Bloomberg Looks at Netflix's Churn Rate

Jonathan Weil on Bloomberg.com takes a close look at Netflix's churn rate, At Netflix, Sexed-Up Math, Half-Truths and DVDs.

The upshot: By my math, using the dictionary definition, Netflix's second-quarter churn rate was 16 percent, or 5.3 percent a month. And get this: For the 12 months ended June 30, Netflix's churn was 63.4 percent.

In other words, almost two-thirds of Netflix's customers canceled during the past year. You won't find stats like that in Netflix's disclosures -- unlike at TiVo, which shows both annual and monthly churn rates and labels them accordingly.

Comments

[gir]

63%? Cha! So over the last few years, more customers have quit NetFlix than customers they currently have?

Statistics can be tricky. I don't think those statistic can be taken at face value.

[CJ]

Early on in my MBA program years ago, I learned you can lie with statistics and claim anything you want as "facts." Who cares about churn rates anyway - NF is obviously still growing customer base.

[leonardodicrapio]

"Netflix has withdrawn its 2012 target and slashed its 2007 goal to a range of 6.8 million to 7.3 million.

Reaching Low

Even the low-end goal looks tough."

The author submitted this article at about 12:01 a.m. on Wednesday, August 1st.

Jonathan fails to mention that Netflix lowered the monthly fee for a majority of its customers by $1 on Sunday, July 2nd.

Jonathan also fails to mention that on Thursday, July 26th, Blockbuster raised the price of its Total Access program by $7 and then changed the name of Total Access to Total Access Premium. Blockbuster then introduced a new program called Total Access with limits on exchanges at price points that were on average $0.50 higher than the old Total Access program price points.

[leonardodicrapio]

"Jonathan fails to mention that Netflix lowered the monthly fee for a majority of its customers by $1 on Sunday, July 2nd."

I meant to say on Sunday, July 22nd.

[scott]

It looks like you are multiplying 5.3% by 12. You can't do that with an AVERAGE. You have to add the total number of cancels for the year and divide that by the total # of subscribers for the year. That will give you the number you are looking for.

You can't do it by month and multiply...

[FrankLev]

Scott, gir,

You are incorrect. Bloomberg is using the accepted standard churn rate determinants. this method is the standard used in service business like mobile phones etc for decades. This is quarterly or even yearly, not monthly.

For example Sprint mobile's churn is about 2% per quarter-- which is quite bad, worst of all the majors. Netflix churn is indeed 16% per quarter or over 60% in the past year. That is extremely bad for any subscription service.

Bloomberg also notes customer acquisition costs are rising. it looks like costs are rising 10% per year unabated. combined with high churn this is a big problem and fatal if not reversed.

Of note Hastings has publically many tmes citied blockbuster's churn rate as damging to blockbuster -- and Hastings has ALWAYS used bloomberg's method of calculating blockbuster's churn. Notably blockbuster's churn has been dropping steadily over the past year at the very same time as Netflix's has been growing!

[FrankLev]

"63%? Cha! So over the last few years, more customers have quit NetFlix than customers they currently have?"

Exactly.

[leonardodicrapio]

"For example Sprint mobile's churn is about 2% per quarter-- which is quite bad, worst of all the majors. Netflix churn is indeed 16% per quarter or over 60% in the past year. That is extremely bad for any subscription service."

For the most part, cellular companies lock people into 1- and 2- year contracts. Sure, there are some people with prepaid plans, but Sprint probably has about 20% of their customers who can leave in any given quarter (without paying a steep penalty of approximately $199).

Netflix's churn is high, also, because of the number of gross additions every quarter (which are mainly people checking out the service on a free trial). Netflix has gross additions every quarter of approximately 1 million customers. So, Netflix typically has gross additions that are about 15% of their paid subscriber base.

I don't know too many people who sign up for cellular service with Sprint (or Verizon or AT&T, both of who I know gives you a 30-day money-back guarantee) without the intention of actually keeping the service. The last time I was in an AT&T (not for an iPhone) I was there for 1-2 hours. Not exactly my idea of fun.

It is easy to sign up with Netflix (and try it out for free). It is also easy to cancel service. This causes Netflix's churn to be much higher than Sprint. A better comparison would be to another internet service whether it be DVD rental or music download (i.e. Napster).

"Of note Hastings has publically many tmes citied blockbuster's churn rate as damging to blockbuster -- and Hastings has ALWAYS used bloomberg's method of calculating blockbuster's churn."

Hastings can only give estimates of what Blockbuster's churn since Blockbuster doesn't release this data. As such I don't know how you know what method he is using to estimate Blockbuster's churn.

Netflix provides a lot of information about their churn, and their method of calculating is explained simply. Blockbuster provides no information on churn, but Netflix's (churn) critics always like to give Blockbuster a pass on that.

[Old Timer Too]

NF's churn won't equate in any manner with someone like a phone provider (or cable). This goes all the way back to record clubs where you were shipped a new recording each month and you could cancel after buying so many the first year. The churn NF is providing is remarkable in that it is as high as it is. I suspect a number of users are repeaters who sign up after xx days of non-membership for another "free trial" and then quit. Though I have to think that kind of "churn" can't account for 60+% of its membership. Losing over 4 million members in one year? (out of 7 million - someone can correct me if my figures are wrong)?

That's waiting for total disaster to strike. That figure, if accurate, says a lot about NF's caring about its customers. If anything, Hastings and crew should be figuring out how to raise the retention rate!

Maybe 20 million have tried NF over the years and all they have left are the 7 million or so still with them... Think about that one... Ack!

[Old Timer Too]

Ouch! I just took the time to link to the article. Bloomberg is well respected and an article like that is like a bad financial rating for an insurance company. There is no more "up" on the charts. It all goes downhill from here on out.

I would be interested in seeing a second "churn" rate - one that does not include non-subscribing free trial users and seasonal users. That would provide a better picture over the long haul.

If it weren't for the horrendous churn rate (no matter which one you use - Bloomberg's standard or NF's unique one - the future is not bright.

Hastings and crew: start figuring out how to retain your customers. Some of us left and may never come back...

[leonardodicrapio]

"I suspect a number of users are repeaters who sign up after xx days of non-membership for another 'free trial' and then quit."

I'm a happy Netflix customer, but there are definitely subscription services (i.e. EMusic) that only require you to have a new credit card and e-mail address to open a new account and get another free trial. If you haven't already, you should download new albums by Paul McCartney and Arcade Fire for free over at EMusic.

[superfunhappy]

They only let a person have a free trial once, son.

[superfunhappy]

and most people aren't interested in gaming the system.

[leonardodicrapio]

superfunnyhappy,

i paid for my second "free" trial at emusic. i subscribed to emusic's service for over a year before i let my subscription lapse. i bought about 30 emusic songs for half price from the circuit city outlet store on ebay, then found out it was actually just a "free" trial that i had paid good money for (no wonder it was half off -- it was worthless to begin with).

i had to create a brand new account to get the songs that i had legitimately paid for. since i bought the item off of ebay i don't think that all the disclosures were made by circuit city in the item description -- if i had picked up the packaging at a circuit city store i would have realized what it was.

that is how i found out how you could get a second or third free trial at emusic. when i found this information out i wasn't trying to game the system.

[superfunhappy]

Yes, you can definitely make it happen, inadvertently or not. I'm saying that most people are just going to go own a regular path whereby they'll only get one shot at a free trial. And aren't dedicated enough to try and get a second (or third) free trial.

By the way, I've seen people trying to sell the regular Netflix trial on eBay too. You'd think eBay would require full disclosure since it amounts to fraud.

[Edward R Murrow]

Bloomberg is a pretty reputable organization and I'm sure that Jonathan Weil must have had his math vetted by others at Bloomberg prior to publishing.

[Mike Madich]

Why would anyone rent for $25 + per month when you can get DVD's for a Dollar at the local video store? Netflix is a ripoff! Too expensive.