The Netflix Churn Challenge

| About: Netflix, Inc. (NFLX)
This article is now exclusive for PRO subscribers.

The critical question here is: "Can Neflix (NASDAQ:NFLX) continue to sustain its growth with its high churn rates?" There seems to be a possibility that Netflix is going to exhaust its potential subscriber base. I previously wrote an article about Netflix to review its revenue growth, which is very impressive. Not only has NFLX managed to grow revenue each quarter over that quarter a year prior, but the growth rate has almost without pause increased. This growth has risen from single-digit growth to over 30%. This is truly an impressive feat; however, as I noted earlier, it cannot be sustained indefinitely.

Netflix's revenue growth is driven by significant subscriber growth and tremendous acquisition of new customers. This is offset by declining revenue per subscriber, but this is often noted as part of a broader business strategy by shifting away from DVDs delivered to homes and to a streaming service. Table 1 shows a summary of the last nine years of subscriber additions.

Table 1: Historical Subscribers 2002-2010 (Thousands)
Year Beginning Period Subscribers New Subscribers Lost Subscribers Ending Period Subscribers Lost/ Beginning
2002 456 1,140 (739) 857 -162.1%
2003 857 1,571 (941) 1,487 -109.8%
2004 1,487 2,716 (1,593) 2,610 -107.1%
2005 2,610 3,729 (2,160) 4,179 -82.8%
2006 4,179 5,250 (3,113) 6,316 -74.5%
2007 6,316 5,340 (4,177) 7,479 -66.1%
2008 7,479 6,859 (4,948) 9,390 -66.2%
2009 9,390 9,332 (6,454) 12,268 -68.7%
2010 12,268 16,301 (8,559) 20,010 -69.8%

Data sourced from NFLX 10-K filings at and Some data may be slightly off due to rounding.

This is a very interesting set of data that generates a couple of critical challenges about NFLX's ability to continue to sustain growth, which is probably a key factor in driving its valuation.


  1. The ratio of lost subscribers to beginning subscribers is quite high.
  2. The ratio of lost to beginning declined until 2007, but shows signs of worsening now.
  3. The straight sum of 2002's begining subscribers and all subsequent new subscribers is now over 50 million, which represents close to half of all U.S. households.

This data raises questions around the sustainability of Netflix's subscriber growth, but there are some possible explanations. Table 2 shows some simple projections assuming 20% Ending Period Subscriber annual growth and continued 65% lost subscribers each year. Historically, Ending Period Subscriber growth has ranged from a low of 18% to a high of over 70%. The average growth has been near 50%. This data is projected analytically and does not attempt to account for any competitive business pressures, customer alternatives, etc.

Table 2: Projected Subscribers 2011-2014
Year Beginning Period Subscribers New Subscribers Lost Subscribers Ending Period Subscribers Lost/ Beginning
2011 20,010 17,009 (13,007) 24,012 -65.0%
2012 24,012 20,410 (15,608) 28,814 -65.0%
2013 28,814 24,492 (18,729) 34,577 -65.0%
2014 34,577 29,391 (22,475) 41,493 -65.0%

Data is based on my own projections with the assumptions provided above. This data also shows a markedly lower level of subscribers than my previous analysis. That analysis focused more on customer acquisition costs and marketing budgets.

Some quick comparison data Nielsen (NYSE:NLSN) estimated that in 2009-10 there were 114.9 million U.S. television households. The Worldbank shows Canada's population in 2009 at approximately 33.7 million, or assuming 2.5 people per household, that there are 13.5 million households. This gives NFLX a total target market of about 128.4 million households, assuming that at most one household would have one subscription.

This is starting to present a problem in that these assumptions are potentially inconsistent. In order to meet the above numbers, Netflix would require another 90+ million new subscribers or, from doing the math from 2002, 140+ million subscribers -- which would represent more households than the U.S. and Canada combined, even assuming some modest growth in households from 2009 to 2014. There are several possible explanations for the above presumed inconsistency:

  1. Lower growth. This growth is simply not going to occur, which could have implications to NFLX's valuation. An annual year-end subscriber growth rate of 10% and the same lost subscriber rate would provide year-end subscribers in 2014 of just under 30 million.
  2. Lower loss ratio. The ratio of lost subscribers to beginning subscribers could be much lower. However, putting this at -30% would still require another 50+ million new subscribers over 2011-2014.
  3. "Indecisive" subscribers. Perhaps many NFLX subscribers churn themselves multiple times. Perhaps someone is a subscriber one year, drops for a while, returns as a subscriber and repeats. This way the same household can represent more than one new subscriber spanning over the time frame. This would be interesting data for NFLX to provide. Possibly, the same household has been counted as four or even five new subscribers throughout the years.

Another interesting question is whether this level of subscriber loss is common. For comparison, I pulled data from Sirius XM Radio Inc. for Table 3. First, it should be noted that Sirius acquired XM Radio in 2007, dramatically increasing its subscriber base.

Table 3: SIRI Subscribers 2004-2009 (Thousands)
Year Beginning Period Subscribers New Subscribers Lost Subscribers Ending Period Subscribers Lost/ Beginning
2004 261 987 (105) 1,143 -40.2%
2003 1,143 2,519 (345) 3,317 -30.2%
2002 3,317 3,758 (1,051) 6,024 -31.7%
2007 16,234 2,337 (1,222) 17,349 -7.5%
2008 18,921 1,713 (1,630) 19,004 -8.6%
2009 18,516 1,883 (1,626) 18,773 -8.8%

Data sourced from 10-K filings for SIRI at

This data shows dramatically lower losses relative to base, even during the early years when growth rates were very rapid. The later data seems to show a more mature business characterized by lower growth. One significant other fundamental difference is that SIRI data shows stable to increasing levels of revenue per subscriber.

NFLX seems to have a very high risk of slowing growth. Its historically high churn rate is possibly exhausting its potential subscriber base. Eventually, individuals will decide whether they are NFLX subscribers or not. However, additional information would be required to conclusively determine this, given the potential for indecisive subscribers to join, leave, and return over and over. With NFLX's stock price pushing a 52-week high at $223.20 per share, investors might not even care.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.