Reframing the problems with “Freemium” by charging the marketing department

Seems like every third startup nowadays is using the “Freemium” business model: The lowest service tier is free, and the business is designed to get those users hooked and then upgrade to a paid plan.

free breakthrough

It can work wonderfully of course, but usually it crushes and destroys companies, not only because it costs more than anticipated but because the founders didn’t realize the business model itself caused them to make incorrect decisions.

Foibles of Freemium

Let’s dispose of some misconceptions about what Freemium actually does for you and how much it costs.

Freemium is not customer development

Just imagine how much you’ll learn once you have 1,000 real, active users of the system: Everything from behavioral statistics (which features are actually used?) to democratic product development (voting on which features customers would like to see next?).

Trouble is, those freemium users are not like those who will actually give you money. Frequently the features that paying customers want don’t show up on the freebie’s radar.

Think about it: Almost none of the freeloaders will convert to even the smallest tier. If you suddenly started charging only $1/mo for your service, most wouldn’t come aboard. Why is that? Because the need, the interest-level, the value to the end user isn’t pressing enough for even a pittance. But those who will pay $10/mo or $100/mo do have needs, and it’s not just a matter of scale (i.e. because they require more resources), it’s a difference in kind.

Your problem is that freemium users outnumber your paying customers 100-to-1, so in feedback forums they drown out the voices of those who actually matter.

Freemium conversion rates makes marketing expensive

A really good conversion rate for free-to-paid is 4%, like Dropbox. Awesome for them, but normal rates are more like 1%, and that’s if users are reasonably active.

I surveyed a dozen small startups who don’t use freemium, and on average they see a 1% conversion rate from web traffic to a (real, not “free”) purchase. [UPDATE: Andy Brice has much more data which supports this simple assumption] Even assuming you can get a higher website-to-signup rate for a freemium offering (you’d better, right?), you only get paid on on a few percent of those, which means your total conversion rate of web visitors to actual money is 20-100x worse than other startups.

Which means it costs 20-100x as much on marketing campaigns to achieve the equivalent revenue.

This essentially takes paid-advertising and almost all other forms of marketing off the table for driving growth in a freemium business, unless you’re willing to take big losses to get things rolling.

It also means you essentially have to build a viral product, because you can’t afford advertising. Getting true viral behavior is very hard — again most companies who attempt this will fail — and even so you need to seed it at the beginning, so you still have the marketing expense problem.

The few companies who were clear winners here also raised millions in funding, in part to get over this hump.

Freemium tech support is expensive

It’s easy to say “We’ll just direct everyone to forums,” but when people email tech support they want a response.

It’s easy to say “We won’t provide tech support for the free tier — they’ll understand since it’s free,” but if you really do ignore them they’ll be less successful with your tool, which means far less chance of them converting, and less change they’ll evangelize to friends and coworkers.

It’s easy to say “We’ll provide a lesser grade of support for the free tier,” but that means every tech support email and chat session and phone call has to be coupled to an account, so you’re wasting time figuring out what level of support this person “deserves.”

Are you prepared for people who say, “If you help me through this, then if it works I’ll pay.” Are you hard enough to shut the door in their face, even knowing that in fact they probably still won’t pay?

If you care about good support — one of the few true competitive advantages a small startup can have — can you really segment who gets treated well and who gets the cold shoulder? Should you really turn your back on the benefits great tech support brings? Is that conducive to converting free accounts to paid accounts? Is it helping your company’s reputation?

Advantages of Freemium

Obviously freemium also has important benefits which cannot be denied:

  • Easy to upsell. They’re already using your tool, so whether it’s by special offer, changing the pay scale, or the user just outgrows the confines of their service tier, there’s many ways a person can start giving you money. That’s something few other business models can boast.
  • Stats for selling. It’s awesome homepage marketing to be able to say “Join 1,000 other happy users.” It’s social proof, just like the RSS counter in the corner of the blog implicitly saying “if 40,000 others think it’s worth their time to read this every week, maybe it’s worth yours too.” It’s also useful when bagging bigger customers because it proves your system can scale, both in technology and in training new users with minimum effort.
  • Easy to start. Even a “30-day free trial” or “money-back guarantee” is a much bigger barrier than “free.” Getting a web visitor to stop perusing and start using the product is a critical step in any customer acquisition, and you’ve just diminished the barrier as much as possible.
  • Not using the competition. Well, they might still be, but at least you have a horse in the race. One more user for you is one fewer user for them.

So in the face of the positives and negatives, how do you decide whether it’s right for you and, if it is, how do you think about it so that you’re reaping the benefits while mitigating the costs?

Charge Freemium to the marketing department

A pattern emerges from those “advantages” bullets: It’s all marketing. It’s lead-gen, reducing barriers to conversion, and competitive advantage.

Retool your expectations of Freemium: It’s a marketing cost. It’s more expensive than you give it credit for, but it could very well be the best marketing strategy available.

So how do you decide whether those costs are worth the benefits? My technique is to “charge the marketing department” exactly like AdWords or any other lead-gen campaign: measuring the total cost of acquiring new paying customers.

The reasoning is: You have a theory that by spending the money to support these freeloaders, you’re in fact building an efficient path to real, paying customers. That goal — revenue — could possibly be attained other ways: AdWords, blogging, or any other marketing technique. So just like any marketing campaign, the marketing department should pay, measure the results, and compare the ROI against other methods.

(And make sure the cost is much less than the total revenue — the equation that most startups fail to achieve, exactly because they don’t honestly consider the total cost to acquire.)

How do you decide how much to charge the marketing department? Suppose you really were charging those customers, but also that you’re only looking to recoup costs and not make a profit. Amortize those costs — servers and tech support — and come up with a correct cost-replacement price for that tier.

That’s what those users should be paying (minus profit), so that’s the amount the marketing department needs to “reimburse” the rest of the company.

Obviously you shouldn’t get carried away with the bookkeeping. But you have no excuse to not be measuring that per-user cost so that you’re running this Freemium program with full knowledge of the cost and comparing it to other forms of marketing.

Further Reading

More great articles on the subject:

  • Andy Singleton on how to “make those cheapskates pay.”
  • Andrew Chen on the precise financial model of freemium (spreadsheet included).
  • Eric Ries on three strategies for “free” that still lead to making money.


  • Deacon Bradley

    Love this article Jason, and love the proposed solution of accounting for all those hidden costs. One thing I love about the Lean Startup community is their emphasis on charging from the start. It’s easier to let yourself off the hook just to get some users in the door, but it’s the wrong users! Thanks for the reminder ;).

  • Steven Dupree

    Jason, good take here but you’re mixing two very different models: the prevalent “feature-based” freemium (where upsell is based on solving at least two customer pain points) with the stronger “capacity-based” freemium (where upsell is based on surpassing a usage threshold e.g. Dropbox). I wrote a significant article on this last week and encourage you to check it out and circulate w/ your readers.

    I also believe it’s overly simplistic to focus on the conversion rate and hold everything else equal. 4% is greater than 1%, but let’s not forget that it’s the product of conversion rate and transaction size that matters. When I managed this at LogMeIn, we saw sub-1% conversion rates to a hugely valuable support tool (LogMeIn Rescue) that was relatively expensive. This was expected because the features in LogMeIn Rescue only applied to a subset of free users. The business model was highly profitable.

    Furthermore, advertising free and freemium have a much higher response rate than advertising paid products. In some cases, clickthrough rates are 10x higher. Thus the conversion rates may be lower when there is a free offering, but this is offset by a much lower CPA (acquisition cost). The net result of these offsetting effects to ROI is unclear and depends on the product.

    Last point for now, freemium is not a pure cost center. Free users (at least a proportion of them) are future customers – and keep in mind that the products which they eventually purchase may not even be on the product roadmap yet. It’s hard to know the value of a free user today. If you knew the net present value of all transactions that customers would someday purchase, then you would know precisely whether or not freemium makes sense! There are ways to get at this value but it depends on the company and the market.

    Thanks for sharing!

    • Patrick Foley

      Good read, Steven … but I suspect Jason is trying to caution people against using freemium at all, or at least not to use it in a lazy way.
      I had a conversation with Ash Maurya about this very subject … his message was simply that you get better information by STARTING with paying users – you can always add freemium later, since – as Jason notes – it’s a marketing tactic.

      I liked your article, but it addresses a different question IMO – how to do freemium right once you decide to use freemium at all.

      • Steven Dupree

        Thanks Patrick, I see your distinction. Of course: starting with pure paying users provides pristine information in return for slower growth! Free users are easier to accumulate and they become a base that you can learn from and bounce ideas off – but this may take your company further from profitability.
        We may agree that once you go freemium it’s hard to go back if customers are accustomed to receiving a product or service for free. That’s a strong argument in favor of trying pure paid first.

    • Richard Garand

      Doesn’t it make sense then to understand the full cost of acquiring a paying customer, using free versions as marketing channel? If it turns out that it is good for your business then you’ll see that clearly it beats all other marketing channels and you’ll keep doing it. If it isn’t good then you won’t waste your time.

  • Pingback: Tu ți-ai da produsul pe degeaba? | Cristian Pascu

  • Francis Moran

    One of my favourite quotes, from Shopify founder Tobi Lütke, summarizes many of the issues you raise here, Jason. “Twenty-four dollars a month is a slightly more annoying version of free,” he said in recounting how Shopify experienced swift and steady growth from real users once it shifted to a minimum monthly subscription model. The tire-kickers with their expensive server cycles and support calls all went away, and those who could see real value in the product were happy to pay a modest amount for it. Break out the hockey stick growth curve.

  • Pingback: Should you worry about your IP (Intellectual Property) and other silly questions — Startup Desi

  • Kumara S Raghavendra

    Insightful piece, Jason. Pegging the costs on the marketing department makes good sense. But in doing so, are you suggesting that going freemium is a purely marketing call?

    • Jason Cohen

      Perhaps nothing in a startup is “purely” one thing or another. Rising prices isn’t purely around profit for example — it changes your ability to spend money to acquire, which changes the marketing channels you can use, and it also segments out different customer populations which then also might demand different product and service.

      In the same way, Freemium can mean many things of course. My suggestion is that the value for things like customer analysis is not as useful as it might seem, and that its greatest function is that it’s a way of attracting and sustaining potential customers, which is also a way of describing social media and other things that are generally “marketing” functions.

      And of course that the ultimate goal is the conversion-to-paid, and thus it makes sense to compare it to your other methods of getting paid customers.

  • Pingback: The insanity of giving your service away for free - Flashissue Blog - Flashissue Blog

  • Pingback: Seriously, Startups Shouldn't Give it Away for Free | Mark Evans Tech

  • Pingback: Reframing the problems with “freemium” by charging the marketing department | VentureBeat

  • Pingback: Reframing the problems with “freemium” by charging the marketing department » Business Newsgator

  • Remington Robertson

    Freemium is best used for “social” products (not “viral,” as that’s a relationship between the magnitude and intensity of the social product) with features that you can up-sell to free users.

    Free users provide marketing by sharing; free users provide greater feedback on iterating new features, but you can segment your focus groups on premium/free user voting. You have a room with a bunch of potential customers, and you may only see the 4-6% pay, but you get to talk to the others in order to deliver more value, resist growing attrition rates and develop your company.

    Understand some of Jason’s concerns but freemium is one of the best, most predictable models out there. Has to be right for your product, I agree.

  • Pingback: Reframing the problems with “freemium” by charging the marketing department | Dumper

  • Pingback: Reframing the problems with “freemium” by charging the marketing department « Music RSS

  • Pingback: Reframing the problems with “freemium” by charging the marketing department |

  • Andy Brice

    >I surveyed a dozen small startups who don’t use freemium, and on average they see a 1% conversion rate from web traffic to a (real, not “free”) purchase

    So did I and got the same result:

    >It’s a marking cost

    I assume you meant ‘marketing cost’?

    • Jason Cohen

      That’s awesome data, thanks for sharing. I fixed the typo in the article, and I assume you don’t mind that I also added a link over to your data.

      • Andy Brice

        Thanks for the link!

  • Pingback: Software Marketing Tweetables - 15 April 2013 | Smart Software Marketing

  • Gregor –

    To me freemium spurs the same amount of loyalty in users as Groupon does for those just those who move from place to place looking for a bargain. They have zero loyalty.

    You can’t build a sustainable business off the back of zero loyalty. If you can get people to pay and come back there is much more of a relationship. This gives you more insightful feedback, which you can then use to improve your service or product.

    I can understand at the very beginning you need some users and some feedback, but you have got to figure out what people will pay for and what they will not. Only by charging can you figure this out.

  • Pingback: About Products – A Few Things I Think I Think | A Random Jog

  • Pingback: Reframing the Problems with “Freemium” by Charging the Marketing Department CapitalFactory

  • Carlos Tico

    Really enjoyed the article: well done Jason!

    As for our business experience, I must say freemium is working quite well for us at in essence, a simple, yet strong, innovative method to establish proof that the content of a particular email has been delivered to a third party at a verifiable time.

    Obviously, individuals and small companies are to be using such service just now and then, thus we concluded there weren’t real chances to get significant revenue from them. We still wanted them in though, as once they tried they become leads. We decided to offer the possibility to certify up to 50 emails a year for free, which in our opinion is a reasonably good free value offering for a reasonably high value service.

    We consider free users as marketing cost, which lets us focus on targeting businesses for sales. The service is simple enough to be used by anyone, but also simple to become part of any email-related workflow within enterprise software solutions, such as SAP.

    We’ve encountered a huge range of premium opportunities beyond free users, as well as within developers. With a freemium conversion rate above 13%, it seems we must be doing something well. In our case, it’s important no note that free licenses are for single users, when flat-rate premium licenses can be shared by as many users you like. What we’ve seen is that 89% of free users going premium share their licenses with co-workers. No question this is the kind of free users we were looking to get.

  • Pingback: Reframing the problems with “Freemium” by charging the marketing department by @ASmartBear | Startup Product

  • Pingback: Growth Hacking is B.S. Follow these Marketers instead | Luke Thomas

  • Pingback: Reframing the problems with “freemium” by charging the marketing department | BaciNews

  • Pingback: Visualizing the Interactions Between CAC, Churn and LTV by @ASmartBear

  • Pingback: Freemium to Premium and back again

  • Pingback: Visualizing the Interactions Between CAC, Churn and LTV | Crapkin News

  • Pingback: Visualizing the Interactions Between CAC, Churn and LTV | Shockwave Innovations