Are you building a brand or an ad arb business?

Are you building a brand or an ad arbitrage business?

If this were a poll, I suspect 100% would claim they are building a brand. I believe the real number is less than half. Let me explain.

From 2011 to 2020 I was deep in the facebook performance advertising space and part of my job was managing the global facebook relationship for nanigans. Our customer base were mobile gaming companies such as Zynga, EA, Supercell, and Machinezone, ecomm companies such as eBay, Zappos, and Wayfair, DTC 1.0 brands like Warby Parker, Glossier, Peloton, and Bonobos and lead gen and edu companies. These companies used our software to manage north of $750 million of facebook spend per year. Their objectives? Drive immediate sales. ROAS was the golden metric.

In my experience, much of DTC online chatter mirrors the conversations of 10 years ago. ROAS, Meta (then facebook) performance, outages, lack of attribution, should I try channel X, Y or Z…It’s the same topics. The reason it’s the same is the underlying strategy is the same - buy a first time customer at a low price and monetize the customer at a higher price.

What’s wrong with that? Two successful founders recently talked about this problem. The first post to catch my attention is a piece of content from Preston Rutherford, LinkedIn post: ROAS is like a drug. Preston is a co-founder of Chubbies which built a successful business in men’s apparel (hard) with an initial focus on shorts (even harder) and exited for $129 million (incredibly hard) to Solo Brands. Preston makes the point that ROAS was like a drug that he and the team chased at the expense of doing what they really wanted which was building a brand. Their real goal should have been building loyal customers. I encourage you to read the whole post.

The second piece of content was the The Operators podcast episode 45 where @seanecomm talks about a key pivot in his strategy for growing Ridge. What perked my ears was Sean’s assessment of his success to date, what has changed and what he believes he must do now. You should give the episode a listen to hear, but to summarize Sean’s point:

1. Ridge’s competitive advantage over the past 10 years has been their skill at paid advertising. Being so good at this enabled them to grow quickly and profitably.
2. The arbitrage that skill provided is “shrinking, shrinking, shrinking” as Meta and all the other platforms are getting faster at making it easier for everyone to be good at advertising and thus performance is rapidly moving to the average. Arbitrage opportunities are going away.
3. To take Ridge to the next level, he needs to focus on better telling the story of the brand and in his case, that meant doing a partnership with Marques Brownlee.

The commonality of these two insights is two founders realizing that the ad arb game was a dead end and ultimately distracting them from what they really needed to focus on which was building their brands.

Not either/or

I am not arguing that brand or ad arb is an either/or decision. In the early days of building a business, it’s critical to drive sales and get customers. Companies often look a lot like ad arb businesses as they focus on customer acquisition and immediate sales conversions. So ad arb and brand building is really a spectrum. Where you are on that spectrum has a lot to do with your stage.

Where you are on that spectrum also has a lot to do with what you sell. At one end of the spectrum are the Amazon Seller types that Thrasio tried to unsuccessfully use for multiple arbitrage under the greater sucker theory. Next would be undifferentiated resellers of Chinese goods with cool logos and websites. And next to them would be single purchase companies like the mattress in a box companies.

My goal with this post is not to disparage business models or companies but rather to shake some founders into honestly answering two questions:

1. Am I really a brand? Or am I more like a commoditized product or single purchase focused company?
2. If I really can be and want to be a brand, have I been pursuing an ad arb strategy for too long and am I at the stage where I need to shift focus to brand and away from ad arb?

I am doing this because I want founders to maximize their outcomes. Brand builders should focus on brand. And the ad arbers should throw away the pretense of brand and get on with ad arb. Not being true to yourself is what will lead to failure.

Specifically, I am going to cover:

1. My definitions of a brand and an ad arb business
2. Why ad arb works and when it fails
3. Why ecomm is a lousy monetization method for ad arb
4. The flawed thinking that led DTC into ad arb
5. Famous ecomm ad arb flameouts
6. Why the market doesn’t like ad arb models for ecomm
7. Why accounting treatment confuses the issue
8. How to make brand building actionable

My brand definition

A brand is the predisposition of a group of people to buy from a specific company.

I am talking about the sum of conscious and unconscious ideas an individual has about a company that pushes them towards or away from a purchase from that specific company. I am not talking about marks and logos. I am also not talking about product because it’s tautological that a great brand must have great products. Besides monopolies and low cost providers, who can deliver crappy product experiences but increase the predisposition to buy? This isn’t to diminish obsession on the product experience. This is exactly what you should do to build a great brand. But the product is simply the physical manifestation of the brand and what helps to enable that predisposition to buy.

Companies are machines that deliver returns based on capital inputs. As founders, our goal is to maximize the returns per capital input. We build brands because that is the single best lever for moving the return / capital equation.

Some measures of strong brands are:

• Pricing power. This means being able to price at a premium to competitors and well above utility value. You can raise prices without having a prayer session to use Warren Buffet’s words. You sell through at full price and don’t discount or discounting is minimal.
• High repeat purchase rates.
• Increasing contribution profit from repeat customers
• High organic traffic.
• Steady or declining CAC’s.
• A customer pyramid that is heightening and widening

My ad arb definition

An ad arb business is focused on the immediate monetization of ad clicks. The more immediate the focus, the more of an ad arb business you are. Signs of ad arb focus are:

• The CEO and leadership team obsess over the ad spend / first sale cycle as evidenced by spending most of their time reviewing ad accounts, ad analytics, ad campaigns and ad metrics like ROAS and MER.
• The main benefit of repeat customers is improving LTV’s to justify higher CAC’s.
• Borrowing lots of money to fund ad campaigns.
• The business has little understanding of repeat customer mechanics and monthly cohorts is as deep as it gets; there is little or no concept of contribution profits by repeat frequency or knowledge of products that lead to high repeat contribution profits; there is a lack of understanding of who, how and why customers become high repeat and high contribution profit customers let alone how to acquire more of them.

Why ad arb models work

Ad arb businesses can be fantastic companies. The model can be super powerful when applied in the right way. One of my favorite companies of all time is Red Ventures who started with the purest of the pure ad arb model, executed it superbly and used that foundation to build a multi-billion powerhouse which is also insanely profitable.

Ad arb models have two parts: (1) acquisition and (2) monetization. The simpler, the better and the best ad arb models are super efficient at converting acquisition into profits. This is what makes lead gen so profitable so quickly. Use online ad skills to acquire lead, sell lead for more to company that can’t do this for themselves, rinse and repeat. Red Ventures did this to perfection with DirectTV and others in their early days. There was virtually no friction between the paid ad and the profit.

Examples of ad arb models

Many new media companies are simply ad arb models. Create content; acquire cheap traffic; monetize traffic through higher priced ads. Part of the genius of facebook was in getting the audience to not only create all the content for free, but to also use them to bring in new customers and traffic. Where Meta went next level was in closing the system, eschewing third party ad monetization and building out perfect ad units and targeting. Travelocity started as (still is?) an ad arb business. Get really good at SEM and SEO, use UGC for free content, monetize the traffic through ads and lead gen.

Mobile gaming companies can also be great ad arb businesses, but they have more friction because you first have to build a good game and second have to be good at acquiring whales who monetize versus free or low value players who don’t. But once you have a good game, dialing in the monetization is straightforward and can lead to years if not decades of massive profits.

Why ad arb models fail

Acquisition and monetization. Ad arb models fail when one or both become too hard or inefficient.

On the acquisition side, ad arb models are often built on arbitraging a single channel. While ad arbers will try lots of channels, they typically follow a power law where the large majority of traffic comes from a single channel. 30 years ago it was search. Then for many it was facebook/Meta. When these platforms make changes or competition becomes too intense, the arbitrage opportunity disappears. For most lead gen, losing their edge in a channel means game over as they struggle to master another channel. Coincidentally this is another reason Red Ventures is so impressive. They mitigated their biggest weakness, reliance on someone else’s channel, by become content owners and producers.

Inefficient monetization can be overcome by massive arbitrage on the acquisition side. It doesn’t matter that much what your machine looks like when a dollar of ads results in ten dollars of sales. But as the arbitrage shrinks, the inefficiency of the machine becomes more perilous until it eventually becomes deadly.

Ecomm as ad arb

Ecomm, by the above standards, make for mediocre ad arb businesses because of the high friction between the acquired traffic and the monetization. Think of the complexity of the monetization mechanism. You have product development, international supply chains, logisitics, customer support, inventory planning and finance in addition to your customer acquisition. You have multiple channels such as DTC, Amazon, marketplaces, wholesale and own store, all with their unique cash cycles and nuance. Effective monetization means mastering multiple disciplines across multiple channels which means lots of expertise, people and platforms. It also requires a ton of working capital. And at the beginning of it all, you have to be great at customer acquisition. If other ad arb models are variations of riding a bike, then ecomm is like riding a unicycle across a high wire with no net while juggling a bowling ball, a tiki torch and a rattlesnake.

How did we get here?

We know the story of ZIRP, the rise of VC in DTC and funneling capital into facebook ads. But why? The underlying and flawed hypothesis behind it was that instead of building a brand over decades or even generations, with Shopify, some apps, design savvy, enough capital and facebook ads, pretty much anything could be turned into a ‘brand’ in just a few short years. This isn’t to say it could not be done. There are examples of companies that were able to start as ad arb and transition to brands and then ultimately exit at premium valuations. And every one of those inspired another ten thousand founders. But the darlings of that era like Allbirds, Outdoor Voices, and Away are today’s cautionary tales.

The truly great brands take time. Facebook and easy VC money were an accelerant for some. But for most, the wrong lessons were learned and the longer they chase ad arb as opposed to brand building, the longer and harder it will be to get premium valuations.

Cautionary tales from famous ecomm ad arb flameouts

Remember Fab.com? How about two others recently in the news, Zulily and Wish.com?

I imagine that when Zuck feels blue, he dons his VR goggles and logs into a secret room where he gets to relive watching the dashboards showing the billions of ad dollars from past ecomm ad arb businesses flow in.

Fab.com was the first big ecomm spender. Their $25 million annual facebook budgets wouldn’t even get them a real rep now. But back in the day, they were feted by Sheryl and the most senior facebook execs. They raised $336 million at valuations cresting a billion and ended up selling to Publishers Clearing House for around $15 million.

Wish.com took it to another level. They spent over a billion on ads in 2021. They raised a billion from investors but recently were acquired for $173 million by a Singapore company Qoo10 which is a 99% decline from their 2021 market cap.

Zulily spent hundreds of millions on facebook and parlayed that into a $2.4 billion acquisition by Qurate. After blowing a hole in Qurate’s balance sheet, they were passed to a PE fund and then went bankrupt. Their IP (i.e. their ‘brand’) was just purchased by Beyond for about $4 million. Literally hundreds of millions in ads converted into a brand worth $4 million.

This isn’t to bash the founders and teams at these companies. The point is that even billions of dollars of ads on channels like Meta won’t make you a brand. Fab, Wish and Zulily sold third party products. Outdoor Voices, Allbirds and Away sell their own. And all these examples have unique contributing factors to their falls. My argument is that their single biggest factor was chasing an ad arb model.

How the market views ecomm ad arb

As founders are painfully aware, investors and acquirers have awakened to the danger of the ad arb model applied to ecomm. The reliance on channels like Meta means the acquisition machine can stop at a moments notice (as the events of the past couple weeks have shown). The need to continually pour capital into the acquisition machine to keep the company alive combined with the complexity of the monetization engine equates to low Return on Capital and high execution risk. In other words, little interest in investing and acquiring. The answer and the only solve, is building a real brand.

Accounting practices make brand building harder

Imagine you had the choice to buy one of two companies. Company A has a ton of inventory, but no repeat customers whereas Company B has a ton of loyal customers, but no inventory. Which problem is easier to solve? Which company leads to more certain profits, faster? I would buy Company B because it’s far easier and more profitable to buy the inventory you know the loyal customers already want as opposed to acquiring the customers who may want the inventory you already have. Company B is the more valuable company.

But how does accounting view this? Company A has inventory which is an asset. Company B has a customer file. Accounting views Company A as more valuable because Accounting does not have a way to understand the value of a customer base until a company is sold. When a company is purchased for an amount above the book value of its assets (e.g. cash, AR and inventory) the remainder is called goodwill (technically goodwill and intangibles).

I believe this practice of thinking of marketing as an expense and having no way to show the growing customer base asset in the financials exacerbates ad arb thinking. If we thought of marketing as investments into building an asset, we would naturally use better metrics to evaluate the effectiveness of our customer acquisition and move away from ROAS and MER which are just P&L focused metrics. And coincidentally, that focus on the balance sheet would also improve our focus on Inventory Turnover, Free Cash Flow and Return on Capital which would lead to better management practices than a simple focus on profit.

How should we think about customers?

Your customers represent future cash flow streams. Each individual will generate a different cash flow stream depending on their purchase frequency, contribution profit of what they purchase and the cost to maintain them. You know this intuitively. Some customers being Loyal Fans buy repeatedly and at full price and other customers being Discount Hounds buy only when incentivized with deals. Some come back infrequently, the One and Done’s you never see again and some even cost you money through fraud.

Cohort analysis is not enough. Time-based cohorts are simply an average of all the types of customers grouped together by the date of their first purchase. Each cohort is a mix of Loyal Fans, Discount Hounds, One and Done’s and Fraudsters. This is useful for understanding average trends over time, but it’s nowhere near enough to understand the value of your customer base.

The customer pyramid is the most useful heuristic for thinking about your customer base as an asset and how well you are investing to grow that asset. At the top of your customer pyramid are your most loyal customers who buy repeatedly and at full price. Next layer down might be repeat customers who buy a little less frequently and maybe sometimes on sale and then your bottom layer is customers who have bought from you once. Even below that are customers who are aware of you, but have not yet converted. This is your brand. It is a group of people with a predisposition to buy from you.

Over time and as you invest in more marketing, does your customer pyramid grow? How much? And are all your layers growing equally? It would be a troubling sign if your pyramid’s base were expanding, but your higher layers were not keeping up (i.e. your pyramid was getting wider, but shorter). The ideal shape of your pyramid will change over time depending on your segment, business model and stage.

There’s a terrific description of Ferrari’s customer pyramid in this Business Breakdowns podcast.

Brian Lum - Ferrari: Magic from Maranello

The pyramid I describe above is layers of groups of customers with various commonalities around contribution profit. Getting clean enough data to do this analysis on an individual customer basis can be a challenge. So while the ideal analysis is individual, simplifying the analysis into large groups works if that’s all you can do. And regardless of underlying method, presenting the analysis in groups helps people understand and act.

Whether you model it individually or by layers, the concept for valuation is the same. The customers represent future cash flow streams. Each forward stream will be different based on purchase frequency, contribution profit of the products purchased and cost to maintain. Sum all those predictions and you get the forward value by time of your customer base. Discount those back at your cost of capital (or other rate - this is an entire post in itself) and you get the present value of your customer base.

How does this compare to cohort analysis? The cohort analysis predicts the purchase frequency of the cohort into the future. This is an average of that particular group. It then applies an average AOV or if more sophisticated, an average contribution profit for the entire business to that purchase frequency. This is certainly simpler to do. And greater precision may not yield more accurate predictions. But what the cohort analysis won’t do is help you understand who, why and how. Who is becoming more valuable? Why do they become more valuable? How do we effect that? And how do we acquire more customers like them?

Why is brand important to investors and acquirers?

There are four main reasons brand has value for companies.

1. Brand gives pricing power. It is the only defensible moat in commerce. What is the difference between a Birkin bag and another high quality, handmade leather bag? Could the vast majority of Birkin buyers point out the differences in two unbranded bags? Do they actually care? The brand enables Hermes to charge $50,000 more for that bag than the unknown artisan can. That pricing power helps Hermes knock out an astounding 48% EBITDA margin and 26% FCF/Total Revenue which investors have loved to the tune of 33% CAGR of share price over the past 5 years easily doubling the S&P 500 (14%).
2. Brand means repeat purchases. As Jeff Bezos told Anne-Marie Peterson from Capital Group back in 2010, habits are incredibly hard to change. But once the consumer has changed their habit, you can get the benefit of that change for a long time (Invest Like the Best podcast episode March 12, 2024). Costco has done such an incredible job at creating this habit that it can even charge for it directly in the form of a membership fee. This membership fee alone was $4.6 billion in revenue in 2023.
3. Brand means more efficient marketing because customers tell others about the brand which has a far more powerful impact on their predisposition to buy than any ad can.
4. And all the factors above plus the ability to control or reduce ad spend as a percent of sales means brand enables margin expansion.

How to move from ad arb to brand building

If you have come this far, you are likely interested in brand building. If you suspect you are more ad arb than brand, how do you make the switch? This transition will require new metrics, new skills and new ways of running the business. It will also require relentless messaging to stakeholders. During this transition period, I would focus on four things:

1. Switching the mindset from sales growth to asset building. Your greatest asset is your customer base.
1. Use the customer pyramid heuristic and brand metrics above to track the value of your customer base and improve it.
2. Generating Free Cash Flow is critical. You can’t make the transition without the ability to generate your own cash and capital.
3. Get inventory under control and understand how much of your future inventory buying is expected to be bought by newly acquired customers as you make the switch from a focus on inventory for newly acquired customers versus a focus on repeat customers.
4. Return on Capital becomes your guiding principal. Product, Customer Acquisition and Distribution should be measured and judged by metrics that drive ROC. You can read about these metrics in detail at The Payoff Playbook.

Successfully making the move to brand building will give you three huge benefits.

1. One is the mental and emotional benefit of building what you actually want. Did you start your company because you had a vision for a brand or because you wanted to master Meta?
2. The second is increased speed and quality of decision making as you and your teams use the easy to understand brand filter.
3. The third benefit is financial as you will build a more valuable company.

Let me know your questions and feedback. And I wish you success!