic to an eCommerce site.
In this case, we will use Paid Media platforms to send targeted Ads to our eCommerce platforms.
I will perform due diligence to study how much it costs to send 1 Visitor to the eCommerce website. Once I know the cost of 1 Visitor, I can begin to calculate how much it will cost to acquire 1 sale.
Once I understand how much it costs to acquire 1 sale, I can begin to increase the Paid Media budget with confidence that the advertising budget will produce a knowable, predictable, Return On Investment.
Over time, as the number of Reviews increases, the Site Wide Conversion Rate increases, and as the Conversion Rate increases, it will cost less to acquire 1 sale.
The entire effort is to calculate the Cost of Acquiring A Customer, then to calculate the Lifetime Profit Per Customer over set intervals, such as Quarters, Weeks, or Months.
I have previously created my own custom, proprietary software using Python, AI, Machine Learning, and Ruby to perform many functions. The goal would be to re-create this Software, and all of its Reporting features.
My software performs many functions, and the Paid Media Software can: Calculate Ideal Bids on Paid Media Platforms, Adjust Bids, Recommend New Ways To Run Campaigns, Recommend Changes To Existing Paid Media Campaigns, and overall, the Software performs complex Mathematical, and Data-Driven computations to find the lowest.
The software also has a Reporting piece, and it is designed to Report on crucial KPI information, such as all of the calculations discussed, and many more channel, and segment, specific information.
Wow, great info!! Thanks for sharing. Very inspiring stuff.