Winning Strategies and Business Models

You’re an entrepreneur with the greatest idea since sliced bread. It’s even better than the invention of the wheel. And you don’t understand why those clueless VCs are hesitant to fund you. As a former CEO of a startup I share your pain. What are you missing? Well, your wheel may hit ruts in the road ahead if you haven’t paid attention to:

  • The size of the opportunity;

  • The profile of your potential customers;

  • How to maximize revenue from those customers; and

  • The soundness of your business model.

In this post and an upcoming keynote to the Canadian Financing Forum in Vancouver on February 21, I will present an analysis addressing these issues for ecommerce companies. However, that same thought process applies to all companies.

On the subject of maximizing customer revenue, it’s embarrassing to admit, but I was the subject of a very creative (but actually worthless) upsell to a transaction I was making online. The basis was a recent traffic ticket. Of course I’m innocent (sort of)! I had gone about 20 years without a ticket and hated to see my record blemished. Therefore, I decided to do an online traffic school course so the point would not go on my record. While the ticket was very pricey – I live in California, need I say more – the course was only $20. When registering, I was offered the ability to be notified by email as soon as the state was told I had passed. This option was $8 extra and it seemed worth it (increasing the value to the seller by 40%). The embarrassing thing is that this proved to be only a slightly more complete notification than the one included with my original price and thus was entirely unnecessary! What this points out is that it pays for merchants to test various ways of upselling! People like me may be tempted to buy.

Commerce Models that Work

We believe people shop in three very distinct ways.

  • Some know exactly what they want to buy and are relatively price insensitive. Many of them prefer shopping online as they easily find what they want without traveling from physical store to store. Even a superstore like Walmart is limited to 140,000 SKUs (stock keeping units or choices where each size and color is one SKU). While this may seem like a lot, Walmart is missing over 99% of the 240,000,000 SKUs available from Amazon. As long as they believe the price is fair, these consumers are not tempted to keep searching etailers to find the lowest price. To such shoppers, it’s more important to buy from a trusted source rather than one that might offer a slightly lower price.

  • Other shoppers may also know what they want but feel compelled to continue searching through multiple sources until they discover one that appears to offer the best price. Amazon, like Walmart in the brick and mortar world, has attempted to position itself as both a trusted source and a low cost provider. In actuality, Amazon, like many brilliant retailers before them, charges more than other sources for many items, but has created the image of being lowest cost on ALL items. My wife and I most frequently shop on Amazon as we believe it:

  1. offers most items we shop for,

  2. has a price that is usually (but not always) within 5% of the lowest available, and

  3. offers free shipping to Prime members.

Since many others behave in a way that is similar to us, this means that it is difficult to compete with Amazon for customers that know what they want unless you are a more trusted source (in a category) or develop a reputation of being an even deeper discounter (like bargain shopping sites, eBay or closeout sites).

  • The third category of shoppers want an enjoyable shopping experience and want the store to offer items they might want to buy. Because of limited shelf space, physical stores must “curate” what they carry by selecting the items they believe typical shoppers at their store will prefer. Some do this by becoming a superstore in a vertical (like books, electronics, home improvement, etc.) which makes them a place where you can find most items you might want within that vertical. Others try to curate around style so their customers will enjoy the shopping experience. On the web, the concept of curation around being the trusted source for some type of goods, is a newer concept. Yet, the web is really quite ideal for this as there is much greater opportunity to match goods to customer needs and desires. Some examples of curated commerce include:

  • Trunk Club: Helps men to pick their wardrobes (even advising on what to wear on a particular evening out)

  • Curates workbooks and games for helping kids in school where they can match to both topic and grade level

  • Shoe Dazzle: Provides women with a personalized shopping experience around fashion shoes

  • Aha Life: Curated marketplace for creative, inspiring objects (higher end items)

  • Chairish: Offers a curated marketplace to buy quality used furniture

  • The Real Real: Offers a curated marketplace where women can buy used high end branded clothing

  • Open Sky: Offers products that celebrities endorse creating a “trust” if one believes in that celebrity

  • The Honest Company: Offers products for young children that are chemical free

None of these companies are trying to be the place you go when you know what you want to buy unless you’re a repeat customer. Rather, they are trying to create a fulfilling shopping experience where you can find things you might want to buy. Several of these companies have recently been acquired by larger ones. and Chairish are in the Azure portfolio so we believe enough in curated commerce to invest in it ourselves.

Any savvy retailer will tell you that increasing the lifetime value of a customer (LTV) is one of the key elements in creating a successful enterprise. For a brick and mortar retailer, this means getting customers to return to the store and getting them to spend an increasing amount when they do. Etailers actually have many ways to get customers to increase their LTV, including:

  • Have them become subscribers to a product set. This guarantees revenue each month.

  • Increase the invoice value of purchases by getting the customer to buy additional items related to the one they are already buying (This tie would look great with that suit; do you want shoe polish for those shoes, etc.)

  • Have them become members with small monthly fees that provide extra benefits like advance opportunity to buy new merchandise. (The revenue this guarantees is not the point of this, rather it is that paying money creates incentive to return to the site)

  • Have them become members with no fees but with an expectation that they will receive a monthly email. (Email, despite being viewed as old school, still generates more transactions than any other method)

  • Engage them through social media (Facebook, Twitter, Pinterest) and/or run campaigns with special privileges

  • Sign them up to a free shipping program (ala Amazon) where for an annual fee they get all purchases shipped for free

  • Feature users on the site (Betabrand has your picture show up when a Facebook friend comes to their store)

  • Provide personal customer service that makes them view you as a friend. Even something as simple as a personal thank you note in the shipping box creates a stronger bond. Extraordinary customer service enabled Zappos to become the leading online shoe retailer.

This list is not meant to be comprehensive but rather to point out things to think about. It also helps to A/B test everything so that you can determine the best path to drive site visitors towards to optimize monetization.

How Do You Measure Your Business Model?

As we said, we expect to do a series of blogs on the individual metrics we are about to discuss as there is some complexity in how to measure them. This post will focus on the various elements that go into calculating all variable costs associated with a sale. When we work with companies, we try to get them to correctly estimate two of the most important metrics in determining whether a company will be successful. These apply to B/B (business/business or enterprise) entities as well as B/C (business/consumer) companies. What is critical is that:

  • LTV (lifetime value of a customer) is greater than CAC (cost of customer acquisition) by “a lot”

I know “a lot” seems pretty vague. The way we think about it is that this excess must eventually exceed all the other costs (which we will refer to as Operating Cost, or simply OpEx) in order for it to be profitable. So, if OpEx is low, then “a lot” can be smaller than if OpEx is high (like a large R&D budget, sizeable executive salaries, etc). There is also a question of how long it takes to recover CAC. If the first year’s value more than covers CAC then the company will need less cash investment than if it takes much longer. We will spend time on cash flow issues in a subsequent post. For now, let’s try to give more color on the composition of LTV and CAC.

LTV starts with lifetime revenue (LTR). LTR is the expected total of all sales to that customer for every purchase they will ever make from that merchant. To get to LTV, LTR needs to be diminished by all variable costs associated with that revenue, or LTVC (lifetime variable costs). The most apparent of these costs is COGS (cost of goods sold). For a retailer selling physical goods, COGS should include all costs associated with purchasing or making those goods and the cost of storing them. Companies will usually include the shipping cost for receiving inventory but some companies don’t include the cost of shipping to their customer in COGS. Since these costs are variable (unless the customer pays them) they are part of LTVC. Other variable costs should include:

  1. Credit Card Charges

  2. Warranties (It’s very important to associate this with the item when sold rather than when the warranty takes place)

  3. Returns (These should both be deducted from LTR and any cost incurred added to LTVC)

  4. Credits given to a buyer like coop marketing (Often happens in B2B transactions)

  5. Customer support

The following case should help us better explain this analysis. It is based on an actual company but the name, product and numbers are made up.

ABC recently pitched Azure on an investment. Initially we were very excited by the company as it had tools that enabled a customer to personally design a high end dress specifically to her desire. The tool allowed you to pick from any of 30 base designs (for shape, style, etc). One could specify 6 personal measurements like waist size, height, etc that would allow the tool to understand how to create a custom pattern. Next, there were 10 material types, each in 3-10 colors for different parts of the dress. Finally, trims could be added to further design something completely unique. The price of the finished product ranged from $125 to $250 depending primarily on the materials selected. An average customer spent $175 on the first purchase. The company told us that gross margin (revenue minus COGS) was 50% (an average of $87.50), that the cost of acquiring a customer was $75 and that an average customer would buy 2-3 items over time. To us, this was the ultimate in curated commerce, helping women to completely design their own dresses. Revenue was also ramping nicely and would be about $4.5 million in 2013 up over 100% from the prior year. Since the initial purchase appeared to generate enough profit to more than cover the CAC and multiple purchases followed, we were excited to move this forward in our process.

We embarked on our detailed diligence process. First we used “cohort analysis” (we have used this method hundreds of times; it was first shown to us by the BillMeLater team in 2001 and will be fully explained in a subsequent post) to estimate the LTR of a customer.

The company had 4 years of data that indicated an average customer made 1.6 purchases (rather than 2-3) over a 4 year period. While it is possible that they’d buy a lot more over the next several years, this seemed unlikely since they had only made 0.6 additional purchases over the first 4 years after the initial purchase. So we estimated 1.6 x $175 or $280 as the average LTR of a customer. We next evaluated all variable costs associated with this revenue. We identified:

  1. COGS was 50% and included the material used in each dress, the manufacturing labor costs and manufacturing facility rental;

  2. Shipping and packing cost averaged $8.00 per dress;

  3. Customers paid via a credit card and the credit card fees to ABC were 3% of the charge;

  4. 10% of customers needed to have their dresses adjusted. The company paid for them to take it to a tailor at a cost of $25/dress;

  5. Customer service was outsourced and the company was charged $10 per service call and averaged one call per 5 dresses sold;

  6. 10% of customers returned the dress for a full refund and these dresses could not be resold since they had been custom made. The customer did pay the return shipping on these items.

While returns were subtracted from revenue, the cost of the returned goods was not added to COGS. All other costs in items 2-5 above were not reflected in COGS but rather in P&L items like customer service, shipping, etc. Our math indicated the following should be part of LTVC:

  1. COGS of 50% or $87.50

  2. Shipping and packing of $8.00

  3. Credit card fees of $5.25

  4. Dress adjustment cost of $25 x 10% or $2.50/dress sold

  5. Customer service occurred 20% of the time at a cost of $10/incident or $2.00/dress sold

  6. The subtotal of 1-5 is $105.25 variable cost per dress sold

  7. The cost of the returned dresses (10%) needed to be added to the 90% not returned. For every 1 dress returned there are 9 dresses not returned. Therefore the $105.25 in cost for the returned dress needs to be divided by 9. This adds $11.69 in cost per dress not returned, bringing the total cost to $116.94.

  8. So the variable profit on each sale would be $175.00 – $116.94 or $58.06

  9. LTV for a customer would then be 1.6 purchases x $58.06 or $92.90

Given the cost of acquiring a customer is $75, this leaves net variable profit vs. acquisition cost of $17.90 over the life of the customer. There was additional risk that some of the fabrics would never be used, reducing this number further. What we did know is that the second purchase typically occurred 12-18 months after the first. Since profit on the initial purchase did not cover the CAC, it meant that ABC would require substantial capital since it took over a year before CAC was recovered adding to the burn from OpEx. If creating more scale required lower pricing then the model could become much weaker, never yielding profits at scale. So while we really liked the concept we passed on the opportunity due to business model weakness.

This does not mean that the company could not be successful. If the model held up or improved (for example perhaps a 5 year period would show a higher level of repeat purchase) the company might eventually be profitable but only after consuming quite a bit of capital. On the other hand, the company could work on improving the model by lowering the CAC, increasing follow on purchases, increasing the initial invoice amount through offering add on purchases, etc. What the analysis indicated was that without action the company would most likely fail.


The ability to offer an almost unlimited number of products has made Amazon an even bigger gorilla in online retail than Walmart in brick & mortar. But having unlimited products also makes shopping very difficult when the potential customer doesn’t start out knowing what they want to buy. Curated commerce turns the model on its head. Like many successful physical stores, such etailers select a limited amount of inventory and provide an enjoyable shopping experience to help their customers find items they might want to buy. Importantly, etailers need to make sure that their business model works over time. The key to this is maximizing a customer’s LTR and keeping variable cost associated with that revenue to such an amount that the LTV of a customer is well in excess of the CAC.

On a final unrelated side note, a number of you (about 30-40) have sent their comments on prior posts to me via email. These comments were great. And please feel free to comment on the blog as well!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s