About Tom Evslin

Video Profile of Tom Evslin

Follow Tom Evslin on Twitter


Add to Technorati Favorites!
Powered by TypePad
Member since 01/2005


« DARK WATERS: Programming in Small Places | Main | How the Telecom Mega-Mergers Might Make Sense (but probably don’t) »

Subscription Pricing

Customers like subscription pricing.  In fact, they are willing to pay a premium for predictability and simplicity.  If you can find a way to price a service at a flat monthly rate, you can make a better profit per customer and attract more customers than if your pricing is based on reading a meter.  You also save a fortune in detailed billing, dispute resolution, and issuing credits.

Netflix is the most recent example of this I’ve seen.  My daughter gave me a trial subscription (viral marketing at work) and it’s a big hit in our house.  For $17.95 per month, you get all the DVDs you want to watch.  The hitch, which isn’t very important to us, is that you only get to have three at a time.  You go online, make a list of the movies you want to see, and very quickly three arrive.  You can keep them as long as you want.  There’s no such thing as a late fee.  But you don’t get any more DVDs until you return one – they come with a handy postage-paid return mailer.  Almost as soon as you send one back, you get another flick.

That’s simplicity itself.  I don’t think I used to spend $17.95/month on movie rentals but that’s irrelevant.  I hate late fees.  I hate being rushed.  I like having a couple of unwatched movies around for spare time.  Of course, having it all come by mail is also great and could have been done without subscription pricing but, to me, the subscription model with no minimum commitment makes this a winner.  If I had to worry about getting to the Post Office in time to avoid late fees, I don’t think I’d be a customer.  And Netflix gets more out of the Evslins than Blockbuster ever did.  I just noticed, by the way, that Blockbuster now has a similar service priced at $14.99.  Netflix will have to fight for their lives but they got first mover advantage.

Back in the early days of popular use of the Internet, ISPs including AOL and MSN charged by the minute for dialup access with no cap.  People were afraid to sign up because they thought they might leave their PCs on and get huge bills.  Some small ISPs experimented with subscription pricing; all the access you could use for $19.95.  When we started AT&T WorldNet Service, we borrowed that idea and popularized it with the still hugely powerful AT&T brand behind it.  Some said we’d go broke; others that we would ruin the Internet.

To hedge our bets, we also offered a metered access plan.  We didn’t want to lose out on people who planned to spend less that $19.95 per month.  To our surprise, people typically converted themselves from metered access to subscription when their monthly bill was around eleven or twelve dollars.  And their usage didn’t spike after conversion.  People were paying a premium for predictability and simplicity.  With a subscription plan, they didn’t feel they had to keep track of minutes to make sure they weren’t being overcharged and they weren’t worried about surprises.

The rest of the industry was forced into monthly subscription pricing.  There were some teething pains, especially at AOL which didn’t have enough ports available for the extra hours its young customer base had time to use. However, I think monthly subscription pricing was one of the most important factors in the explosive growth of Internet access during the second half of the 1990s.

Of course, the subscription plan also had much lower costs for us as a provider.  We took a hard line that we would not provide usage detail to subscription customers so we could truncate usage information early in the processing cycle.  There were no disputes with subscription customers over whether they actually were online at any particular time.  The subscription customers were also more forgiving of occasional slow service because it didn’t increase their cost and they knew we had no incentive to keep them on longer.  (For more on WorldNet, see Lessons From The Crypt #1, #2, and #3)

Today a flat monthly rate for domestic phone calling is offered by all US VoIP providers that I’m aware of.  Traditional service providers have been forced to respond with unmetered subscription plans of their own.  Customers like simplicity.

There are two good arguments against subscription pricing:

  1. If the subscription rate is high enough to return a profit for customers with typical use, it will be too high for customers with low usage so you will lose them.

  1. Some customers will “abuse” the fixed price and overuse the service; perhaps even resell it.

At the price of some complexity, wireless operators have temporarily overcome these problems by establishing subscription tiers of use with different monthly rates entitling users to different size blocks of “free” minutes.  My guess is that competitive pressure will collapse this into just a couple of tiers with the top – and very popular tier – unmetered.  The complexity cost even with blocks of “free minutes” is still too high.

There is always room, however, for a metered low tier for budget conscious customers – if you can serve them profitably!

The problem of “abuse” is overstated although it is real.  With many businesses, the highest variable cost is the cost of subscriber acquisition.  Following that are a number of per subscriber costs like billing, bad debt, and customer care which are not directly related to usage.  In many telecommunications applications, there is no incremental cost to the provider when the customer uses the application.  (Note that “access charges” levied by monopoly providers of a last mile phone connection are an important exception to this.)  What is important is that the provider has to invest capital to size the network for peak usage.

A provider whose capital cost is determined by peak usage, doesn’t really care whether “abusers” camp on the service during nonpeak times since the provider has extra capacity available then.  During peak times, the normal, nonabusing users are on in such numbers that they dwarf the small number of “abusers”.  The experience of providers I’m familiar with is that small percentage of “abusers” (I’m not sure they’re really abusing anything) account for a relatively large percentage of total usage.  But remember that, since the “abusers” are a small percentage of total users and peak time is when a high percentage of total users are on line, the “abusers” can’t physically account for a high percentage of peak usage and therefore don’t significantly drive capital costs.  If you are planning a service, you need to do a very careful spreadsheet on this.

Subscription pricing for telecommunication is nothing new in the US.  Except in New York City and a few other places, unlimited local calling has been an option for as long as I can remember.

It is important that there be some gate to control usage even if there is no meter.  For example, Netflix only lets you have three DVDs at a time.  With unlimited calling as it offered today, you still can only make one call at a time.  When you buy subscription Internet access whether it is dialup or cable or DSL, the total load you put on the network is limited by the amount of bandwidth you have subscribed to.

Simple pricing, I think, always drives out complex pricing.  Finding a way to price your product or service on a subscription basis is worth a lot of thought.   Netflix sure has a clever way.  Sometimes it is necessary to be as inventive as they have been to find a way to offer your customers the benefit of subscription pricing without risking having your own costs driven uncontrollably through the roof.


TrackBack URL for this entry:

Listed below are links to weblogs that reference Subscription Pricing:

» A Good Argument For Renting, Not Owning, Music from digitalmerging.la
A couple of days ago, I posted an entry about owning downloaded content, as opposed to just renting it. Well here's a blog entry by Tom Evslin, who started flat-rate pricing in the ISP field with AT&T's Worldnet, that makes a good argument for the subs... [Read More]

» Customers Like Subscription Pricing from robhyndman.com
Tom Evslin with yet another meaty post - this one on why subscription pricing is a good thing. [Read More]

» Subscription Overdose? from Opine Online - The Rubicon Blog
Fred Wilson points to a great read by Tom Evslin on subscription based pricing. He brings up Netflix as his first case study then goes on to talk about the flat pricing of dial-up services at AOL and AT&T WorldNet or the pricing of most VoIP service ... [Read More]

» The Benefits of Subscription Pricing from AnumatiNews
Tom Evslin offers his thoughts on competitive strategy in an article on the benefits... [Read More]


Tom OKeefe

Very nice post! According to Bizak.com subscription models average $1.02 Earnings per Visitor (EPV). This is much better than most advertising models that range from a low of $0.06 EPV (Google Adsense) to $0.19 EPV (In House Advertising.) Even though subscription models tend to be more profitable than advertising models they still are only used by 8% of the startups listed on Bizak.com. Advertising makes up 58% of online revenue models with Adsense representing 21.5% and In House Advertising 25%.


what ever u have to sell u should keep in mind that u have to serve good thing

Andrew Odlyzko

It's great to see this account of how subscription pricing came to AT&T

It's amazing how often AT&T ran into the power of subscription pricing,
and how each time it had to relearn the lesson. As far as I know, the
people at AT&T Wireless who came up with the AT&T Digital One Rate plan
did not know about your WorldNet experience, nor about the half a dozen
other similar cases in AT&T history.

I have a paper on this topic, by the way, "Internet pricing and the history
of communications,"


that was published in 2001 in Computer Networks, vol. 36, pp. 493-517,
but I was writing on this topic already in 1996, for example in the paper
"Fixed fee versus unit pricing for information goods: competition, equilibria,
and price wars" (with Peter Fishburn and Ryan Siders) that was published in the
July 1997 issue of First Monday,


but I could not get people inside AT&T to pay attention to this question.
(In particular, they were then deeply involved in the Mondex scheme, which
this paper showed was unlikely to succeed.)

Charles Kemper

Once you add it all up, the average consumer (head of household) can spend anywhere from $250 to $500 per month on subscriptions services - for communications & media alone!

Andy Havens

Having worked in marketing in the cellular industry for 10 years, during the decade of most extreme growth, I can tell you that flat-rate pricing (as we called it then) was very seriously considered... and rejected... for the forseeable future. It represents the ultimate comoditization of a product/service and the ultimate in price transparancy. Which simply makes for a very easy race to the bottom. You know the old saw... "In a pricing war you're at the mercy of your dumbest competitor."

There are times when a subscription model makes perfect sense. When the service is essentially an "always on" product, and the provider isn't in any way bearing an operating cost or suffering a bandwidth degredation when folks stay "on" all the time. Cable TV is a good example. It doesn't matter if everyone watches TV all day long. It doesn't cost Time Warner anything more to produce than if you watch 1 hour a day. Unlike Netflix and unlike cellular companies and unlike AOL and cable modem folk and electric companies etc. If there's a cost that scales, the price has to scale... somehow. Or you'll go out of business.

Can you build a subscriber model "above" your scaling costs? Sure. Local loop phone service does that everywhere. People pay $24-$35 for a service that basically costs the local phone company about $10 to keep going. We don't complain because, frankly, we have great local phone service in this country. And we've had lousy results with the few major attempts at competition in the local loop.

If subscription service was, acrosss the board, a wonderful pricing strategy, it would be used in many more industries. But people don't subscribe for very many services when you think about it. Food, clothes, most entertainment, vacation, tools, appliances, books, cars... we buy in discreet units. Why? Because some months we like to dress with lots of clothes (December), and some months we like to wear just a Speedo (August).


There are areas where subscription makes the most sense and there are areas where it doesn't. Paying for ISP minutes is lame but paying for blocks of cell phone time is not. DVD rentals for $7/month and $2/rental with no late fees would be quite a bit more appealing to me. The problem with subscriptions is that consumers don't want to lock in a monthly in perpetuity, hence the large churn at Netflix and AOL, for example.


I think the commenter above is making a great leap to label heavy users of Netflix as "abusers."

The most profitable element of the video rental base was people with families. These were the heavy renters. Younger and single people went to theaters, older people watched more cable television.

Heavy renters with families were the first target of Netflix in its competition with Blockbuster bricks and mortar. Their own finanacial reports directly noted this.

It would b very problematic to target your competitor's meat and potatoes and then categorize them as "abusers" or give them worse servic because they are, as expected, taking advantage of stated, expicit and implicit terms of their agreement.

The suggestion heavy renters are, "usually burning the dvd" is in my view unfounded and and uniformed. It is a smear as well.

DVD's are more and more like magazines, especially anything beyond top 10 new releases. they are very diverse and targeted to more and more specific market segments. Forgive the generalizations, but if a wife is getting a A&E Madam Bovary, the husband is getting an action film, the 12 year old daughter getting an Olsens twin flick, the five year old son is gettng Bob the builder, the mother in law is getting a 1950's film they aer not film pirates. Even a single mom with a kid who wants to get educational programing for their kid so that 5% of his average time in front of the TV is educational would fall into the above commentor's category of alleged "burners"?

My own feeling is Netlfix has inadaquate traction to comete with Tivo. They simply do not have anything but themost shallow effort at VOD where they are neither a rights holder not a bandwidth provider. ISP's are very diffent they provide an actual pipeline.


can't agree more - the comfort factor is particularly pertinent to encouraging take up of new services for which the customer have little understanding over what their usage might be.

mobile content is the first thing that comes to mind.......the existing pricing regimes insanely confusing (there are cellcos out there charging on a per kilobyte basis!?!?) the abuse factor you mentioned is also mitigated somewhat by limitations of mobile data networks (even the different variants of 3G) and handsets and the walled garden model around most of the content services......

but then it's not hard to see why many companies are still resistent to subscription pricing even when it makes sense......financial models r more difficult to fudge when there are less variables :-)


Great post......consumers love cost certainty and will pay a premium for it, as you have indicated. I am quite familiar with the netflix model and, imo, they do have a problem with "abusers". There are subs that will go through 30+ movies per month, usually burning the dvd and then shipping it back same day. The incremental cost of postage alone often eliminates what little profitability there was. I think, eventually, you will see netflix move to a model whereby you are limited to a set number of movies per month, say 20 @ $17.99. Still a great value proposition.

Post a comment

If you have a TypeKey or TypePad account, please Sign In.

Now on Kindle!

hackoff.com: An historic murder mystery set in the Internet bubble and rubble

CEO Tom Evslin's insider account of the Internet bubble and its aftermath. "This novel is a surveillance video of the seeds of the current economic collapse."

The Interpreter's Tale

Hacker Dom Montain is in Barcelona in Evslin's Kindle-edition long short story. Why? and why are the pickpockets stealing mobile phones?

Need A Kindle?

Kindle: Amazon's Wireless Reading Device

Not quite as good as a real book IMHO but a lot lighter than a trip worth of books. Also better than a cell phone for mobile web access - and that's free!

Recent Reads - Click title to order from Amazon


  • adlinks
  • adsense