I have covered the key elements of Netflix’s business model in great depth with a focus on key metrics, such as (free cash flows, customer acquisition cost, lifetime value, contribution margin) and business model / micro economic principles (economies of scale and diseconomies of scale).
Today, we are going to link these elements together using the concept of the flywheel:
- Customer unit economics (customer acquisition costs, CAC) and (customer lifetime value, LTV)
- The investment cycle as the source of value creation:
- content
- technology & development and
- marketing
- The competition
- The outlook for Netflix
- and finish with The Flywheel
The Netflix Flywheel
The Flywheel consists at the macro level of three processes:
- Value creation through decisions made in the investment cycle
- Creation of assets and a user base and the monetisation thereof (the latter part also being called value capture)
- External funding (liabilities) and competitor moves (external risks)
You can feel free to divide it further or even differently to how I am doing it so long it captures the underlying elements that we will discuss in the following.
I will be capturing important elements of this in a somewhat arbitrary order and conclude at the end with a few more words on the Flywheel as a whole.
About us: We share MBA-level innovation and business management knowledge with Entrepreneurs, Founder-CEOs, App Creators, Startups, Professionals, Executives, MBA Students/Graduates, Business School Professors and other smart people.
Customer Lifetime Value (LTV) and Customer Acquisition Cost (CAC)
Customer Lifetime Value (LTV) and Customer Acquisition Cost (CAC) are a different way of looking into the cost and revenue side. They are what we would call customer unit economics. I.e. the projection of key costs and revenues on a single (marginal) customer. (It does of course not reflect any individual as such.)
LTV, CAC formulas
How do you calculate the lifetime value of a customer (or LTV)? I have explained the LTV calculation in-depth here using Netflix as an example.
How do you calculate customer acquisition costs (or CAC)? I have explained the CAC calculation in-depth here using Netflix as an example.
You can look up the formulas there (they are not repeated here).
LTV, CAC overview and methods
As you know from your accounting lessons, there are often many different ways to calculate the same thing (leading to different results). Same holds true for LTV and CAC. Here is a succinct overview of two methods. One is more lenient the other more stringent. Both have their merits depending on where your company is in their lifecycle.
LTV, CAC method #1
Method #1 excludes certain cost items that are considered one-offs (technology & development, T&D) or not part of the core value proposition (such as general & admin, G&A) or in accounting speak fixed / overhead costs. The logic behind this is the assumption that these costs stay largely put as the company scales and more customers join (we speak of zero marginal cost of provision). This method includes only the direct costs of revenues. The biggest cost of revenue is content amortisation (I have covered the details in a previous article).
But this is only correct in the short-term. It really does not hold true in the long term. T&D, as well as G&A, too scale as the company scales up. I have shown this in our article on diseconomies of scale using again Netflix as an example.
On the content front, things are also not easy. For licensed content, the judgment call is clear: it should be captured in the costs of revenue. But what about self-created and acquired content? This content is likely to have additional future value for licensing out or consumption by future users. This is a tough call and who knows what happens to user preferences in a few years time given the massive acceleration of content creation across many platforms. Anyone can have their own judgment call on this – the conservative way to go about this is to fully take it into account which is what is done here and be happy about the upside should it materialise in the future.
Overall, this approach still has its merits. I would recommend this approach for early start-ups. If you check out my article on economies of scale (again on Netflix) you will see that G&A made 40% of the costs of revenue in the first year of being public. T&D ate another 25%. In this situation, we can be certain that economies of scale will settle in as the user base grows. Therefore, method 1 is good to use and won’t distort your results too negatively.
LTV, CAC method #2
Method 2 allocates all costs that are included in the operating results. It means though that it still excludes come costs, such as debt servicing costs or taxes. Given most of the debt is being taken to fund content creation you could be even more strict and allocate the debt service costs to costs of revenue (but I am not doing this here).
In the example here I have arbitrarily allocated the T&D costs 25%/25%/25%/25% to: (1) cost of acquiring new customers (domestic); (2) cost of customer retention (domestic); (1) cost of acquiring new customers (international); (2) cost of customer retention (international). I have done the same for G&A. The exact ratios are not important. For your own business, you would have a good feel how to do this. You could also allocate some of the content costs to customer acquisition, esp the marquee shows and movies (I have not done this and it would not change the top-line numbers much).
There is definitely a case to include costs that do not follow the classic sloping-down curve of economies of scale. I have shown that G&A as well as T&D have grown as a percentage of revenues (at least until about 2 years ago) even as revenues themselves have grown nicely. Thus, some if not all these costs should find their way into the calculation.
I am also using the net customer addition for the calculation of the CAC. Using the gross new customer addition is useful to assess the marketing efficiency. Using net customer addition assesses the overall effectiveness.
We need to be aware that we are being very hard when we do this as we now have the churn rate penalising both sides the LTV as well as the CAC. But that might not be wrong after all: it is possible that beyond a certain, let’s call it “critical” churn rate, the virtuous cycle may turn into a vicious one.
Using gross customer addition here would not change the numbers dramatically. CAC = $230 and LTV/CAC = 2.5, CAC/ARPU = 20 months. Feel free to use gross customer addition within method #2 if you feel it reflects better what you are after.
For a company of Netflix size, I would use method 2, especially for the mature US segment.
I hope you have found this discussion valuable. It should have helped you to understand the idea behind it. Decide for each cost how much of it should be taken into account for these calculations. There are other ways of calculating these metrics here.
What is a good LTV to CAC ratio?
One recommendation for SaaS type subscription models is LTV/CAC >3:1 and CAC/ARPU <12months.
I would say a ratio of LTV/CAC >3:1 caters for the fact that we do not include all costs into the equation. You could argue that if we apply method #2 we could reduce this ratio to say 2:1. My clear answer is: possibly. No metric should be used just by itself. If other metrics such as free cash flow or contribution margins and others as relevant to your business look good, then it may be acceptable.
In Netflix’s case, free cash flow is not good. And it indicates that there are other costs, esp debt servicing which are not included even in method 2. That is a point to watch over the coming years.
And, yes, as you incorporate most costs into these calculations you can adjust the ratio.
But our aim is not to assess how good things look for Netflix – we want to learn this stuff so we can apply it to our own ideas. So, moving on, we will be looking at what we can do with these metrics.
Essential financial education
I have met so many people who desperately want to understand corporate finance better. Inability to do so holds people back in progressing their career or embarking on their own business (sometimes even both).
I have learned a lot of finance in the last 10 years of my corporate career working with many Heads of Finance. But I also realised the importance of unit economics and spent thousands of hours in my free time over a period of five years studying this. It has paid off: unit economics is what you need to know to profoundly understand the trajectories of innovative companies.
One thing that really annoyed me and vastly delayed my learning curve was the quality of examples. Each concept was explained using a different example and without tying it to the big picture. This was very frustrating (I’m sure you know what I’m talking about). Each concept was explained using a different (teeny-weeny) example.
I decided to do it differently by using a single longitudinal example to explain these important concepts. Netflix has >20-years worth of audited, published financial data. With that, we now have articles on crucial financial metrics (Contribution Margin, Free Cash Flows, the investment cycle); unit economic metrics: Customer Lifetime Value/Customer Acquisition Costs, Economies of Scale, Diseconomies of Scale). All on Netflix!
These articles are very valuable. But the Netflix Premium Resource adds to it in many ways: it fills the gaps in between the articles, adds more details & knowledge, it’s presented better, is updated and gives you 11 resources, templates and downloads with lifetime access.
You will not only be learning how to calculate these metrics, more importantly, you will also be learning what they mean in the context of managing a business.
I have had feedback from many, many finance students, lecturers, professors and entrepreneurs alike on my approach: “Wow - that’s a thrill! Thank you so much for your immense and so valuable work” Financial Lecturer at IESEG School of Management.
It will be deeply empowering for your career and business. Learn more here ...
Investment cycle
(1) Content investment
There is some interesting stuff we can do with the customer unit measures in our investment cycle that is more useful than the aggregate (accounting) numbers.
We are going to look at how we can use these methods to make concrete content investment decisions. I mean as in: “show we invest in this show?” not as in “how much should we invest this year on content?” Let’s say, we can make more granular investment decisions using the customer unit measures.
On the latter question, Netflix made a clear statement that the overall content spending envelop is constrained by contribution margin targets. So, let’s look at the granular decisions.
The first step in this decision making is customer segmentation. But it is not the traditional geo-demographical segmentation that you are being told about in business school.
Netflix says “We have seen that where you live, gender, age and other demographics are not significantly indicative of the content you will enjoy, Time after time, we see that what members actually watch and do on the service transcends the predictions of stereotypical demographics.”
Instead, they divide their audience into 2,000 taste clusters (this number has gone up since it was first reported on so don’t be surprised if it still changes up or down).
Here are some interesting numbers in the context of content:
- 2,000 taste clusters is what Netflix divides their audience into
- 80 Netflix original films in 2018
- 700 (!) original TV shows in 2018
- 27,002 (micro) genres
These are pretty staggering numbers. Armed with the knowledge of taste clusters and the size of each one of them Netflix can make decisions to invest (or not) into content.
Take as an example “Friends”. Netflix tested the waters in not extending their licensing of the show but ran into a full-blown user revolt. In the end, they paid $100m to license the series for another year. This is the annual revenue they receive from 731,000 users. Equal calculations can be made for other, smaller shows.
Recommendation system
Closely linked to the taste clusters (thus a more sophisticated segmentation) is the recommendation system. At its core, it determines to which degree a viewer of one show will like another show.
“More than 80 per cent of the TV shows and movies people watch on Netflix are discovered through the platform’s recommendation system. “
This helps Netflix to determine the probability for any given show to be watched and attach a monetary value against it.
The aim will be to:
- Acquire new customers
- and extend the lifetime value of existing customers
E.g., if they were to create a show that caters to the taste of 0.5m users then they can use some of the profits generated by these people. The number of monthly revenues that this content may keep the users hooked is one indicator of how much can be spent on the content. But it will not be a simple calculation as some suggest you wouldn’t assume a person watches just one show per month/year or even at a time. You would also not invest all the profits of one customer on content, you would also take into account the future value of the show beyond immediate use and more. But the gross profit per user generated during the time of the consumption of the show is still one strong indicator of what we can invest in the respective content.
Ultimately, CAC, LTV are unit economic measures that can support investment decisions. One way to do this is by applying the unit customer measures to segments such as the various larger and smaller taste communities.
And when original shows like Stranger Things, Narcos, Orange Is the New Black, the docuseries Making a Murderer, and the platform’s Marvel titles become hits, significant numbers of sign-ups follow. Netflix needs more of its own must-watch media to woo subscribers and keep them streaming. Otherwise, they can easily cancel their memberships, and Netflix’s on-demand entertainment empire crumbles” [Buzzfeednews]
(2) Technology investment
Looking from a traditional accounting perspective, we could argue that technology costs are a one-off fixed cost and be done with it (i.e. they would not feature in the above calculations). But we all know that this is not the case any longer. If you stand still technologically, you can pack up and go within a few short years. Technology & development requires constant reinvestment.
Here is a really comprehensive comparison across the largest competing streaming companies (you can soon add Walt Disney and Apple to that). Apart from pricing and content, technical features are the third element playing a big role in customer acquisition and retention. With switching cost being almost zero, people would start switching away if their streaming platform was to fall behind in the feature set for a prolonged period of time.
Further, technology investment can lead to engagement and more consumption, thus higher LTV.
(3) Marketing
In their annual report FY18 (pg. 6), Netflix lays out the reasons for cancellation (i.e. churn):
Members cancel our service for many reasons, including
- a perception that they do not use the service sufficiently,
- the need to cut household expenses,
- availability of content is unsatisfactory,
- competitive services provide a better value or experience and
- customer service issues are not satisfactorily resolved.
And their response to churn:
We must continually add new memberships both to replace canceled memberships and to grow our business beyond our current membership base. If we do not grow as expected, given, in particular, that our content costs are largely fixed in nature and contracted over several years, we may not be able to adjust our expenditures or increase our (per membership) revenues commensurate with the lowered growth rate such that our margins, liquidity and results of operation may be adversely impacted. If we are unable to successfully compete with current and new competitors in both retaining our existing memberships and attracting new memberships, our business will be adversely affected. Further, if excessive numbers of members cancel our service, we may be required to incur significantly higher marketing expenditures than we currently anticipate to replace these members with new members.
The last sentence here also shows their risk mitigator in case of excessive cancellation: increased marketing spend. Of course, if they endure persisting high churn rates they will need to invest more into content and features. Ramping up marketing beyond what was planned can only be a short-term measure. High churn would need to be tackled through various long term actions.
Virtuous cycle
This then nicely closes the loop to a virtuous cycle of investment, subscriber growth and lower unit cost. It was prominently featured on page 1 of their 2010 annual report:
“Our core strategy is to grow our streaming subscription business within the United States and globally. We are continuously improving the customer experience, with a focus on expanding our streaming content, enhancing our user interfaces and extending our streaming service to even more Internet-connected devices, while staying within the parameters of our operating margin targets.
By continuously improving the customer experience, we believe we drive additional subscriber growth in the following ways:
- Additional subscriber growth enables us to obtain more content, which in turn drives more subscriber growth.
- Additional subscriber growth leads to greater word-of-mouth promotion of our service, which in turn leads to more subscriber growth at an increasingly cost-effective marketing spend.
- Additional subscriber growth enables us to invest in further improvements to our service offering, which in turn leads to more subscriber growth.”
This is how the described cycle looks like:
It does show the three elements (that we just covered above) where Netflix invests their money in:
- Content
- Technology
- Marketing
You will see these elements in the larger flywheel when we come back to it at the end.
Netflix & the Subscription Business Model
Netflix Case Study: A real-world innovation, finance and unit economics education book presented with Netflix's 20-year journey as the real-world innovation example.
Probably the most engaging and empowering "textbook" you will find on these often-dreaded topics (without being a "textbook" as such).
You will also get access to the portal of 11 tools with downloads and templates.
Your top resource to understanding the Subscription Business Model including the crucial financial and economic concepts that its built on.
Must-know knowledge for any entrepreneur & intrapreneur!
What’s the outlook for Netflix?
What you believe about Netflix future depends on your assumptions. I have chosen two articles from opposing ends:
- This is a piece that highlights that Netflix vast negative free cash flow is no problem for the reason of revenue generation lagging 1-3 years behind content investment (though as mentioned above this is only true for self-generated content, not licensed or acquired content). So to be fair to Netflix you would need to compare revenue growth today to some extent to investment spend made 1-3 years ago (pro-rated for investment into created content vs the other two types).
- Here is another piece that highlights exactly the same point. However, it states that revenues have not grown anywhere near content expenses which, in their view, constitutes a vicious cycle
The latter source also lists a number of alternate services:
- Facebook Live – Free
- YouTube – Free
- Twitch – Free
- Roku – Free
- Crackle – Free
- CBS all access – $6
- Hulu – $6
- Starz – $9
- Amazon Prime Video – $10
- Apple – $10
- Showtime – $11
- Netflix – $13
- HBO Now – $15
- Sling TV – $25
- and many other
More competition emerging
Add to this list the recent announcements of Walt Disney and Apple (not exactly light-weights) of developing their own streaming content / platforms.
Take Walt Disney with a greater focus on kids/families (with Walt Disney content) and sports (ESPN+). Will people be willing to have Netflix and Walt Disney which you could argue are somewhat complementary? Or will people eventually opt for one? Will they subscribe to Netflix for one month say every 2-3 months to binge-watch what they have missed (or binge-race new shows) and then opt out again for a while? Just some questions that are hard to answer (and more to come below).
Other types of content
And then, of course, there are other types of content that also compete for the user’s share of wallet.
Some important questions
Who is right? Who is wrong? We can’t know that. It really depends on how assumptions pan out in reality and the actions that the participants take (esp in terms of investment decisions). As you have seen from Netflix free cash flows (i.e. funding of content via debt) there are important assumptions around future user growth and further content costs.
Your assessment of the outlook depends on how you think some of the key assumptions will develop:
- How much further room is there to attract international subscribers?
- Will churn rates stay low compared to other OTT offerings (esp cable)?
- How much more money will Netflix need to invest in content (will it taper)?
- How much revenue can Netflix make from their content assets once fully amortised?
- How many subscribers will defect to Walt Disney, Apple or other streaming providers?
- What will happen when other studios/distributors start their own streaming services (e.g. TimeWarner)?
- How churn evolve when others don’t licenses out their content to Netflix (e.g. Time Warner’s Friends)?
- How many more competitors will emerge and what pricing will they have (US market and internationally)?
- How will Netflix’s pricing power develop (will there be any if everybody creates massive amounts of content)?
- How will the talent pool of filmmakers and actors develop?
- Which other revenue sources can Netflix tap in (tiered access / channels)?
- How will content tastes evolve over time?
- How will user-generated content evolve?
- How will the known competition evolve?
- Will there be new types of competition (unknown today)?
- and more
Lot’s of questions that make for an exciting industry.
Premium Content
We have two great product bundles that cover the platform business model:
- The Social / Search platform Super Bundle with several case studies on the platform business model.
- And the Sharing Economy Super Bundle covering platform business models based on tangible goods.
The Netflix flywheel
Let’s close with a few more word on the flywheel.
(1) The value creation cycle starts with the injection of cash. For more mature companies this should be generated by profits. And to some extend Netflix’s dependency on external funding is surprisingly high but potentially justified by the high growth rates that it commands (and their ambition to be first mover – whether or not that will translate to a sustainable lead is a different question).
(2) The investment cycle allocates these funds to concrete decisions in order to create value for the customers. Major categories for Netflix are content, technology and marketing. The former create assets that last for a number of years and translate to revenues (value capture). Marketing aims to bring in new customers well in excess of churn.
(3) The flywheel can be humming well and be a virtuous cycle. But if the red items become large (churn, content that doesn’t stand out enough compared to competing content) the flywheel loses momentum and liabilities and dependency on external funding becomes a burden rather than fueling the machine.
It is ultimately the investment decisions in relation to the competitor moves that determines Netflix’s trajectory.
Replace the word “content” above with what is the key ingredient to your innovation and you can basically use the same model for your idea!
I hope you have found value in this article – please help us to fuel our virtuous cycle by sharing!
Related articles (Netlifx archives)
Strategy & biz model
We are using Netflix as an in-depth example of the subscription business model. Learn about:
Financials
We are explaining important financial concepts using Netflix as our deep, real-world example:
Unit Economics
We are explaining important unit economic concepts using Netflix as our deep, real-world example: