Jim Barksdale once said: “Nothing happens until somebody sells something.” This challenge is made harder by another business truism: if the cost of acquiring customers exceeds the ability of the business to monetize those customers, the business will inevitably fail since it has not created product market fit. Not only does someone need to “sell something” but is must be done in a way that is cost-effective.
Some people forget that a functioning business model is an essential part of product market fit. Andy Rachleff has identified the elements a business must create or discover to have product market fit: “the features you need to build, the audience that’s likely to care, and the business model required to entice a customer to buy your product.” If you do not have a scalable and repeatable business model, you do not have product market fit.
Paying too much to acquire customers is not a solvable problem. But not paying anything to acquire customers is not an option for a real-world business. Sometimes you will hear a person claim that they pay nothing to acquire customers. They may be telling a very nice story, but it does not reflect of reality. These storytellers have often simply shifted to a “proprietary product distribution” model, which inevitably has costs that appear somewhere other than in a line in the income statement called “sales and marketing” spending. I should emphasize that in this blog post I am not writing about what some people call “organic” distribution since the term has been made less meaningful than it once was due to the marketing approaches of some companies that sell “paid distribution.”
Many businesses today are so reliant on paid product distribution that they retain little or no value from their value chain. Paid product distribution happens when the business pays another business a fee for acquiring the customer and may not even own the relationship with end customer or the data generated by that customer. The problems caused by relying on paid distribution often create a version of a “wholesale transfer pricing problem” that I have discussed many times on this blog in the past. The supplier of distribution in a paid arrangement has this power and shifts the profits to itself.
“Proprietary product distribution” is a customer acquisition system that is within the control of the business itself and which generates a customer relationship that the business owns. Owning the data generated by this direct customer relationship is increasingly valuable since it can be used to fuel machine learning and other modern business data science processes. To say that so-called “co-ownership” of the customer relationship is problematic is a significant under-statement. Do you really want another business owning the data that your customer generates? The data “exhaust” produced by direct customer relationships is increasingly what makes products more valuable since it is the fuel for machine learning and modern data science. If Business A has data from a direct relationship with a customer since it was acquired with proprietary distribution and Business B has no relationship with the customer since there is an intermediary between it and the customer, Business B is very probably walking dead.
A test for whether distribution avoids being paid is simple: Would the customer have arrived arrived regardless of marketing spending? Is your distribution paid or proprietary? Some businesses have a mix of the two approaches. Disney streaming its content direct to customers is one type of proprietary product distribution. A Disney advertisement on a third party web site is paid distribution. Additional examples of proprietary product distribution will be discussed later in this post. Businesses that can create proprietary product distribution are modern day alchemists. The best business alchemists can turn modest spending that is usually considered part of “cost of goods sold” or sweat equity, into a substitute for marketing spending.
Proprietary distribution channels always have some cost, even if it is opportunity cost. If that cost is reasonable relative to the revenue generated, it can create a very attractive business. In short, even with proprietary distribution the cost of acquiring the customer must make sense from a unit economics perspective. When a business files documents when it becomes a public company the public can see for the first time how much money is being spend on sales and marketing. The founders and investors have known this for some time. Very high sales and marketing spending is part of what Bill Gurley calls the “game on the field” right now in many businesses. The markets love revenue growth especially if it generates an annuity-like flow of cash. If your competitor is growing based on “pedal to the metal” sales and marketing spending, what can you do other than matching them by “turning it up to eleven” as the guitarist did in the movie Spinal Tap?
The entire business model of venture capital depends on grand slam home runs. A business does not produce a grand slam home run for founders and investor by only acquiring customers with paid product distribution. That does not scale. For example, Dropbox, Atlassian and Docker do not acquire their customers by purchasing advertising in newspapers and magazines or television advertising. Proprietary product distribution is the only approach to acquiring customers that can generate the necessary scale to create businesses with multi-billion dollar valuations. While it is hard to overemphasize how important growth is to a venture capital backed business, growth is just as important for an existing company trying to grow an existing or new line of business.
The return on investment on paid product distribution has dropped significantly in recent years. In a recent post I quoted Andrew Chen as saying: “Addiction to paid ads is bad.” Paid advertising on additional problems in addition to what I discussed in the previous paragraphs. The internet has created enough transparency that people aren’t as influenced by traditional advertising. It is harder to fool some of the people as many times as it was in the past. Consumers have access to too much data for paid advertising to be as effective as it once was. Jeff Bezos nails key aspects of the nature of the change here:
“The balance of power is shifting toward consumers and away from companies…the individual is empowered… The right way to respond to this if you are a company is to put the vast majority of your energy, attention and dollars into building a great product or service and put a smaller amount into shouting about it, marketing it. If I build a great product or service, my customers will tell each other…. In the old world, you devoted 30% of your time to building a great service and 70% of your time to shouting about it. In the new world, that inverts.” “Your brand is formed primarily, not by what your company says about itself, but what the company does.”
The shift away from paid product distribution is not binary but it is significant, especially when the product is software as a service (SaaS). Advertising is unlikely to disappear as a means of acquiring customers. Zenith expects global advertising expenditures to grow 4.1 percent in 2018 to reach $578 billion by the end of the year. The United States (perhaps due in part to its scale combined with diversity, which places a very high value on targeting) also remains the largest market in the world, accounting for ~ 38 of global advertising spend in 2017. Dentsu issued a forecast for advertising in June of 2017 as follows:
“In 2018 digital will be the top media in terms of global share of spend, taking over television for the first time. Digital’s share of total media spend is predicted to reach a 37.6% share in 2018 (up from 34.8% in 2017), versus 35.9% for television (down from 37.1% in 2017), amounting to a total value of US$215.8 billion. Reflecting the continued disruption by digital technology of the print media industry, Paid Search (advertising within the sponsored listings of a search engine) is forecast to overtake traditional print media (newspapers and magazines) in 2018.”
Business are obviously spending a lot on advertising. Why is proprietary product distribution so important? The key words in this tweet below are “most efficient traffic acquisition possible.”
If a competitor has a more efficient traffic acquisition engine than your business, you have a significant problem. This greater customer acquisition will enable greater growth, which can create network effects. Ideally the growth feeds back on itself, creating more growth n ways that are nonlinear. The “game on the field” has forced many businesses into a battle for growth that is essentially “go big or go home.” Gurley points out:
“You could sit around and say, ‘We’re going to get to profitability,’ but you’re not going to matter. You might as well lock the door and leave the building. So you’re forced into a game of capital warfare that you may have not been ready to play. And so I don’t know that any one person is responsible. Silicon Valley and venture capital have always been cyclical. And so there’s something about human nature that causes us to be increasingly risk-seeking until someone comes along and really punishes everybody.” “…there’s this old saying about selling dollars for 85 cents. But there’s a truism to it. You can create infinite revenue if you sell dollars for 85 cents. And if you give consumers more value than you charge them for, they will love you. And I remind entrepreneurs all the time that Webvan had the highest NPS scores of any company I’ve ever known. It wasn’t that the consumer proposition didn’t work, it was that the economics didn’t work. They weren’t charging enough for the service level.”
In some markets the competition to acquire customers can ruin a business. Charlie Munger talks about that set of issues below and has said that he has not yet been able to figure of why this happens in some businesses but not in others:
“Over the years, we’ve tried to figure out why the competition in some markets gets sort of rational from the investor’s point of view so that the shareholders do well, and in other markets, there’s destructive competition that destroys shareholder wealth. If it’s a pure commodity like airline seats, you can understand why no one makes any money. As we sit here, just think of what airlines have given to the world—safe travel, greater experience, time with your loved ones, you name it. Yet, the net amount of money that’s been made by the shareholders of airlines since Kitty Hawk, is now a negative figure—a substantial negative figure. Competition was so intense that, once it was unleashed by deregulation, it ravaged shareholder wealth in the airline business. Yet, in other fields—like cereals, for example—almost all the big boys make out. If you’re some kind of a medium grade cereal maker, you might make 15% on your capital. And if you’re really good, you might make 40%. But why are cereals so profitable—despite the fact that it looks to me like they’re competing like crazy with promotions, coupons and everything else? I don’t fully understand it. Obviously, there’s a brand identity factor in cereals that doesn’t exist in airlines. That must be the main factor that accounts for it. And maybe the cereal makers by and large have learned to be less crazy about fighting for market share—because if you get even one person who’s hell-bent on gaining market share…. For example, if I were Kellogg and I decided that I had to have 60% of the market, I think I could take most of the profit out of cereals. I’d ruin Kellogg in the process. But I think I could do it.”
To say right now the market loves growth is a huge understatement. But the standards for that growth being sufficient are high. For example, in SaaS markets Rory O’Driscoll writes:
“most venture investors prefer to invest in companies that have at least the chance to become standalone public companies (which is not to say most achieve this objective). Looking at the realistic low bar of what it takes to be a public company, this implies being at run rate revenue (ARR) of $100 million at the time of IPO, while still growing at 25 percent or greater in the following year.”
In consumer markets, Ryan Metzger has created a chart which shows growth rates that look like this:
An investor or founder of a business which desires to grow quickly will want to see cost per customer acquired acquired dropping over time. If customer acquisition cost (CAC) is increasing over time it may cause a yellow or red light to start flashing for an investor or founder depending on the rate of increase and whether customer lifetime value is staying the same or growing. Customers are getting more expensive to acquire. Data supporting this view is in this chart:
Tom Tunguz explains why this increase in CAC is happening:
“Patrick at ProfitWell sent me his survey results across about 800 companies. The chart above shows the increasing cost of customer acquisition on a per company basis. Those surveyed have observed a ~65% increase in cost of customer acquisition over the last five years. This increasing customer acquisition cost likely has two root causes. The first is competition. The second is scale. As a company grows, the initial customer acquisition channels become less efficient with saturation. Consequently, the business must develop a portfolio of customer acquisition channels. Typically, each marginal channel has a higher cost of customer acquisition.”
Mary Meeker’s data pointed out this past week that customer lifetime value in increasingly being used as a decision making tool:
I have written a few blog posts recently about the problems with paid advertising generally and I am not going to raise and discuss all the points again. In this blog post I am going to focus more on the ways in which a business can create proprietary distribution, most notably in a software as a service (SaaS) business. Before moving on I should make two key points. The first is described by Gurley: “Users typically have a higher NPV, a higher conversion rate, a lower churn, and more satisfied [if they are not] acquired through marketing spend.” Gurley is saying that customers you acquire through paid distribution channel are less valuable when you look at the unit economics. David Skok argues: “Owning the customer base is an important way to control your own destiny. It will also earn your company a higher valuation.” We are seeing more and more businesses like Disney going direct to consumer since they can’t take the risk of having a distributor between them and their customers. As I have also written before, the telemetry that can be generated via direct customer interaction is far too valuable to be given to an intermediary in many businesses. That telemetry can be used as a basis for experiments powered by machine learning that can make the product better. Of course, there are some markets in which advertising has never been effective. As an example, Peter Thiel has pointed out that “for most small business [customers] you can’t really advertise. If you can’t solve the distribution problem, your product doesn’t get sold—even if it’s a really great product.”
A tweet from Rick Zullo describes what will be the topic of the rest of this post simply: “Knowing how to create/cultivate proprietary distribution can be huge lever.” Achieving proprietary distribution can take many forms:
- Freemium:The essence of the freemium business model is to reduce the firstion associated with initial usage and to as a result reduce customer acquisition cost (CAC) when compared to the paid alternative (e.g., based on paid advertising). Steven Sinofsky points out that “The fundamental tenant of freemium is experiencing the product is the best advertisement.” For this reason, freemium can lower CAC, sometimes at the cost of higher cost of goods sold (COGS), in order to produce a higher LTV and increased growth. In a freemium model there two very different types of services: (a) Free services from which that are upsold or cross sold to an existing revenue generating customer; and (b) free services offered when the business is starting from zero in an attempt kickstart the same motion as in (a). Starting from zero will result in a higher cost of customer acquisition (CAC) since there is an additional step. A classic example of freemium strategy is Dropbox which was described in a Harvard Business School case as follows: “Dropbox provided remote-storage over the internet of any type of computer file, along with file sharing, synchronization and backup. Using a freemium pricing strategy whereby a basic service was free-of-charge and a premium service was paid.” At the end of last quarter Dropbox has 11.5 million paying customers, but supports more than 500 million users in total.
- Community: Nextdoor is also a classic example of community-based proprietary distribution. Here is a slide from Mary Meeker’s slide deck released this week that shows how the Nextdoor community has grown :
- Referral: Business like Dropbox have offer referral credits to create proprietary distribution. For example, “Dropbox will automatically reward you with 500 MB per referral, resulting in a total of 32 x 500 MB = 16 GB of extra online storage.”
- Content Marketing: Much of the content on Twitter and Medium is created by people who are participating in the platform’s service either partially or completely in order to directly indirectly market their product. For example, wealth managers and venture capitalists that write blog posts are engaged in content marketing. There are many providers of SaaS that work hard to have a strong presence on social media and platforms like Medium. Elon Musk is an example of an entrepreneur who uses social media to generate passionate customers and shareholders. He also attracts his share of critics. The resulting controversy creates free brand impressions.
A CEO or founder who is able to make the right points in the media is able to create free brand impressions. Another form of content marketing/referrals is harnessing “influencers.” As just one example, platforms like Instagram can enable a brand to turn fans into referral engines.
“it would be easy to describe Fashion Nova as ‘an Instagram brand.’ Saghian likes to call it a viral brand, but something about those descriptions make the company feel ephemeral and illegitimate. Fashion Nova is a bona fide apparel, accessories and beauty business that’s attracted women with body measurements that don’t always fit within standard sizing — although Saghian says the clothes are for everyone….. By working with a network of more than 3,000 influencers, he’s made them pervasive. It’s easy to assume that Fashion Nova only exists online, but it has five stores within malls across Southern California.
Influencers operate outside of consumer markets since they can promote and provide referrals to consumer, small and medium business and even enterprise markets.
The proprietary product distribution strategy will be assembled from these components. In doing so they must decide how to account for this type of marketing and the decision over whether you include it as part of CAC or generic brand marketing. Here is how a Ilya Fushman describes the goal in an enterprise software as a service (SaaS) setting:
Investors carefully look at spending and want to know customers acquired by source and cost for each over time. Tom Tunguz writes: “In each of these cases [of proprietary distribution] the startup owns the distribution channel. Sometimes the channel is integrated in the product, sometimes the channel is adjacent. Challengers may try to replicate the success of the distribution channel by copying it, but cannot compete within the channel, which means it’s an incredibly cost-effective channel for its owner.” This is a set of sentences that is worth taking the time to unpack. If a business has developed a community that generates customer growth that community can be recreated but that community itself belongs to the business that created it. This fact gives the company that created the proprietary distribution an important competitive advantage.
Ben Horowitz has a great post on which type of distribution is right for a given product. He sorts businesses in terms of their distribution characteristics in this way:
“can be delivered online, no assistance required: GitHub; Dropbox, Slack (for small companies)
can be delivered online, minor assistance required: Okta, Salesforce (for small companies)
can be delivered online, major assistance required: Oracle Financials, Palantir; Dropbox, Salesforce, Slack (for large companies)
not delivered directly, no assistance required: Anki Overdrive, Apple Watch
not delivered directly, some assistance required: Nest Thermostat
not delivered directly, major assistance required: EMC Symmetrix”
Of course, the nature of the assistance that must be provided will impact the business model. Is the assistance required to get the product into use, to enhance value, to manage, to get the most out of, or to sell them more products? Different answers to these questions will impact the unit economics of the business model in different ways. Horowitz then adds:
“…you need to design your sales channel to meet their needs not yours. In this context, we can think about some example targets:
I. Individual — a direct consumer or a single decision maker in a company
II.Small group — a small engineering team deciding, such as with Trello
III. Entire small company (<1000 employees) — for example, deciding on HR software like Zenefits
IV. Large group — multiple decision makers, with different economic and technical motivations for deciding on the product
V. Multiple groups simultaneously — for example, sales and marketing both needing to agree on the right marketing automation solution
VI. An entire large company — for example, deciding on an HR system like Workday
…Targets I and II involve the same simple decision-making process: The customer asks herself, is this something that can help me and is worth the money and effort? These targets therefore can often be sold entirely via marketing; viral distribution (if you have a product that inherently travels by word of mouth or other organic spread); and optionally telesales …Target III decisions generally involve multiple people so are more complex. … Targets IV-VI [also] represent complex decision-making processes…”
This blog post is about to hit 4,000 words, so I need to stop writing now or I will turn into a pumpkin. More later. The linked material is excellent. I put a lot of effort into the End Notes and I hope people benefit from this work.
Andy Rachleff: https://www.youtube.com/watch?v=7G9Cb6sCjL8
David Skok: https://www.forentrepreneurs.com/startup-killer/
Tom Tunguz: http://tomtunguz.com/controlling-your-destiny/ http://tomtunguz.com/cac-increase/
My Previous Posts on Growth Topics.