The First Law Of Internet Physics

I’ve written a lot about free vs paid here on AVC. There was a time I was obsessed about this topic. We even coined the word freemium here at AVC (thanks Jarid) during that phase. I’ve moved on to other obsessions but I still think a lot about it. I got in a long twitter discussion with a bunch of people yesterday about this topic, spurred by this tweet by Jonathan Weber:

That led to a long twitter discussion in which a bunch of people joined. I can’t find a good way to showcase the discussion so I’m not going to. But during the course of the discussion I tweeted this. I made it up last night but it’s something I’ve experienced many times over the years.

Isaac Newton observed some things about motion and encoded them into his three laws of motion. I think we should do the same thing with the Internet. There are some things that just are, and we should acknowledge them. I posit that one of them is this:

many users * low arpu >>>> few users * high arpu

I’ve seen so many people try paid content on the Internet and the result is less users, a lot less. You can extract a higher average revenue per user (arpu) from a paid model, but you get so many less users that is it better to extract a lower arpu with a free model and get many more users. I guess a corollary to the first law of Internet physics is that you can implement a freemium model on top of a free model and turn some of your users into high arpu customers, but they will always be a small portion of the total number of users.

That’s a long way of saying that you can do paid, but you had better have a free tier first and foremost as most users will go for that. And if you put too much of your content behind a paywall, you’ve effectively turned your core product into a paid one and you are back to {few users*high arpu}. So be careful with the freemium offering.

#Web/Tech

Comments (Archived):

  1. jason wright

    inside a black hole do Newton’s laws hold true?

    1. fredwilson

      good question

      1. jason wright

        blockchain = black hole?

        1. Chimpwithcans

          blackholechain

          1. Guest

  2. SD

    The entire TV indusry today is paid. The relatively small free consumer businesses (broadcast, hulu, YouTube) are essentially marketing efforts supporting a b2b license fee business.

  3. Guest

    I used to be an advocate for the paid business model for most startups.. Generate revenue and get funding to scale. Which does work for “some”. After reading more and more on freemium and experiencing the grind of attempting to generate profitable revenue, my tune has changed. I see all my peers that chose the freemium model growing at an alarming rate. Much faster than the majority of my friends grinding it out via the paid model. I am definitely a convert for starting your product or service out first as a freemium model then monetizing down the road. @wmoug:disqus

    1. William Mougayar

      The linkage between free and paid must be clear, and easy to understand and value.

      1. Liban Mahamed

        Great point, easy said than done This is the billion dollar answer If you know where the line line.

  4. Steve Gurney

    I understand this from a consumer model where other sources of revenue are possible. What’s the best model for business focused service, where demands for support are higher and tolerance for advertising/sharing of data are low?

  5. David Semeria

    Just for the record, freemium is not a business model, it’s a marketing model.The underlying business model is still subscriptions, you just consider the cost of servicing the free users as the marketing spend required to gain conversions.That said, it can be a very effective model because it’s based on people actually using the product.

    1. JimHirshfield

      And if you monetize the free users with advertising? Isn’t that a revenue/business model (IOW, not exclusively or necessarily a marketing model as you describe)

      1. David Semeria

        That’s simply two business models running in tandem.

    2. JamesHRH

      I don’t know that I agree. Your costs structure is impacted by your revenue model, in this case.

      1. David Semeria

        Can you explain your question better?

        1. JamesHRH

          When you look at your business, you think: how am I going to attract customers and then generate revenue. If you opt for full subscription, partial (freemium & paywall) or none, you affect how much content you can produce. So, its a business model decision.I think.

          1. David Semeria

            Up to a point. Some sites let you read above the fold for free, or read X number of articles per week in their entirety.In other words, the content is produced for and funded by the paywall. By allowing limited free access to what you’ve already got, the hope is to enlarge the paid base.The trick with freemium is knowing how much (or what parts) of your inventory (be it software, music, or media content) to give away for free.

          2. Mario Cantin

            Exactly! And to do it with the point of view in mind of the user who wants the free offering, in such as way so as to make her happy — not frustrated — all the while finding the right triggers that will make her feel she needs to upgrade to the paid offering without feeling she has been manipulated or coerced.

  6. Guest

    I used to be an advocate for the paid business model for most startups.. Generate revenue and get funding to scale. Which does work for “some”. After reading more and more on freemium and experiencing the grind of attempting to generate profitable revenue, my tune has changed. I see all my peers that chose the freemium model growing at an alarming rate. Much faster than the majority of my friends grinding it out via the paid model. I am definitely a convert for starting your product or service out first as a freemium model then monetizing down the road.. @wmoug:disqus

    1. Guy Lepage

      @wmoug:disqus, this was posted by me.

  7. Chimpwithcans

    If there are certain laws around fremium and media access developed or understood, then we might well see more cases like Apple Inc., which in my opinion is heading towards using media (music) simply as an entree into other more profitable services or hardware products (Apple pay, iMac etc).Gotta nail down those laws first though – I would say Spotify needs to have a think about this – where is the next step? could they build some hardware? Or stick with software that takes advantage of the captive market more effectively than the current business model does. Great post – a thought provoker!

  8. Seth Godin

    This is the SECOND law of internet content physics.The first law is that the only thing that people with a choice will do is something that they think is worth doing.Things that people decide matter >>>> things that you want to have happenWhich means that you have to matter to a small group, the smallest possible group you can sustain, before you can even begin to deal with the second law.Which means that the best way to test if you matter is to build something that SOME people would even pay money for.If you can do that, then sure, make it free and let the second law take over. But if you can’t, merely asserting that people should use it because it’s free is a little narcissistic, which is something MBAs not worth investing in are good at.

    1. Twain Twain

      The first law is that people will consume what they’re already used to consuming unless impacted upon by a piece of content that changes the direction of their thinking and tastes.That impact can happen randomly, stochastically.The challenge for content providers is to try to figure out what it is within that content and the context of its delivery (timeliness, relevance, accuracy etc) that impacts the content consumer to change their thinking and tastes and to adopt the path of their content.

    2. Matt Kruza

      Good point on the “first law”. However, I think one thing people don’t do often enough is try to robustly indicate why something should matter to users. I don’t mean in a dopey, moralistic, idealistic manner. Just that you should believe enough in WHY people should pay for content and then see where it goes.

    3. JamesHRH

      That’s a first law of humanity, a great test and a very good point.

    4. ShanaC

      So should everyone charge and then go free if they go bust, then go charge again once they matter again?People can be fickle. It can be a while until you get critical mass numbers in audience development terms that make sense to charge. So when exactly should audience development flip the switchAnd don’t get on my most content marketers don’t realize they might be running publishing arms. They won’t charge since they are marketing a product.Hard to be a publisher these days.

    5. fredwilson

      Well now we have two rules although you failed to provide the mathematic equation seth ;)If we can get a third we can get into science books for the next five centuries. But of course there won’t be textbooks anymore

  9. William Mougayar

    A second rule (or observation) is- paid content isn’t always “better” content.

    1. Adrian Palacios

      Any examples that come to mind of paid not being the “better” choice?In many cases I’d much rather go for paid:HuffPo/etc vs NYTimes/Financial Times/etcYoutube vs Netflix/Amazon InstantPandora vs SpotifyIn all those cases I opt for the latter options (and pay for them).

      1. William Mougayar

        I was focused on the web content (excluding music/video). Take startup/tech/VC content, and the best stuff is all free and wildly available on blogs such as this one. If the NYT or FT write something about startups/tech, and they put it behind a firewall, that doesn’t mean it’s better. There are exceptions of course, e.g. Harvard Business Review, but even they are giving a lot of free stuff online too.

        1. awaldstein

          You raise something interesting with this type of content.There are only two model free and advertising at scale. And subscription.This is not changing any time soon.

    2. LE

      Will add that sometimes content that isn’t widely read or shared (perhaps because it is paid) is better than free content that everyone has access to. If you believe that knowing certain information vs. your competition is an advantage. Which I do.

  10. Twain Twain

    Haha, this is a great post intellectually and appeals to the science nerds amongst us!So…I tend to think of the content payment model in terms of the laws of thermodynamics and as it applies to electron stimulus in atomic shells.The core is where the electrons (content) is free. The next layer out involves providing some excitation energy to get the content to a higher quality and for people to WANT TO jump to that level. Subsequent layers out require even more excitation (payment) to gain access to exclusive content which is beyond the core level. An example would be an interview with Fred talking about things he doesn’t share on AVC.Each level has a latent energy as well as additional excitation energies. Latent would be found in lurkers who read the content, up vote and absorb others’ comments. Additional excitations would be when one of us decides either to be provocative and challenge what Fred or anyone else has said……..That then makes the content migrate from the core free part written by Fred to other levels and realms of thought; sometimes to the uncanny edges of fractal space.Anyway, I’m thinking about this free vs paid content in particular this morning because…I just signed up to Condé Nast’s first-ever International hackathon.________In 2014, I’ve completed 5 hackathons spanning 5 different sectors (educational non-profit for SF Treasury Office, financial services—CapitalOne, telco—Tata, auto—Edmunds, FMCG—Unilever) and won prizes in all 5 hackathons, with 5 different teams. That maybe says something about my versatility and ability to work with other people. The prizes have been a mix of cash and incubation.The Condé Nast one will be my first media hackathon so I’ll probably borrow Fred’s Newtonian law of ARPU and put it into the product pitch!

  11. John Revay

    As I read this post, I was thinking about Taylor Swift and her decision to pull her music off spotify this past week.I like her music and will find another way to listen.

    1. Maureen Scott

      why won’t you buy her music or is it that you prefer the delivery method (streaming) model?

      1. John Revay

        We actually did, my wife purchased on iTunes and then I shared it to myself.

        1. Maureen Scott

          The great “sharing” option, love it.

          1. John Revay

            I wanted to use the phase “Ganked it” in my comment (I once saw that term used by Fred on a photo he once posted)…but after looking it up – probably not the best term.re: Sharing – Apple has made sharing easier w/ iOS8http://www.apple.com/ios/wh…

    2. Anne Libby

      Individuals live on the right hand side of Fred’s equation.And individuals create the pixels behind the left hand side of the equation.There’s something missing. (Dark matter?)

    3. JimHirshfield

      You’re 13. Stay young John, stay young.

      1. John Revay

        Shake it up 🙂

      2. JamesHRH

        She has the only platinum album of 2013 and is ruthlessly hard on herself to be better (where better = more popular).My 12 yo calls her music ‘completely thought out’, which is not really a diss or a compliment.

        1. Chimpwithcans

          I keep wondering whether she is the one behind this decision – the industry being what it is, and record company exec’s working behind the scenes – they manage careers in the background – I’m sure there is more than meets the eye with this decision. Why walk away from $6m??

          1. JamesHRH

            Comes from a high achieving family (Xmas tree farm story not untrue, but not family’s main income…..was a hobby for her Merrill Lynch employed father).Basically forced her parents to drive her to Nashville @ 14.She is of a personality type that likes to control outcomes, achieve and be recognized for doing thing the right way.I would be more surprised if she was hands off than if she was hands all over everything.

          2. Chimpwithcans

            Interesting character info, thanks. The reason for my remark is that Spotify pays out 70% of its income in royalty payments to labels and publishers, who then pay musicians according to the terms of their contracts. I suppose Spotify should have negotiated a more binding contract!

          3. LE

            Basically forced her parents to drive her to Nashville @ 14.I don’t know if that is true or not. But one thing I do know is that the dog and pony show story is often not the real story.Things that happen in the past (or in the present) can and are spun any number of ways to achieve the desired effect publicity and image wise.To me it doesn’t matter if it’s something that is simply repeated or if you happen watch an interview on 60 minutes where the story and words come out of their mouth directly. And are even acknowledged by family members.I know of a case (one of several) where a family member of someone who became quite well known in a particular area had told me flat out that he didn’t support what the child did in any way shape or form. But yet once the child became successful the PR machine around the child spun a quite different story of parental supervision and help and upbringing. The “fuck no way he should do that career” never came out in any stories. It didn’t fit with the image they were trying to create for that person. (In some cases it might have of course but not in this case).In the end of course it doesn’t really matter but makes for interesting stories and background color for sure.

    4. fredwilson

      taylor and U2 and the people around them need to stop thinking about iTunes and mp3s. those days are in the past. the future is SoundCloud, Spotify, YouTube, and Pandora (aka streaming). We are in a valley where the revenues from iTunes are declining and the revenues from streaming are not yet large enough to make them up. but there will be billions of people streaming music onto their phones within a few years (we may be there already with YouTube) and the manyusers*lowarpu is going to drawrf the revenues made via mp3s and iTunes. they should not get off the bandwagon before its even started it’s turn.

      1. Twain Twain

        How will Beats’ music streaming change the prospects for iTunes?

        1. Chimpwithcans

          Will Beats hardware provide any sort of incentive for the user, the artist or the record company execs?

          1. Twain Twain

            According to this Mashable article, the acquisition was about more than Beats hardware:* http://mashable.com/2014/05…Wrt music streaming services, I’m mostly interested in how they do their recommendations algorithms.Whoever manages to get that right will likely win the market.

        2. fredwilson

          Well beats should probably be on that list along with Pandora SoundCloud Spotify and YouTube. It hasn’t found a massive user base yet but with apple behind it you have to assume it will. I would say it is apple’s successor to iTunes

          1. JamesHRH

            Totally agree re: Beats role in future.

          2. Chimpwithcans

            And as a kicker, there seems to be a lot of room for improvement in the hardware/headphone department.

          3. Twain Twain

            It’ll be interesting to see how Apple adapts Beats’ contextual music recommendations (pls see graphic) with SIRI voice recognition, Apple’s gyroscope-accelerometer-healthkit context toolkits to detect if the listener is running/driving/sitting and whatever Apple will learn from IBM Watson partnership wrt Natural Language.The contextual music recommendations piece shouldn’t be under-estimated.In 2012, I took part in a Music Data hackathon wherein EMI (now Universal) gave the data scientists access to part of their 1 million interviews with music lovers.Post-hackathon, they invited me in for a meeting.I was the only person who’d tried to do something different from the standard algorithms of predicting what people will like to listen to by music genre, age group and length of recording.The contextual piece of Beats counts………

      2. William Mougayar

        we’re talking “money talks” for them, vs. “principles”. Taylor wants to make money, and she’s following the money, selfishly or not, because she can.i thought U2 was more open to it, given his Web Summit interview.

      3. Bruce Warila

        I tired of listening to people in the music industry bitching about streaming. What I have not seen is a smart lifetime value calculation on a per-song basis. People will stream a music collection for 50-75 years, and each song will have an interest curve (peaks and valleys, or one peak?). Billions of potential listeners * many, many years = more than they think…

        1. SubstrateUndertow

          I like your idea of zooming out on the value calculations. Examining the larger aggravate value to both the listener/artist, all framed around longterm “to-good-to-pass-up” consumer value. That would be a good starting place to contemplate a coherently workable pricing model.An industry-wide “to-good-to-pass-up” aggravate pricing model is what it would take to get a lot of edge consumers like myself back to the music pay window.

      4. Jon Michael Miles

        I’d venture they will get back on the bandwagon, with big fanfare, a press conference and a better financial deal.

      5. JamesHRH

        Why is both not the answer?I get the argument that Taylor should be on Spotify et al.But, there are times when my internet connection goes funky and having the content on a local device is required.Classic benefits of owning v benefits of not owning issue.We live and die on iTunes. Love it & will be footing the media bills for our kids for the next decade.

        1. JimHirshfield

          It’s both to same extent that my parents buy the occasional CD still. #transitional

        2. Marty Gottesman

          You’re right. And premium subscriptions on streaming services let you download playlists for offline listening (even YouTube as of this coming Monday). They’re limited downloads though so you can’t take them with you if you leave.

          1. JamesHRH

            That’s one of those things that poorly managed operations do.If you are a streaming service, that is what you are. Ltd downloads are a dumb idea.They are addressing a market reality in a limited way – never good.

          2. Marty Gottesman

            True, but only if it’s executed poorly. Good music services bundle streaming and offline listening into one great user experience. That’s what the market really demands. Not more fragmentation.

      6. JimHirshfield

        Spotify paid $1B to record companies/artists in the last 12 months. That’s friggin’ huge and nothing to cough at. The issue, IMHO, is the shit between artist and fan. That’s in need of some fixing.

      7. Marty Gottesman

        You’re correct that it’s still too early in streaming. Everyone complains about the payouts. But radio works because it’s 1 broadcast to millions. Streaming is 1 broadcast to 1 person. As adoption goes truly mainstream (and we get some relief on rate regulation) your formula should ring true for this industry as well.

      8. Q

        Streaming services have to get subscription rates up to $19.99/mo. (at least) with a much broader selection of artists and better interfaces (well, Spotify anyway)

      9. ErikSchwartz

        Do users really give a shit about the technical mechanics of the delivery channel? As long as they can see/hear what they want do they care if it’s stored on their local drive, or a CDNs edge server, or at some server spotify owns?The logical limit of a CDN is that everyone has everything on their own private local edge cache. Storage will get to free and infinite long before bandwidth gets to free and infinite. When storage is free there’s no point in streaming the same piece of media to an individual more than once.

        1. Mike Bestvina

          No, but they care about having a limitless music library and not running out of space on their phones in order to take selfies.

          1. ErikSchwartz

            Define limitless.My phone right now (64G) can fit about 3 months of continuous, non repeating music. In a few years a phone will have 10X that storage for the same price. The half terabyte SD card in already here (albeit expensive). A 1T phone could hold 4 years of continuous, non repeating music.Storage issues are rapidly becoming moot, data transfer issues, much less so.

      10. Johnny

        Streaming will never work for the artist. Why?1. Because music has always been a niche business and the market is very small. Remember “all you can eat” cd subcriptions?2. Record companies know this fact and negotiate the lowest rate so they don’t have to pay artists. The money is clearly on the front end via advances.3. Record companies are now tech companies leveraging catalog for ownership of Soundcloud, Spotify, etc.4. Record companies don’t care about the long term, and that’s why catalogs are being leveraged for just about anyone with $$$.

      11. Stephen Bradley

        many_music_streamers*low_rpu legacy_download_revenues has a good shot.But…many_music_streamers*low_arpu <<<<< legacy_music_industry_revenues for N=infinity.That’s (still) the problem.

      12. Matt Zagaja

        My general sense is that artists and content creators are agnostic to the mode of delivery of their content provided it gets them paid. Superstars have much more bargaining power than most and thus are able to make moves like de-listing from Spotify to try and make more money by forcing people to other venues or getting the service to pony up more. If Spotify can’t make itself a better deal then the alternative than it’s going to just lose out. Maybe Spotify’s user base will eventually be high enough to make it make sense, but Taylor isn’t trying to make money in five years, she’s released her music last week and the window to make money is now so she has to do what works best today.

      13. Dave Hendricks

        Perhaps if Artists could pixel their listeners and figure out how to market to their ‘engaged’ fans through streaming services, maybe that would fix things. But for now, Artists can’t even sell Merch to people who just listened to their tracks for peanuts.First record companies and payola decided who would hear them, now they can get huge distro but it’s all blackbox – Do they know they are big in Cleveland?I like supporting artists by seeing them play live (so do you) and buying their stuff in outmoded formats that sound better than 96kps.And I want to listen to things out of the range of wireless.What is a content/music fan to do? Are we relegating our artists to busker-dom?

  12. TanyaMonteiro

    and I was thinking how much I’ve been able to educated myself thanks to this kind of ‘free’.

  13. William Mougayar

    But you’ve got to be careful that these paid users are enough to sustain your business. If you’re not well funded while you’re trying to become sustainable, that limits what you can do, and you stay small or look like a lifestyle business..But the flip side is that if you wait too long to figure out your paid sustainability, it could become more difficult to figure it out later.

  14. Maureen Scott

    Apple’s in-app purchase proves the theory also. Few consumers buy before they try for any new/unknown game/musician/artist/writer/etc. Free to download apps; then monetise through in-app purchase is the leading business model for Apple App developers…..

    1. Dave Pinsen

      The tricky thing about this is that if you build an app, you may not know at first what a good, premium feature would be. That might only emerge after you’re up and running, and have gotten feedback from users.

      1. JimHirshfield

        SO TRUE.

        1. Maureen Scott

          We launched a beta iOS app in early 2010 with only paid content (in-app purchase) and were crucified by users. We quickly started adding free content and learned/proved that free does drive paid in-app purchases. Painful lesson.

    2. William Mougayar

      Let’s be caref not comparing different things. Apple’s in-apps purchase is a toll model on top of other app-specific models. Whether the app is crap or not, Apple gets paid. Same with ApplePay, they make money if you buy an item, or if you return it. That’s a transactional fee model.If you get lost on the freeway, and take the wrong toll entry or exit, you don’t get a refund when you find your way.

    3. Matt Zagaja

      I still think it’s strange Apple does not allow free trials.

  15. pointsnfigures

    always wonder in a straight B2B model if this holds. Or SaaS model. The other piece of advice I would give is if you are using freemium to get to a higher level of paid service, you better be good enough for most on your free service. I am invested in a company that has the free/pay/pay more model. Their basic free service is really really good. The pay service is awesome.

    1. William Mougayar

      Yup. The free that betters the experience or value is great. Like Docusign, I had to move to paid because of the features & frequency requirements. Their free level wasn’t good to me.

    2. JimHirshfield

      Quality matters, for sure.My feeling in re SaaS B2B is that the end users don’t pay. The corporation pays. End users in corp environment don’t think about price or they have much lower price sensitivity. So this can be = many users * high arpu….counter to Fred’s First Law of Internet Pricing.

      1. pointsnfigures

        At this years Avedmed conf, I was on a panel and suggested to device and drug startups that they figure out a way to give their product away for free, and monetize all the data around the product similar to what Google does with search. I didn’t advocate for it-but encouraged them to start thinking that way to see where the path might go.

        1. JimHirshfield

          Likely a tough crowd to handle that message, no?

          1. pointsnfigures

            Not when you tell them that if they are going to design a test, device or drug that is going to cost over $100 a pop, and the average monthly income in a place like India is far less than the cost of one dose. They need to think about how to monetize market share, not cost per dose. Not sure how that exactly works in medicine, but if they start thinking that way, someone will find a way.

  16. Sebastian Wain

    I think a simple part of the formal demonstration is: paid content => no social media share => less audience.

    1. JimHirshfield

      Definitely a consequence of a paywall. The leaky paywall (FT, NYTimes) handles that to some extent.

      1. Sebastian Wain

        Yes, for the leaky paywalls special case it can simply be modified to paid content => less social media share => less audience.

  17. William Mougayar

    Funny that we don’t expect print magazines to be free. Like if you print it, you can charge for it. Imagine if magazine stores charged people for walking thru the aisles or flipping thru the pages.It puzzles me when convenience stores put a sign “Don’t flip magazines if you don’t buy them.” They should say “You can flip for 5 mins, then you need to buy something, or get out of the store.”

    1. JimHirshfield

      You are mistaken, sir. The world is full of free magazines and newspapers (e.g. Metro, A.M. NYC)

      1. William Mougayar

        I know …but these are referred to as “alternate” segment. The quality is generally lower, or it’s for city guides, special issues, etc…I’m referring to the magazines you see in a respectable magazine rack at Barnes & Nobles or in a news/magazines store.

        1. JimHirshfield

          Retail…yeah, they’re just happy that people still go there. Browsers convert to buyers like the freemium model; ya gotta let some of them ride for free to get the buyers buying.

    2. LE

      It puzzles me when convenience stores put a sign “Don’t flip magazines if you don’t buy them.” They should say “You can flip for 5 mins, then you need to buy something, or get out of the store.”You’re joking, right? You don’t really think it’s “better” to phrase it that way (do you?).

      1. William Mougayar

        I was half-joking 🙂

  18. William Mougayar

    On the Internet, free content users pay with their time on free sites. It’s another form of payment.

    1. Anne Libby

      Users pay with respect and attention, too. But that currency doesn’t keep the lights on.

      1. JimHirshfield

        I just blockchained you 5 bitrespects (and 1 upvote).

        1. Anne Libby

          +annelibby is my Bitcoin user name.We’ll have to ask the good people at onename if we can transact in bitrespect. I’d send you some back. (And I’d double it when there’s humor involved.)

          1. JamesHRH

            If one of you figures out how to send bitlove thru the blockchain, umm, I will invest.

          2. Anne Libby

            Aww! Yes.

          3. JimHirshfield

            That could get kinky.

          4. JamesHRH

            investing can be that way…

          5. JimHirshfield

            Oh? I see investors more as the pimps.

          6. JamesHRH

            no comment.

          7. liban mahamed

            Thin line between promoting an pimping

        2. William Mougayar

          Jim- micro-tipping is coming. soon, you’ll be able to micro-tip with millibitcoins instead of just upvotes.

          1. JimHirshfield

            I know. I look forward to that day.

          2. William Mougayar

            day? maybe really soon….stay tuned and load-up your bitcoin wallet. i expect big tips from you 🙂 and i will tip you back for every good joke you make.

          3. Drew Meyers

            haha. lots of tips going jim’s way then.

      2. Tom Labus

        How many ad companies can the world support. Another route, please

    2. JimHirshfield

      Can you transfer me some time by blockchain?

      1. Anne Libby

        Timechain.

    3. JamesHRH

      Time saving is the next great internet hack.

    4. Jon Michael Miles

      Time is actually a form of currency. Good point.

    5. G

      And their privacy …

      1. William Mougayar

        To some extend, yes.Benefit to the NSA 😉

  19. Liban Mahamed

    Freemium model is not sustainable, also it is next to impossible to go from free to pay model,. Once people get used to free stuff hard to change habits.Right from the start make sure they pay for services. Eventually, the dust will settle and you will end up with the right number and type of customers.The paying customers.With good content it will work, look at WSJ, the content is 90% reserved for paying subscribers.Look, there is so much ad revenue to go around.Every internet company can not count on surviving through ad revenue..

    1. JimHirshfield

      As Fred pointed out in his tweets, Business Insider has more readers than the WSJ. So, it’s not just about the quality of the content.Also, you first paragraph is off-target as regards to what a Freemium model is. It is not “offer it for free and then ask everyone to pay at some later date”. It is offer a free version that is sustainable (usually thru an ad model) and offer a premium (as in add’l features) for a paid subscription. IOW, cater to two types of users: those who will pay for extra value and those who will never pay.

      1. Liban Mahamed

        I note your correction, I was describing the general revenue problem with websites that have free services.This is what I see as a problem with many internet companies; excessive reliance of ads, and wildly optimistic projections about future revenue and profits based on number of users.I am not sure no users even in the billions will ensure enterprise will become profitable. They may never become paying users. Also, they could leave and go to the next popular web site in few months.This is a problem with many prominent companies , Facebook, Twitter , and many smaller companies.They just do not have any specific, unique service that you need and can not get anywhere else.Remember AOL, 15 years ago they said it has 20 millions users and will be the dominant media, entertainment co. in the future. We know what happened since then.Twitter has 35 billion market cap and just 400 million reveune , all based on no users and hope of making money in the future.I think this type of analysis is a joke and some of these high flying companies may be more of a ponzi scheme.The valuation of many internet companies is based on fantasy such as no of users.

  20. Chimpwithcans

    Another observation – as per Spotify losing Taylor Swift – internet companies need to be crystal clear who their customers are.

    1. JimHirshfield

      Did you read the Spotify founder’s blog post? Very well written and clearly artist interests are their top priority. The post is transparent about their number of paid subscribers as well as the $2B they’ve paid artists, $1B of which has been paid in the last 12 months.

      1. Chimpwithcans

        I saw that – Taylor missed out on $6 million or something? I get the feeling the artist is a pawn in the bigger game. Very interesting times in the music industry. I never understood Apple’s acquisition of Beats, but now I think hardware/software combinations are looking like a factor for the future. Spotify is stuck in that regard without an acquisition – media/music only takes you so far i believe. But it is also a huge marketing tool to leverage off – ala iTunes.

        1. JimHirshfield

          I’m not sure I get your point. I’d be delighted if my company made so much money that the rev share we paid out was $2B.

          1. Chimpwithcans

            I guess my point is looking at what is on the horizon, and I question whether Spotify can maintain their position – I think the model is easy to replicate. If that is true they need to keep artists and record companies on board. And they need to gain market share.With Apple and Google far more ingrained in people’s everyday lives, and more than capable of copying Spotify’s model, they pose an imminent threat. Hardware will help them in this regard because it’s another lever to pull in the wooing of content owners, and in the chase for market share.That said, of course Spotify is clearly a huge success. Just postulating. 🙂

          2. JimHirshfield

            Google has definitely not succeeded at every tangential market/product they’ve gone after. They’re more likely to buy Spotify for bajillions than to see their own offering scale, IMHO.Apple? Well, what you pointed out with Beats is them acquiring to make up the loss they’ve seen in downloads (where the market has moved to streaming). Not sure Beats is the answer, but it’s a move in the right direction.Not convinced any of this has anything to do with hardware increasing user conversion to the hardware manufacturer’s music service.

          3. Chimpwithcans

            Makes a lot of sense. Sounds like I am under-rating the Spotify offering then.

          4. JimHirshfield

            I’m impressed by them. No skin in that game, but I admire how they’ve scaled.

          5. Chimpwithcans

            I am getting more and more hung up on audio quality – but that seems to be a niche factor in this game. Although these guys are onto it as a POD http://www.theguardian.com/… (again no skin in the game for me- just interested)

          6. JimHirshfield

            Quality a bandwidth issue?

          7. Chimpwithcans

            It does take more bandwidth to broadcast higher quality files, but most broadband networks more than capable of at least CD quality. But they don’t do that:

  21. Sebastian Gonzalez

    That’s probably because it’s still a hassle to pay for content and the prices are too high. If you could set up your payments once and get access to every paid content provider of the internet and any article you read costs you $0.01 wouldn’t you do it?That could be done via a plugin of your browser that automatically pays the content provider in bitcoins.

    1. JamesHRH

      Micropayments were thought to be the answer 5-7 years ago, but no one has ever made a breakthrough.

      1. Sebastian Gonzalez

        Has ever existed a solution that provides the best content with only one subscription at a reasonable price?

        1. JamesHRH

          Not an expert, but I don’t think so.Like where that idea seems to lead, Seb.

        2. JimHirshfield

          There is/was a startup that you could pay…like load up credits…and then they would deduct and payout to content providers. Or it may have been like $5 per month and depending on what you read, they paid content providers in proportion to what you read. Can’t remember the name or if they still exist. Challenges were getting users to pay for news, which en mass is generally available for free AND getting content providers to opt in to their program. HARD.

        3. Martin De Saulles

          Blendle in The Netherlands seem to be getting some traction with their micropayment system for news providers https://launch.blendle.nl/

    2. fredwilson

      Then you are at lowarpu which yields massive users and you win!!!!

      1. Sebastian Gonzalez

        You’re right, but the fact that the users are paying probably increases the revenue by an order of magnitude.

    3. Dale Allyn

      I don’t need bitcoin for this, but I have long said that I’d gladly pay for wide access to news sites (such as WSJ, FT, NYT, etc) or other content categories, if I could subscribe to one service and access the whole enchilada. As it is, I refuse to subscribe to services for info I can find elsewhere. Not forming large syndicates or co-ops is a huge fail in my opinion.

  22. Humberto

    Great topic. – I think for now we’ve seen mostly consumer focused media, which has always been freemium…: highly subsidized media paid for with heavy advertising. ,abalones, newspapers, etc.- when you get to corporate/ niche based media, the story is completely different. Customers actually get an advantage by other customers not having access. Advanced analytics, industry reports, business insights, etc have a strong value. I haven’t seen many great models emerge who compete with McKinsey Knowledge Base, the Economist or Euromoniotor, and these incumbents haven’t fully address the Online opportunity.Interesting years ahead.

  23. JamesHRH

    Every major CDN daily has some sort of paywall.My wife has given in and coughed up for a few.I am surviving without. My fave trick – catch a headline & the author, search it, find the author has posted it on Twit and back door the paper.Too funny.

      1. JamesHRH

        I find the Star to be better than CGY’s dailies. No wonder they can go it on ad rev alone.

  24. Jon Michael Miles

    While I like subscription revenue as an owner, I resent it as a user. Five bucks a month for Hootsuite is just a nuisance fee to me. Don’t get me started on Creative Cloud. That said, I enjoy incremental payments, like a powerup for a game when I’m bored on the train. Yes, it’s a cognitive issue – 1.99 is fine whereas 5.00 repetitively is just another task I have to attend to at some point. Whether its content or a service, I’d prefer the ability to cherry pick AND alternately subscribe if I’m a power user. BOTH. The idea of one model for a business seems counter intuitive to me.

    1. JimHirshfield

      Would you pay $300 one-time for a lifetime subscription to Hootsuite?

      1. Jon Michael Miles

        I’d have to love the service, which currently I don’t. In fact I just canceled my pro level. I need to reset our social campaigns so it’s a bit of spring cleaning. (Mailchimp you’re next) To answer your question: I just don’t see how a company can deliver something over a life-time. Humans inherently know that things that exist in time run out of time one day – so by definition it’s an unkept promise at the outset. I’ll love you forever – until I die. It also has the air of desperation, a one-time cash grab. So as a human I don’t see a life-time subscription to a cloud based internet service as something I’d do.

        1. JimHirshfield

          I agree with most of where you’re coming from…but not the desperation point. I think companies can offer a monthly, yearly, and lifetime price with increasing discounts and not look desperate.

  25. ShanaC

    Third observation free or paid is average amount over all content:on an individual page that logic may not be true

  26. leigh

    Toronto Star (newspaper) is getting rid of it’s paywall. Which i thought was probably a good idea since you only had to open an incognito window to get around it anyhow. It’s in the DNA of the network to go around obstructions!

    1. JimHirshfield

      Do non-webworkers know what incognito mode is? 😉

  27. dan_malven

    I don’t think the right characterization is free vs. paid. Free is really just a constant cycle of trialing behavior (and you can monetize that trialing behavior through advertising…).”Trialers” may convert to paid after a month, a year, 5 years or never. And they may go from free to paid and back to free. It depends on their particular circumstances at that time (e.g. new job, new city, new girlfriend, etc.). I just did it with LinkedIn…I was free for a very long time, then I went to paid (to solve for a particular “job to be done”), then I went back to free. Same with Google Apps.I’ve done it with media too. I watched a free (well, really with a paid Netflix sub) episode of the series premier of Breaking Bad and I liked it so much that I wanted to watch more on a long plane flight so I paid for a download of Season 1 on iTunes. Again, I trialed, had a particular “job to be done” and paid for that. Never would have happened if I hadn’t been able to trial first.In keeping with physics theme, maybe some of you can come up with an appropriate Internet Law of Physics using the Heisenberg Uncertainty Principle. You never really know whether a user is bouncing from free to paid because you can’t precisely measure their place in life at any one time (and hence, what their jobs to be done are). So you build your model with that uncertainty in mind. (And thank you to the creators of Breaking Bad for reminding me about “Heisenberg” 😉

    1. JimHirshfield

      Free samples of blue crystal? Then paid subscription.

  28. ShanaC

    Interestingly the biggest freemium content companies are aol and yahoo. And investors hate them.Why?They scale, and given more leadership/better leadership probably could scale further.(Or maybe business insider needs to merge with a dmp)

    1. JimHirshfield

      I don’t think Aol and Yahoo are Freemium at all. They are free. There’s no ium there.

  29. Jason Gelman

    Is the underlying context here about VC back-able businesses? I think the pendulum has swung so far toward commoditized content sites that something like The Information is a really interesting experiment (but only as a non-VC business, no reason to raise $$ for it).

    1. fredwilson

      Maybe. But I don’t think of it that way. The Internet allows you to serve billions. I think that is what’s important about it. If you want to serve thousands or tens of thousands you can still write books

      1. William Mougayar

        I’m not sure that is entirely true yet, re: books. (Same stage as music channels evolution, old vs. new).If you’re a top author, you’d be hard pressed not to take a $1M-$5M book advance from a book publisher, and decide to publish online only, with a pay-it-forward approach. Plus, a printed book still gives you more notoriety and authority, because of the cachet that publishers still have. (i’m just stating the reality, not saying that where we are is a good thing)

        1. LE

          Plus, a printed book still gives you more notoriety and authority, because of the cachet that publishers still have.For sure that’s a generational thing.For example we both probably can’t think of a time when TV wasn’t “important”. [1] Just like people younger than us don’t remember when all telephones came from the local phone company.[1] Excellent CNN series which included a piece on television coming of age. Time was when Kennedy was shot. TV anchors were still smoking behind steelcase desks, office type desks nothing special.

          1. William Mougayar

            Yup. Twitter has replaced TV for breaking news for sure.

  30. Salt Shaker

    The value prop between free/pay has to deliver strong and meaningful points-of-differentiation to justify conversion. Streaming music services have low free/pay conversion rates cause the benefits on the pay side aren’t large enough relative to the investment for most to take the plunge. Being ad free alone doesn’t cut it. I believe the pay wall has to deliver exclusive content and other benefits (e.g. bonus tracks, exclusive merch, tix discounts, artist meet & greets, sweepstakes/giveaways) to justify the investment. AMX credit card tiering–black, platinum, green, blue–is a fairly good example of successful price/value segmentation. Membership has its privileges. Moreover, price elasticity becomes an even greater challenge to master when there’s a free alternative; in some respects it retards elasticity.

    1. Bruce Warila

      but most music fans don’t care about the other stuff, they just want to listen..

      1. Salt Shaker

        Maybe so, but I bet 20% would care and that’s a sizable rev opp. Further, all these streaming music services need to distinguish themselves from one another, to the extent they even can. Right now, they’re commodities.

        1. Bruce Warila

          Agreed, but I think the 20% are the users paying for ad-free. I believe Spotify is nearing the 20% mark? I think the differentiation occurs elsewhere, as in the service is bundled with a device or a car…

          1. Salt Shaker

            A 20% SOM on pay is way too low. That’s the problem streaming services are grappling w/ now. I was suggesting that 20% of non-pay subs would convert w/ the benefits I outlined. I think pay SOM needs to be in excess of 50%.

          2. Bruce Warila

            “A 20% SOM on pay is way too low.” I left a comment somewhere else on this post about this. I would like to see a really smart lifetime value calculation – on a per-song basis – that takes into account the number of streams that will occur over 50-75 years * billions of potential fans (in the future). If I create a hit today, how many streams can my family expect to profit from over the next 100 years (assuming royalties will extend that far)?

          3. Salt Shaker

            +100. The problem w/ many paid sub services is they’re looking for short-term payout and not assessing their biz (and their pricing) from a LTV perspective. It’s mgt pressure, or shareholders if a public company, that drives this mentality. LTV takes time to measure and requires less focus on qtr-by-qtr financials.

  31. bsoist

    I’m a believer in freemium, and I agree generally with this ..you can do paid, but you had better have a free tier first and foremost as most users will go for that.but I’m not sure we can assetmany users * low arpu >>>> few users * high arpuis true in all cases.Sheer volume is fantastic, but it comes at a cost. I’ve experienced the struggle of supporting millions of users who paid $3.99 to download an app. Since then, I’ve thought a lot about this. Supporting users properly comes at a cost.Perhaps we should substitute profit for revenue in your equation?

    1. JimHirshfield

      Millions of users on an app like a Solitaire card game is different than millions of users on an image-related app (e.g. Instagram). Infrastructure costs are hugely different. These need to be considered when pricing the product.Aside from that, I like the learnings aspect of having a huge userbase (ostensibly acquired thru a free model) to evaluate what features live and which die. Also, learning from the user base what new features they want – either thru feedback or observation of their behavior. This is huge and widely overlooked by pundits shooting down free users as just a cost overhead.

      1. bsoist

        Absolutely agree. I probably should have used investment not cost. Having real people actually use your product in the wild is very valuable. I believe in free and freemium, but I’m not sure the formula proposed today applies in all cases. Your point makes me think, though?Do you think a lot of users at a lower arpu is always better than a few users at a higher one?

        1. JimHirshfield

          Yes, (lot of users) x (lower arpu) is always better. As regards product learnings is simply an issue of sample size…the bigger the better.

          1. falicon

            Not below the ‘break even’ point…after it (assuming you have a model that can get you there)…sure.

      2. Salt Shaker

        “Learning aspects” best done in beta mode. Post hard launch it’s awfully challenging to convince people to upsell to pay, unless the offering is truly meaningful and has strong perceived value. A hard free launch can sully the waters in advance of doing really sound/strong due diligence on product offerings. Learn as you go can be problematic, although often unavoidable.

        1. JimHirshfield

          I’m not sure we’re talking apples to apples.I’m talking about learnings to better a product. Not talking about the free or paid part here. Although if I were, I’m not convinced of your point other than to say that I’d advocate launching a free and paid version at the same time (maybe that’s the crux of your point?). Introducing paid later introduces challenges for sure.

          1. Salt Shaker

            Ah, misunderstood. And yes, that was the crux of my point.

  32. Bruce Warila

    Low arpu = ad-supported right? Makes me wonder how advertisers are going to reach upper-income humans in the future? Premium models are regressive. Ads are a tax on the poor?

    1. JimHirshfield

      You don’t see Porsche ads as you browse?

      1. Bruce Warila

        just retargeting from the AARP

        1. JimHirshfield

          Upper-age bracket —> Upper income bracket?

  33. Ryan Frew

    When did we start using journalism and Buzzfeed in the same sentence?

  34. JimHirshfield

    Case in point du jour…”…the program is invite-only, and beta testers will get six months of free access followed by an invitation to sign up for a promotional price of $7.99 a month.”YouTube announces plans for a subscription music servicehttp://www.theverge.com/201…

  35. DonRyan

    Engineer Fred is my favorite Fred.

  36. thibauld

    As a side note, following on your comment: “I can’t find a good way to showcase the discussion”. Maybe you should give Storify a try: https://storify.com

  37. Adrian Palacios

    I think there’s two things missing from this equation and discussion:First, the supply side: yes we now have the technological infrastructure to serve content to a billion+ users, but IMO *creating* content does not scale. HuffPo, et. al., are giant content factories churning out mostly junk from underpaid staff/freelancers/parasitic summaries of other (premium) content and only every once a while an insanely fascinating piece of content slips through (http://www.buzzfeed.com/mik….Second, why aim for for the largest possible audience? These content factories are the McDonald’s of the internet: scientifically formulated to appeal to the largest chunk of the bell curve. So why ignore the margins to instead chase the biggest audience?

    1. JimHirshfield

      I’d say you’ve chosen an example to suit your argument…which doesn’t make it true.By your logic, Twitter sucks ’cause it’s free and has massive reach.

      1. Adrian Palacios

        First, that’s user generated content. And yes, in the majority of cases, it does suck. I’ve obsessively pruned who I follow for years, and now reap the rewards of that.And like I commented previously to William, in many cases I’d much rather go for paid (the latter option) in all of these:HuffPo/etc vs NYTimes/Financial Times/etcYoutube vs Netflix/Amazon InstantPandora vs Spotify

        1. JimHirshfield

          Understood. However the issue we’re debating (in a friendly way, min you 🙂 isn’t a function of UGC or one individual’s tastes, as far as I see it. Twitter is a huge success in terms of valuation and user adoption (despite pundits focusing on tweeters vs those tweeted at). So, in a nutshell, I’m addressing your point that things at scale are the worst in their segments (at least that’s how I read it). And I don’t think that’s always the case.That said, I appreciate your point that the paid services you favor are of a higher quality in aggregate than the free ones.

          1. Adrian Palacios

            Cool; I think I’m just trying to tease out the supply side of this equation, because it’s very, very hard, and even harder to get “quality” (whatever that unfortunately ambiguous term means) content at scale. For instance, I compulsively check NY Times throughout the day but there’s only a handful of stories that are new each day.

          2. Adrian Palacios

            Another way to word this: I think as output goes up, quality goes down. -Adrian Palacios

  38. InvestingRevealed

    There are just too many free alternatives on the internet. Even if there aren’t for a particular service, I would think that most users have the perception that there are and therefore leave to go searching for the free service. I would think that many are settling with Reuters, Bloomberg or NY Times instead of paying Wall Street Journal.

    1. JimHirshfield

      I think you’re applying the concept to a very simple set of content type–> general news.Think in terms of Game of Thrones; there’s only one (easy) place to get that content. (piracy? yeah, I know).Think Evernote Premium….features.

      1. InvestingRevealed

        I think it comes down to competition and amount of value.In the physical world, it takes time and effort for the user to use a competitive offering. On the internet, it’s one click away, which makes it far easier for users to demand things for free.Creative content, which is unique, can and should get away with charging for it.No provider wants to give stuff away for free if they don’t have to. Because there is so much competition and so easy for users to go to competitors, that is why they do. Or, it’s not viable or efficient to do so, such as charging users every time they do a search on Google.

  39. awaldstein

    I’d love to see some data on the thresholds for Freemium as model as the thresholds seem to keep going up.A million uniques is barely a model.

  40. G

    That rule is only true if your audience is large enough, you are dismissing all niche market were the many users part of the equation is simply not there.

  41. ashafrir

    Having seen this many times, in products i have been involved with that reached hundreds of millions of users – I cannot agree more. another important aspect is that having many (as opposed to few…), also allows you the possibility of coming up with new ways of monetizing, trying out new things (most of which fail…)

  42. Rob Leclerc

    many users * low arpu >>>> few users * high arpu may work in media, but not all industries. Goldman makes ~$243,000 in net income per employee vs $247k per employee at Google and $208k at Facebook. Ferrari makes $3B in revenue each year and they only sell 7,000 cars (Although more merchandise licensing fees). For Airbnb to make $200,000 in revenue they need to sell 20,000 rooms at $100/room * 10% commission. For a small investment bank they only need to do a $4M deal with a 5% transaction fee. And think of the customer service that would be involved in booking 20,000 rooms? That’s 20,000 opportunities for something to go wrong.

  43. Gregory Magarshak

    No way Fred, I don’t believe you guys coined the word Freemium ;-)A few days ago, someone hit me up on Twitter and posted screenshots of a shareware program I wrote for AOL script kiddies many years ago. Brings back memories – I was 14!http://justinakapaste.com/2…But yeah … whether I realized it back then or not, I had all the ingredients come together:1) A close-knit social network, i.e. AOL2) An free app that people would share with each other3) A way for people to pay once they liked the appI remember getting a lot of $3 envelopes, my mom stored them in a drawer and I used to take some out whenever I wanted to buy a pizza or something :)Ah the good old days…

  44. theschnaz

    When things are relatively them same (news) it’s really hard to justify paying for it, when you can get the same thing for free elsewhere.When things aren’t the same, you can get people to pay. People pay Netflix for House of Cards, HBO for Game of Thrones… you can go to YouTube and watch political and dragon videos, but it’s not the same… I can read about any topic, at great length, anywhere on the web for free.

  45. Andy Orr

    Ben Thompson shows one way to make it with paid content that confirms much of what you say here — http://bit.ly/14idVYB. Ben offers good free content but can be supported by a small number of devoted customers on the paid side (I would actually designate them supporters). This model doesn’t scale well.

  46. Dave W Baldwin

    Interesting you mention Twitter since they’re free and now working with IBM which will be something in marketing/ad.

  47. Terry J Leach

    If micro payment with an approate user experience takes root across a wide range of premium content provider, maybe just maybe there will be a middle ground between low arpu and high arpu.

  48. dounts

    i think many users * low arpu >>>> few users * high arpu works for products that are “good enough”, are for the mass & can be cheaply distributed at scale, which is true for most things online.but for very “sophisticated” demand niche, it doesn’t work, i.e. NFLX where “good content” requires a lot of capital to produce & transmit (at least so far). journalism/music is still cheaper to produce & transmit, so there are lots of articles/songs that are free/low cost and good enough (hence soundcloud, huffpost, blogs, etc.). video content will get cheaper at some point but not yet. but in some cases, artists like Taylor Swift can still pull off record album sales, but her target audience was one of those “sophisticated” demand niche of die-hard fans. and for blogs like stratechery, it can create a core following of techies who are willing to pay.Also, abundance of (cheaply produced & delivered) content (like music, journalism & blogs) leads to fragmentation of consumption (even in the TV world). in this sense, i think you can make a case that you want to either go for 1) good enough general mass product with many users * low arpu model or 2) highly sophisticated niche product with a religious following with few users * high arpu model.

  49. falicon

    That’s not an ‘internet’ law, it’s more likely an ‘internet startup’ law…in other words, if you *aren’t* trying to be an internet startup the law doesn’t apply (for example if you are happy to build a lifestyle business few users * high arpu is in fact the VERY thing you are targeting)

  50. Peter

    Does that mean that the theory of relativity will be an ubiquitous payment method for pay per view on the internet. Which would then turn this theory upside down? 🙂

  51. Nik Bonaddio

    ESPN Insider is pretty much the archetype of how to do it right. It mints money for them.

    1. falicon

      Speaking of which – I gotta connect with you on something (will email you in a few) 😉

  52. Steven Kane

    Ah. So on the interwebs, you *can* make it up in volume! 🙂

  53. Semil Shah

    In general, I agree with you — but for startup-related content vis a vis what the big tech co’s are doing, I am a full “Information” convert and happy customer. It is worth it.

    1. Mystic Artist

      I am no internet academic, but am an inventor and exquisitely frustrated to see my ideas over decades come on market by others. How does one find an investor ?

    2. Nancy Ruth

      what do you think about the fergerson case

  54. Stephen Bradley

    “I can’t find a good way to showcase the discussion…”Exactly. I do hope that I’ll someday coax beyond the religious debate about what should or should not live inside Twitter long enough to see how easy and wonderful it is to capture a community “story,” for both live in-context debate AND posterity, in AuthorBee.

  55. Pete Griffiths

    Hence the freemium law:there needs to be a clear bright line that distinguishes that which may be readily characterized as free content vs that which users will understand need (justifiably) be paid.Any mushiness here is a problem

  56. SwitcHHat

    I can liken this to a business I built from 1998 to 2004, in which our company was one of the first & biggest DME (durable medical equipment) suppliers in the nation, with a 100% paid business model. Customers were happy to “purchase” their products from us, and we delivered supreme layers of customer service and diversification of product…unmatched in our industry. And then it happened.New fly-by-night suppliers who were looking to “cash in” on the DME biz, came out in droves in 2004, and started offering the “get it here for free…and we’ll bill Medicare” machine. They became known as “Medicare Whores”, just looking to bilk patients & their Medicare numbers for quick cash. And of course, the industry changed for worse!No longer were companies like mine able to lead with keystone ethics, like, excellent customer service, highly unique & diverse product lines “tailored” for each patient, free house calls, and basically treating customers like a “friend of the family” who’d you go above and beyond for. No, it all became nothing more than dollars and cents, and the customer quickly became a simple commodity. In fact, this is how most companies in the DME field still operate to this very day.My point is this: I am an enterprise level full stack developer for native mobile software, and have been since 2007. Early developers (and I mean the truly committed ones) in the mobile space, really took great pride to deliver a “paid” level of enterprise software (apps), and customers were excited, received great value, and were happy to pay the very nominal fee we charged ($.99 to $4.99ish for apps that took many $1000’s to build). Great business model! Reliable scalability! And NO GAMES being played on customers, in a quasi effort to soak their wallets on the back end. But like so many businesses go, there are those who choose to go the other way, and play another game. Right or wrong, clever or not, one thing is for sure…..all it takes is just a few people to start “whoring themselves out” for cash, and you can say bye-bye to the arms length & intrinsic value associated with a straight forward sale!And this is my short answer.

  57. sigmaalgebra

    There’s the old remark, with only a little truth, that There are only two ways to make money, bundling and unbundling. Extra credit for knowing the (usually credited) source!Well, agreeing essentially fully with Fred (gee, has to be a first time for everything, maybe!), what if? I mean, suppose there was some bundling? Or, maybe, in violation of the spirit of network neutrality, suppose an ISP offered several package deals, the usual ISP service but with, way, Amazon Prime, the new paid HD YouTube, Netfilix/Vimeo, maybe the Knicks, Lions, Tigers, T-Rex or whatever sports team, pro tennis/golf, or exercise video with young, pretty, flexible, energetic girls who don’t need to lose weight, etc.?Maybe some viewers would want an HD yoga channel where the girls are wearing the popular, new yoga pants that don’t need panties! Maybe! Ah, a special case of unbundling!Then the user need just pay the ISP’s bill each month with just the one check for the whole package. And, sure, for the HD video options, upgrade the download bandwidth.So, the attractiveness:(1) The Internet, TV, phone, mobile bill is already ballpark $100, 200, 300 dollars a month so that the extra for a favorite ‘content package’ can look small in comparison. Or, once have paid for all the means of content delivery, why not just go for a little more and also get the content actually want?(2) Don’t have to worry about making separate deals, payments, mud wrestling with customer service, automatic credit card billing tough to stop, etc. for each little source or instance of content and, instead, just solve the whole problem with just one click once. Then write the one check once, or enable just one case of automatic credit card billing.(3) Get just one customer service desk to gripe to.Maybe add a ‘social’ enhancement — if sign up for the platinum NY area sports team HD video package, with the coming features of user real time camera selection, after game slo-mo replays, expert commentary on each pick and roll, give and go, etc., then get an invitation to exclusive birds of a feather after game elbow bending meet-up at a selected venue, right, with a big display screen, a bar, etc.?And, get into the Ballmer Clippers idea of users Tweeting or some such during the game to put their name and picture up on the big, high above center court display screen (HN has the thread with the YouTube URL of his lecture at Harvard’s CS50).And dream up such other benefits, of course, a sweatshirt or polo shirt with coveted logo. And a drawing for a seat in the owner’s box. On and on.’Bundling’!See, this time I didn’t mention E. Fromm! Not so fast! One of the main solutions to the fundamental problem in life is “membership in a group”. So, that’s in part what we’re talking here! Oh, and the best package subscribers get to join the ISP’s ’round table’ on what good packages should consist of!Some such things might be as good as joining the Long Island/Hamptons Ferrari owners club; if there isn’t one, then there should be! By the way, women, if they can pass the application process, get admitted for free, even if they don’t know how a dual disk clutch works or what is good about Brembo brakes or a Ferrari flat crankshaft!PS: Gee, wore out my keyboard! The space bar wasn’t working!So, turn the keyboard over and remove a lot of little screws, some plastic, more screws, some metal and some wires, a complicated sheet of soft plastic, a sheet of plastic with a printed circuit, and discover that that was the wrong stuff and reverse the work!Get out a standard tooth probe tool, that is, with a handle and a little piece of stainless steel wire with a bend and a sharp point and lift up the keys Alt on either side of the space bar. Then lift up the space bar, unhook the wire anti-roll bar (works much the same as in a car and keeps the space bar horizontal as it is depressed), and take a look.Remove a lot of human hair, cat hair, fingernail clippings, and bread crumbs — no sign of insects, whew!Yup, a piece of plastic in the anti-roll bar mechanism was broken. I’d hit the space bar too hard too many times too close to the right end. So, right, the broken piece was on the left end and was broken from too much torque.So, good news: In a box less than 3 feet away, there was a package of epoxy. And it was clear, which is nice. And it was 5 minute epoxy, which for this job is terrific!So, mix up some epoxy, get some tweezers, dip the broken end of the little plastic piece in the epoxy, and position the broken piece where it belongs as the epoxy sets up. Then let the epoxy harden — they weren’t much kidding about the 5 minutes! So, mix up some more epoxy and apply a second reinforcing coat!Now the thing should be stronger than the original!Then with the tooth probe, go between the other keys. Gads! The loose balls of hair, crumbs, etc., about 1/4 cup each, kept coming out from between the keys!Blow out the area with a can of compressed air.Carefully put the keys back.Let the epoxy get harden overnight. Try it out.It works!Saved the keyboard!Do they teach that in Harvard’s CS50!That’s not the first time I’ve broken the space bar mechanism on a keyboard. So, if you ever have this problem, then there is some hope for a fix!

  58. Elia Freedman

    I’ve been thinking about this post all day. With all due respect, I think there are so many ifs attached to this “law” to make it moot. I don’t think its a law at all. If you are a funded business and in a certain type of business then this makes a lot of sense. Otherwise it’s useless and a distraction to attempt to balance this equation.

  59. Michael Chou

    Here’s my 5th grade calculation:Assuming a content site implements a paid model. It has 100 paid customer x $5/customer = $500 (per month)Once it converts to the a model, let’s assume it reaches 100x more people (or, a 1% conversion rate). Because 10,000 x 0.05/person = $500, the site makes equal if it monetizes at 5 cents per free user, presumably and predominantly via ads.Fred’s law holds true if:1. the difference of revenue generated by paid and free users is smaller (ex. $3 and $1), or2. the difference of users attracted by paid and free model are larger (ex. 10 and 1M)And the law breaks if none of the above applies. But, critically, all the factors above affect each other.For example, the law assumes quality content. Because only quality content can attract paid users. This is Mr. Godin’s point. So another way of re-stating the law is “only quality content has the option of choosing paid or free model.”Second, content with higher quality drives more paid users and (exponentially?) more free users. The difference of growth rate is important.Lastly, local restraints and market competition will affect the calculation. For example, English content perhaps has the largest and most affluent user base. But from where I come from, Traditional Chinese is used by at most 30M people. And the average CRM is 15 to 30 cents. That is per thousand impression. These restraints limit the upside of the free user base.

  60. Demian Brener

    Good analysis as usual, Fred. Although ARPU decreases as the number of users increases (and thus increases the denominator), revenue is ultimately a function of the number of users, be it because of network effects, economies of scale or other reasons inherent to internet nature. That’s why you have more leverage to increase revenues by increasing the number of users rather than your ARPU.

  61. Kenny Fraser

    Wonder what the equivalent law will be for enterprise software? The day to day challenge here is the free trial model followed by a struggle to convert paying users. Can we figure out the “physics” for this?

  62. kev polonski

    Dropbox would be a counter-example: a few high arpu with lots of free riders.

  63. Chris Chaten

    Would Bloomberg’s model be the exception to this rule?

  64. jj

    I think there is room to be skeptical about your theory being broadly true. From the perspective of the entrepreneur, it is comparatively easy to build relationships with businesses and become ingrained in their workflow (easy compared to building/selling usefulness to consumers). If you are moderately successful at charging $30/month per user (a tiny amount to a business, but huge arpu by internet standards) and get X,000 users, then you have built a $3M (or $5M or $8M) company. Many of us, given the choice between a 100% chance of finding a “small” customer base and building a $2M company or a 0.1% chance of building a $2B rocketship company are perfectly content to dismiss a theory that huge#of_users*small_arpu>>>>>actual business as only being true from a VC’s perspective.

  65. Alex Bargmann

    Bit of further reading – this article on Ben Thompson of Stratechery from Gigaom goes against this law:https://gigaom.com/2014/11/

  66. OnlyfortheSEO

    We are paying Google constantly to keep our webpages in the first page. I dont see how it is any different to paying for content. The people giving the service needs to be paid. I struggled with the same questions in my head as I started my first website a few months back at http://renwickplatinum.co.zaEventually I realised that the whole internet is just one big market for people that want to make money. Its a tool that millions of us use to get what we want.Whether you use a try before you buy option or subsrcibe to someones blog, the motive is ALWAYS the same and it is called MONEY..

  67. Seth Godin

    As in, “we need you to do this because we need to make a living,” as opposed to, “you want to do this so we built it.”(Good point Twain, btw)

  68. JamesHRH

    Only one of the drivers of what people can think is worth doing.Self righteous?

  69. William Mougayar

    Or label it law of internet “content”? Good point on not mixing content & services, like apples & oranges.

  70. JimHirshfield

    You got free broadband?

  71. JamesHRH

    Hung by own petard, Fred risks, says Yoda of Penn.

  72. leigh

    Keeping with math and science metaphor– Results = Perceived Value – Expectations

  73. JamesHRH

    Youch!Someone got up on the wrong side of the community garden today.Lots of human drivers out there, some noble & some not.Posts are allowed to be works in progress, IMO, when you are cranking out 364 a year for 8 years.

  74. JimHirshfield

    Let’s use baking analogies instead, OK?

  75. JimHirshfield

    I don’t believe that. Artists are taxed by the middle men. Did you read the Spotify founder’s post the other day? Big bucks are flowing out of Spotify. I don’t think they’re the bad guys, if there are bad guys.

  76. LE

    If anyone is underpaying artists it’s for the same reason that Airline Pilots are unpaid and why doctors are paid well (or even in some cases overpaid). Supply and demand.Don’t blame the business people for taking advantage of the fact that there is a great deal of people who are willing to work for peanuts and accept the terms that are given to artists.Right now you are probably the only [1] bread made in Lancaster that is trying to get into supermarkets. (But not the only bread everywhere obviously). If you had 30 other companies in Lancaster making bread knocking on doors what do you think that would do to the terms that you were able to get and money you could make by selling your fine bread? And if one of your competitors or several were willing to lose money on their bread and assuming it was of equal quality to your bread (or even actually not of equal quality) what happens to your business model?Do you think the Whole Foods buyer as a business person is going to just keep the price high for the benefit of you when there are 29 others to choose from?[1] Nearly the only?

  77. JimHirshfield

    Read it. Opinion not changed. I’m not insensitive to the artists. I just think that you can not compare song purchases, especially album purchases to single song streams.”Musician Zoë Keating earned $808 from 201,412 Spotify streams…”That’s like one play of one song on a radio station in Montgomery, Alabama. Not bad, IMO.”Keating earned a mere $13.38 from 387 plays of these songs on Microsoft’s XboxMusic service…”Seems comparable to what she might make busking in Grand Central for 15 minutes…altho in 15 minutes in GCT 5000 people might hear her music.Does she deserve to earn a fair living making music? For sure if there’s an appreciative audience. The more the audience likes the music, the more the artist makes – that’s a tough reality for the lion’s share of artists. Which makes it ridiculous to hear Taylor Swift’s people cry that she only made $500k on Spotify domestically in the last 12 months (or $2M worldwide, if you believe Spotify’s spokesperson).

  78. LE

    because they’re underpaying the artists.Seems that “they’re” is being called “bad” in the above sentence.Would you agree that you have a bias against certain parts of the music business?

  79. Neil

    Charlie Crystle RE: “this is worth a read and might inform your opinion. http://www.theguardian.com/tec…Keating is certainly very saavy…. “Streaming is not yet a replacement for digital sales, and to conflate the two is a mistake. I do not see streaming as a threat to my income, just like I’ve never regarded file-sharing as a threat but as a convenient way to hear music.”I wish I could make this demand: stream my music, but in exchange give me my listener data. But the law doesn’t give me that power. The law only demands I be paid in money, which at this point in my career is not as valuable as information. I’d rather be paid in data.”

  80. JimHirshfield

    Get ye loaf to the big city and we’ll break bread.