Freemium is Not a Business Model

During dot-com (v.2) when it didn’t matter if a startup actually had a business model, a concept that had a lot of street cred was “freemium”.

Coined by Jarid Lukin, and propagated by Fred Wilson, freemium was seed as a solid business model because companies charged for advanced features even though they offered basic services for free. The model was based on the notion that if you gave consumer a taste, they’d happily pay for more.

The problem is freemium doesn’t work for the vast majority of companies, especially ones focused on the consumer market. In theory, it sounds good but in practice few people actually pay for more features. Sure, there are exceptions to the rule but those are few and far between.

Freemium is flawed because most people don’t need more features than what they can use for free. Poll Daddy, which was acquired by Automattic earlier this week, is a great example. If you’re willing to live without customer support and a few small wrinkles, you can create as many polls and surveys as you like. If you’re want the features and more survey responses (>1000/month), there are two premiums packages: $200 and $899/year. For most people, a premium package is unnecessary.

Freemium is another example of how people seemed to forget business fundamentals during the wackiness of Web 2.0. At the time, it seemed like a great idea but freemium probably ranks along with “eyeballs” as yet another flavor of online Kool-Aid that everyone happily drank without thinking too much about it.

If there is a foothold for freemium, it’s the corporate market where customers are willing to pay for high-quality services as well as customer support. This is why companies such as 37Signals and Freshbooks have thrived because enough customers want more than freemium services.

For consumer-focused companies, however, freemium is fool’s gold…and, most important, it’s not a business model to create a viable and vibrant company.

More: Check out Fred Wilson’s thoughts on why freemium is his “favorite business model”.

As well, Steve Hodson (aka “the cranky old fart”) had a post a few months ago asking “Is Freemium the Way to Go?” with this conclusion: “As idealistic as the freemium model may appear it may not be the best answer for any of the parties involved – whether they be the start-up, the advertisers or the users.”

As well: For entrepreneurs and investors, a good read is Paul Graham’s “Why to Start a Startup in a Bad Economy”.

Technorati Tags: , ,

This entry was posted in Web 2.0. Bookmark the permalink.
  • http://KevinDykes Kevin Dykes

    I disagree with you Mark – I think this is a broad generalization. Just because it hasn’t been successful by all who tried doesn’t mean it is not a viable model. Freemium can be powerful and is a proven business model for many web app companies. But, like any successful approach, it has been “me too’d” by companies who didn’t pay attention to the primary requirements.

    A proper freemium model needs to

    1. Allow access to full, core features but with some easily reached constraints that makes the user want to make the jump to paying customer. It is simply an applied buying funnel of today’s marketing world.

    2. It needs to allow the user to get to the ‘can’t live without’ stage – building value and embedding itself into the user’s life/business.

    3. It must start out with a focus on illustrating value for the paid version and that this is the goal. Like any marketing funnel or sales pipeline, it needs to focus on somewhat qualified prospects to enter the free end of the funnel.

    Just because someone creates a new web 2 app doesn’t mean freemium is the right approach. It takes careful definition of what is allowed for free and what is the hurdle to pay to be successful.

  • Taylor Davidson

    Also disagree: the freemium model is solid: our errors are in how we tactically apply the model.

  • Mark Evans


    You make an excellent point. One of the reasons the freemium model can fail is not getting the right balance between free and paid. In other words, you give away too much AND while the premium model is seen as too expensive to migrate to the upgrade.


  • Tom Summit

    How about 37Signals? seems to work for them.

  • Grok2

    I think your theory is flawed or atleast too general. Polldaddy is a bad example because a poll is such a small part of any business that anyone uses it is not likely to want to pay for it for throwaway polls. For such users, having the free services to give a taste of the service is essential — they are the ones likely to convert to the paid users when and if their needs rise. So the base business model of polldaddy was flawed since you can’t run a reasonable business on online polling systems alone.

    I think the fremium model works when you give the right combination of your feature set as free to give a taste of what is available and market to those same users continually about the expanded feature set.

  • Mark Evans



    As I mentioned in the post, freemium works for 37Signals because it’s selling a valuable service to business customers willing to pay for additional features and customer service.

  • Pingback: Freemium Isn’t A Business Model, It’s A Marketing And Trust Strategy » What’s In Peter’s Head

  • Peter

    I wrote a longish semi-rebuttal/semi-agreement here:

    I definitely think you’re more right than wrong about this!

  • Martyr2

    I agree with Kevin and many of the other posters. It is a model, the balance is just not quite right in many of the situations. I think the problem is that these sites think that once people hit their site and try it that they are going to want massive amounts of the service. “If they setup a survey on my site and like it, then they will create 100 of them easy! So lets give them 10 free” The problem fails because even though you like the service, making more than 10 surveys would not fit the habits of 99% of the public.

    Now if they set it so that 1 survey is free, then perhaps someone will pay extra for the additional 4 or 5 they may make later.

    This also brings up a point about price point for services too. I see a lot of them saying “Setup more than 10 surveys for 5.95 a month!” They probably should say “Setup more than 1 survey for 99 cents each a month”. That way instead of getting 100 people all paying 5.95 a month, you could get 100 people who are mixed with some only using 4 surveys for 99 cents ($4/month) but also get those who are using 10 surveys ($10/month). I think when you add it up in the end, you will make more money and also provide the people the power to set their own price.

    But I am sure you know more about marketing than myself. All I know from a tech guy is that if I setup accounts with 5 different sites that don’t have this concept down yet and I can get all the services I want for nothing. An account over at this one site where I can get up to 10 surveys free, another site where I can download 10 free graphics a month, another site that generates HTML for nothing, and another site that allows me to do e-commerce transactions for free if I stay under 1000 per month. Now I have everything to run my own site!

  • dave mcclure

    absolutely disagree.

    plenty of Freemium models working just fine.

    might not be the only way to go, but it’s certainly one of the more prominent ways to go.

    PayPal probably is one of the best examples of this — the early service was free, and then gradually they converted some % of audience over to paid that was accepting cc, and/or doing larger volume.

    actually the BIGGEST problem with Freemium is:
    1) poor segmentation of product features in “free” vs “premium” versions
    2) poor attention to conversion campaigns / features to convert free userbase to paid

    key to #1 is to move selected features from free to premium that don’t retard growth, but are desirable enough to pay for use.

    key to #2 is simple to pay attention to promoting paid features, and perhaps provide “trial” usage to get user to experience more desirable stuff.

    anyway, to cast such a broad stereotypically disparaging statement is both naive and incorrect.

    c’mon mark… you’re usually a good bit better than this.

  • Pingback: What’s In Your Startup? It’d Better Have Revenue, Technology Or Users | tinyCrunch

  • lawrence

    Why is PollDaddy an example of Freemium not working? They achieved exit without raising venture financing.

  • Tony Wright

    I’ll give you a pile of freemium companies that are kicking ass: Skype, PayPal, 37Signals, DropBox, RedHat, Flickr, MySQL, WordPress, etc., etc.

    Your PollDaddy example isn’t an argument against freemium, it’s an argument for carefully deciding where to place your “Pay Wall”. If you give too much away, no one buys anything. If you don’t give enough away, you lose virality and the marketing bump that active/noisy/evangelical consumers give you.

    In PollDaddy’s case, I’d wager that their free offering is so damn viral that being generous on that front spreads adoption enough to compensate for the lack of paid conversions.

    Each business varies– the free version can be a glorified free trial / crippleware or it can be pretty darn feature-rich. I think you’re right that most people are a touch too generous with their free version and don’t offer a compelling reason to upgrade. The location of the “pay wall” can make or break a freemium business.

  • Mark Evans

    Man, there seems to be an awful lot of freemium supporters out there. :) Some excellent points but I think one that resonates with me is getting the right balance between paid and premium. How do you tempt the user with something useful without giving away too much that deters them from upgrading. If you can get that right, your chances of success are improved. Of course, it also helps to have a strong product that offers value.


  • Ken

    At first I was a bit confused, but now I just think the headline is a tad misleading (or rather, a mis-generalization).

    Freemium is alive and well, if you’re doing “free for personal use, companies (generally) pay more”. It’s basically B2B with a free demo.

    So while “freemium is flawed”, “freemium doesn’t work”, or “freemium is fool’s gold” jump out, these aren’t really true. It’s simply that freemium is not a panacea.

    I suggest a better way to look at it might be: “start with MIUM”. If you have a good service people will pay for (“-MIUM”), then make a hot demo so everybody can see it (“FREE-”). But just because you can make free hot demoware, doesn’t mean you can turn that into revenue.

    Dang, now I need to come up with a B2B model…

  • Christopher Golda

    @Mark Evans: Designing a freemium business model isn’t trivial, but it’s certainly valid as many others have pointed out. It’s a balance between abundance and scarcity that has to be found through constant experimentation, iteration, etc; doing it successfully will tie in the “temptation” required. The idea here is that pricing products and services at $0.00 increases the potential to create value (and capture value).

  • David Rusenko

    I also completely disagree with you here, as a founder of a freemium based startup. We’re pulling in quite a bit of revenue (in fact, enough to support our 6 person team), and it’s working quite well for us.

    Obviously, you need to provide a compelling reason to upgrade. It sounds like PollDaddy didn’t. Second, the pricing has to be right — consumers are willing to pay, but they aren’t always willing to pay much.

    To compare freemium with the “eyeballs” business model of web 1.0 is absolutely ludicrous.

  • Mark Evans

    David: Thanks for the insight. Just out of curiosity, what’s made freemium work for you? Any advice to other startups?

  • David Rusenko

    Mark, I think the biggest thing to remember is that, at best, you will have somewhere between a 1%-10% conversion rate. At best.

    Those are some pretty dire numbers, so you have to have a fairly large audience for the model to make sense, especially when you talk about charging < $5/month. One of the ways to achieve that audience is to offer a very compelling free feature set. For us, it also helped to be completely free for a long time, to attract a large amount of users. Remember that users attract more users.

    We tried to find a set of features that was not restricting to most users, but that some users would absolutely kill for. Your price point needs to make sense: Would you sign up for something at that price? In PollDaddy’s case, it doesn’t sound like their price point was well thought out.

    From that point, you need to make the purchasing process as simple and integrated as possible — most web startups already know how to tune funnel and conversion rates, and this is an ongoing process with sales as well.

    It’s not like any of this is is magic. A freemium model will be well suited to certain startups, those who have users who are willing to pay to some extent. It will also be poorly suited to others, who just don’t find enough value in the premium product to pay for it.

    A product of little value that tries a freemium model doesn’t discredit the model itself.

  • dave mcclure

    David Rusenko, Tony Wright, Kevin Dykes: we’re all in violent agreement here. let’s go kidnap Mark Evans and beat him senseless with a wet noodle, until he concurs… then we get him to pay us $25 / month to never do it again.

    (seriously: i thank Mark for teeing up such a great opportunity for us all to tee off on him so hard! ;)

    have a good weekend mark, no hard feelings dude.

  • Vasudev Ram

    Stimulating post, ok :-)

    I agree with those who say the freemium model model is not proven to be unworkable. Previous commenters on this post have given many reasons and real-life examples of the freemium model succeeding.

    And hey – there are zillions of examples of freemium outside the world of tech startups. Go to almost any market (in the physical world, I mean), and check out all the shops and sellers with offers like “buy one, get one free” (*) and variations on the same. Of course, maybe I should call it the miumfree model since they reverse the order of free vs. buy … :)

    (*) It’s another matter that some of them artificially jack up their prices (age-old trick) and then offer the discount or the freebie – about that, all I can say is:

    Caveat emptor (buyer beware).

    - Vasudev

  • Mark Evans


    No hard feelings. :)

    Getting divergent views from people is what making blogging educational, interesting and fun.

    I’m going to follow up with a post on how to make freemium work.

  • Pingback: Wallen's

  • Lenny

    Hi Mark,

    Lenny here from I completely disagree with your post, you should have sent me a note first. PollDaddy has actually done really well from using a freemium model. We were profitable and growing at the time of the acquisition. The freemium model allows us to provide great extra features to our paying customers. Our free users help with our marketing by including a link to PollDaddy in their content. We only ever charge businesses etc. And because we can operate with zero advertising on the site it leads to a much cleaner experience for all of our users. I would urge any start ups to consider this model as it works really well in certain cases.


  • Mark Evans


    My apologies for not touching base. I’d be interested in your advice on how to make freemium work.


  • Mark MacLeod


    A courageous post that you have clearly taken some heat for. Here’s my take:

    I’ll follow up later this week with some thoughts on the pitfalls of freemium just to be fair. Though overall I like it as a business model.


  • Mark Evans


    Hey, good to see I started a vibrant conversation. Looking forward to your thoughts. I’m putting together a follow-up post that will, hopefully, go later this week.


  • Pingback: Two interesting links on venture and startups » Startable

  • Pingback: Another Stab at the Freemium Thesis | Mark Evans

  • Mark Evans

    For anyone interested, I’ve done a new post on freemium:

    Let me know what you think.


  • Pingback: ProjectX Blog » Blog Archive » Xlinks digest - 4 / 11 / 2008

  • Pingback: tinyComb » Blog Archive » What’s In Your Startup? It’d Better Have Revenue, Technology Or Users | tinyCrunch

  • Alex Osterwalder

    Mark, I’m a couple of months late to this discussion, but I would seriously challenge your argumentation and many of the follow-up comments.

    The essential variable that nobody has brought up for the freemium business model is how much the free part costs the company.

    Skype, for example, incurs relatively little costs from additional free users. It can afford to give away and live from 6%-10% paying customers (from a base of 350 million users). Customers that use Skype for free can call as much as they want, the traffic goes through the Internet, independently from Skype, and generates almost 0.- costs for Skype (not considering the development of the initial software).

    A video-service like youtube is completely different it has to pay for storage and bandwidth that free users consume. Each user generates additional costs that have to be covered by paying users. That makes it more delicate and the balance between paying and free users is challenging to establish (as many comments have pointed out).

    So at the end of the day it is always the “cost of usage component” that makes or breaks a freemium model…

  • Laraki

    Damned wrong.
    Look at how many companies were able to generate revenues with a freemium business model.
    Like any type of business model, you need to do it right. Also, I would say that freemium is not for all types of businesses.

  • Pingback: White Collar Game Blog » Debating About Freemium Model

  • Pingback: Value Activation: A Different Business Model for Internet Entrepreneurs - WebPronto

  • Pingback: Value Activation: A Different Business Model for Internet Entrepreneurs - InfoLogs