Comment by cek | Hacker News Reader

Comment by cek | Hacker News Reader

Parent Full thread

That Netflix needs to keep making these changes is simply a symptom of a brutal endemic problem:

    In general, subscription business suck.

Most subscription businesses suck because they are dependent on CAC (customer acquisition costs; the cost required to land a new sub) and churn (the percentage of subs that leave the service).

For example, a typical gross CAC for a business like netflix is $200. At $10/mo the subscriber has to stay a sub for almost 2 years before any profit is made. Meanwhile the company had to spend the $200 up front, burning through capital. Also, meanwhile, 15-20% of the customers leave every month. Those customers will never pay off the costs the company spent to aquire them. It's an never-ending spiral of death. As seen by Vonage, Netflix, etc...

But Netflix has another endemic problem: They are the bitch of the content providers. Netflix owns NOTHING (especially in their streaming only biz). All it takes is one content provider to balk (Starz) and the only way they'll get new subs is to spend more on marketing which raises CAC. Which accelerates death.

Name a few successful subscription businesses. Successful (in my mind) means PROFITABLE. Revenue does not mean sh. Come-on, name 'em. (Those that you do name will have some sort of monopoly or exclusive right to a physical asset).

This is why, as an investor, I will never invest in companies that think they can be successful with a subscription business model.

This is also why Apple is kicking ass. Apple gets paid up front. The only way to go.

Interesting, I would have argued the opposite: well thought-out subscription businesses are goldmines. Once you have a customer, you aren't billing them a single time -- often, you're billing them repeatedly for the option to use your service, whether or not they actually use your service. A customer has to take an additional action to stop paying you.

Are you arguing that non-subscription businesses don't have customer acquisition costs? Or just that CAC for subscriptions are higher?

"Name a few successful subscription businesses. Successful (in my mind) means PROFITABLE. Revenue does not mean sh. Come-on, name 'em. (Those that you do name will have some sort of monopoly or exclusive right to a physical asset)."

What about most SAAS businesses? Or most magazines? Hosting providers? World of Warcraft?

But I totally agree with you about the content providers like Starz calling the shots and ultimately deciding whether Netflix lives or dies.

In Q2, Netflix SAC was $15.09 and churn was 4.2%. [1]

Netflix has been profitable for several years.

[1] http://files.shareholder.com/downloads/NFLX/1290701144x0x485...

That is an interesting perspective, but it runs directly counter to my experience. I've experimented with many different business models; building hardware and selling it (and then renting co-location) charging a setup fee that covered the parts costs (and then charging a just above cost fee for running the server) and traditional renting.

Renting hardware, for me, has been vastly more profitable than any other option.

I can build a 32GiB ram server with an 8 core opteron, 4 'enterprise sata' disks, etc... for well under fifteen hundred bucks. I could probably sell it for two grand, if I spent a lot of effort on sales.

Alternately, I can rent the thing for $300/month. Heck I can rent a three year old server with similar specs for $300/month. This is without really trying; I thought I was going to have to part out those old servers, but people keep emailing me asking for high-ram dedicated servers. I'm honest about the age of the hardware, and yet I still have more demand for old servers than I have old servers. (I'm working on moving my xen customers on to new servers, so I will have more old servers shortly.) Those old servers cost maybe $60/month for me to host; the new ones? $30/month.

As far as I can tell, you can charge vastly higher prices if you rent your product rather than sell it.

From my perspective, renting is more profitable than selling even if I'm paying credit card rates on the capital.

I self-finance by offering people a discount if they pay 3, 6, or 12 months ahead; 5% for 3 months, 10% for 6 and 20% for 12 months. It works quite well, new signups on a server will pay for the hardware the first month. 20% is kind of costly capital, but it still makes more sense for me to borrow and rent than to sell.

"Name a few successful subscription businesses. Successful (in my mind) means PROFITABLE. Revenue does not mean sh. Come-on, name 'em."

USA Today, the Wall St. Journal, People Magazine, porn websites, zipcar, Red Hat, salesforce.com, github, dropbox, evernote.

Oh dear...where to start: * 37Signals * Salesforce * <insert name of food product> of the month club (you said profitable - not Fortune 500) * <insert name of major gym> membership * Does Amazon Prime count? * Costco - you need to pay to be a member, whic his, by definition, a subscription business * Successful hosting providers - I myself use the great Bluehost.com

Though you could argue that paying for and building a gym gives you exclusive right to a physical asset.

Out of hand rejection of subscription businesses is your prerogative, but there are businesses which charge monthly rates that have surperior proritary products or services, or which use the membership as a means to incent more purchases, that can be very successful.

<full disclosure - I love CEK :)>

I have no way of knowing whether 37signals is actually profitable or not, but perhaps Salesforce is a real example. You found ONE.

Your gym membership does not count: Physical asset.

Amazon Prime is not a business. And I guarantee you it is NOT profitable anyway. It's a marketing program.

Hosting does not count: It's a physical asset.

Costco is not a subscription biz. It's a retailer that uses membership as a marketing ploy to drive loyalty.

You said, subscription based businesses. I am not sure what physical assets have to do with anything. Even ignoring Netflix's physical DVD library, if you consider hosting a physical asset, then I think you'll also have to qualify Netflix as a "subscription service which provides automated hosting & licensing of select content for subscriber's convenience".

>Hosting does not count: It's a physical asset.

If, instead of 'subscription business' you mean "businesses that rent access to software without much by way of per-user physical infrastructure" then yeah, I think some people would agree with you. I would. businesses think that because they have a low marginal cost per user, they think they can rest on those high margins. This might be what you are talking about, and if so, I agree. I mean, if you have a business with margins that are very high, investors notice, and they start trying to horn in on your market; either investors are willing to throw gobs of cash at re-writing all your software then saturating the market with marketing, or maybe someone open-sources something that can do a lot of what you can do, and suddenly you have a bunch of small competitors using the open-source software willing to compete with you at a more reasonable margin.

I was thrown off by your use of apple's hardware business as an example; I mean, selling intellectual property divorced from infrastructure is a hard problem, really, it's a different problem from the subscription vs. one time payment business models.

Reply on news.ycombinator.com