TWIS#10 Freemium Special with Lincoln Murphy- This Week in SaaS
Lincoln has just published a brilliant paper- The reality of Freemium in SaaS- the best look at Freemium from a SaaS perspective I’ve ever seen and we’ve got an exclusive interview for TWIS.
JP- Some people think that Freemium is a Business Model. Why do you think it isn’t?
LM- Freemium is a marketing tactic, not a business model, because you’re exchanging your software for your prospect’s attention, the same way you would by purchasing advertising; you buy attention. How can that be a business model if there’s no revenue and it’s replacing a marketing function?
If you want to give away free access to your subscription business in the hope that people will eventually pay to use the product, regardless of how you segment, then this is Freemium. Its just marketing. Any time you give something away for free, it should be in an effort to get them to buy something else. Substitute “give something away for free” with “mark down by 50%” or “buy one, get one” and you can see how this is just a marketing tactic to acquire customers. Marketing is not a business model, but supports a business model.
Just to muddy the water even more, I’d say if you monetize the “free users” in any other way than through premium conversions (ads, for instance) than you are not truly doing Freemium. You’re cheating and making money in other ways. Shame on you. Freemium, at its core, is about getting people to use a product that is free in perpetuity and try to upsell them into paying customers.
JP- Doesn’t it make sense for a company to give away software instead of buying advertising?
LM- Maybe, but only if you understand that you cannot build a business off of “free”… somewhere, somehow, someone has to give you some money. If that money comes from the free users stepping up and paying to use the system, you can say that Freemium, in its truest definition, works. If you monetize those users in any other way (for example selling access to there eyeballs to advertisers or aggregating their network effect data and selling it) then you are admitting that Freemium, as advertised, doesn’t work. If it did, you wouldn’t need ancillary revenue streams to benefit from the glut of free users.
Of course, if you didn’t attempt to benefit from the glut of free users by monetizing them in as many ways as possible, you would be making a mistake and thus, very few companies actually do “pure” Freemium. This likely contributes to the failure of whatever model they chose, by the way, as companies extend beyond their core competencies into advertising sales, for example, or ruin the user experience by running generic Google Adwords in their app.
JP- It would seem to me that we’re collectively obsessed with businesses that rely on network effects to be successful, like LinkedIn, Twitter, Facebook etc and there are few B2B software companies that rely on this the same manner. Do you think this clouds our collective judgement?
LM- Yes. Think for a second about how many companies there are that can raise hundreds of millions of dollars, have a billion dollar valuation and only have two customers. I’m talking, of course, about Twitter. Are there others? Not many, and yet, how many startups, bootstrapped ones at that, are trying to give the product away or bring it to market with no plan. Twitter can do this because they have millions of dollars which has bought them time to figure out a revenue strategy. Do you have the time and resources to take your time to figure it out? Most likely you do not.
But we are obsessed with, and pay attention to companies like Twitter because they are so rare. They are anomalies. More often than not, everything that is known about these companies is what is reported on Techcrunch or other web-focused business blogs. This is fine until startups start to use what they read there to base their business decisions. In many ways, this is like trying to learn about how to break into the entertainment business from TMZ. In the real world, companies without revenue aren’t talking IPO, but bankruptcy.
JP- On what do you base the statement “SaaS Vendors using the freemium model almost always fail?”
LM- Failure comes in many forms, and as I stated in the paper, if the goal of Freemium is to generate more paying customers, then it often fails. I don’t mean to imply that the failure of the Freemium model always leads to the failure of the business, but it has. Even more often it gets companies close to failure and even more often gets them from ever really getting started.
For those companies that have acquired a large user base, but not converted very many to paying customers, my suggestion is for them to figure out how to take advantage of that user base before they bring the company down. It might be time to let go of the notion that they will convert and look for opportunities around Network Effect data and Ecosystem for additional monetization possibilities.
JP- Can you give us examples of companies that publish their conversion rates from free to pay?
LM- First, we should remember that my paper was very clearly targeted at B2B SaaS vendors, not mass market applications. Where I spend most of my time is in helping domain experts bring products to market that are targeted at a very tight niche or specific business problem by leveraging the SaaS Business Architecture. The bigger the addressable target market, the more the “numbers game” of Freemium might land in your favor. My concern is for the niche or vertical SaaS plays that fall into the trap of thinking they are the same, distribution-wise, as mass market apps.
To answer your question, very few companies publish their conversion rates. Most of the publicly available information is anecdotal. I did cite Evernote in the paper (via a Techcrunch article), as it might have application for business customers, but it is just as much a consumer-oriented app. I tried to illustrate how their high conversion rates actually aren’t when you look at simple stats: free users to paying customers. And to get to this point they’ve had to take on $25M in funding. And this is considered a success story.
Some other resources on conversion rates are here, but they are not B2B SaaS:
- Andrew Chen has a bunch of great stuff on his blog about Freemium; I encourage everyone to check out his archives and subscribe to his feed
- Buzzgain as a nice post on Freemium Metrics
- Don Dodge on Freemium – a few years old, but good. Likely where the oft-cited “3% average conversion rate” came from.
JP- Hmmm that’s really intersting. SaaS vendors are slow to share best practice and it’s VC’s like Bessemer who have really been the best knowledge sharing. Are there any SaaS industry figures we can reliably use?
LM- All of this talk about a specific conversion rate for the “industry” is a red herring. SaaS isn’t an industry or a market, it’s a Business Architecture. The focus is on the wrong thing. Here’s the dirty little secret; it doesn’t matter what the average conversion rate for Freemium is because like most things in business, if it doesn’t apply to your market, it doesn’t matter. Who cares what the conversion rate for Evernote is if you have a time tracking solution for the building maintenance industry. When trying to find conversion rates, at least look to the same market, type of application, etc. Finding those numbers published is going to be difficult if not impossible. You might have to talk to people in the know; investors in that sector, current or former employees of your competition, etc. Buy someone a cup of coffee and see what kind of intel you can get.
Ultimately this just perpetuates the problem, though, as it doesn’t even matter what the conversion rate is in a competitive product. The only thing that matters is what your conversion rate needs to be to make this process successful, if you are doing Freemium already what the conversion rate currently is (measure everything, know the difference between users and customers, etc.), and how to bridge the gap or change course if it isn’t working.
Now, if you want careful industry analysis, Softletter, the folks behind SaaS University in Dallas next week 1/26/2010-1/28/2010 where I’m speaking on “Monetizing Multi-Tenancy,” just did a huge survey of SaaS vendors on Freemium and will be publishing their findings very soon. Stay tuned for that; it should be very telling, indeed.
JP- Rick from Softletter kindly donated a pass to TWIS readers but it would be great to see you presenting. I was wondering why did you write the Freemium paper?
LM- My goal in writing this paper is to open peoples’ eyes to the potential pitfalls and to compel companies looking to, or already employing and not succeeding with, the Freemium model to take a second (or third) look at it and make sure its the right thing for them and their market. Some of the greatest feedback on the paper has been from those companies that said “wow, we thought it was just us and our implementation of Freemium that wasn’t working.” It might be them, or not, but one thing is for certain; I got them to take a look at what they were doing. Maybe it saved their business.
By the way, I’m not an industry analyst pulling data from a database to come up with this paper (pretty obvious to some of the commenters!), I’m just giving anecdotes from my own personal experience with SaaS startups of many different sizes and stages. Your mileage may vary and if it does, please speak up! Comment on my blogs, tweet me @lincolnmurphy, whatever… I’d love to hear about Freemium success stories in B2B SaaS.
Just to restate my point, Freemium implementations in B2B SaaS have a poor track record when converting free users to premium customers. When you add in advertising revenue, or other monetization methods, you’ve shifted the focus away from Freemium so the point is moot.
JP- Why did you accuse free of being a cop out?
LM- Because I like to stir the pot a bit! If you look at the mentality of technology startup founders, especially those going out on their own for the first time, there is definitely an attitude that it is easier to pitch investors then to go out into the market and make sales. I think this should be a pretty well understood point if you’ve been around this business for any length of time. If you extrapolate that mentality out to product development and marketing, you can again see where Freemium, and the ability to not have to “make the sale” fits in nicely.
There is often the attitude by the founding engineer who believes in their product so much that if they can just lower the barrier to entry and get people to slide in, once in, the product will “sell itself.” The thought continues that if they get traction, no matter what kind, VCs will come knocking with large checks in hand and then they can hire a proper sales and marketing team to do the dirty work. Even better is that the traction will attract a bigger company to the small startup and they’ll be acquired for a princely sum. None of this is a strategy, its just hope.
JP- That’s interesting- I often see that as a product guy- Normally it’s becuase they haven’t got to product/market fit and can’t convince anyone to pay so instead of solving that problem they make it free. They never manage to cross the “penny gap”.
LM- Exactly. I think, in that scenario, rather than giving the product away for free to avoid dealing with the pain of making sales, the technical founder is better off finding a co-founder more skilled in the art of “business” to help them take the product to market and to make sales. Sure, now they have only half of the pie, but in the long run it will likely be a pie that actually gets out of the oven and makes some money!
In later stage companies that opt for Freemium, I often see this as an act of desperation and/or co-opting marketing methods from other companies without really understanding why they should do it. I certainly like to get involved before they get to this point, but we can help no matter the stage.
JP- I would add a great product person too- but I’m biased. You contend that rarely do existing companies feel the need to default to freemium- why?
LM- What a coincidence, I happen to know a great product guy. So, Phil Wainewright wrote a great piece for ZDNet where he used my paper as a jumping off point to talk about why Freemium is Bad for Business. He agreed with most of what I had to say except for one point; that existing businesses don’t default to Freemium. I think he might have missed the point a bit when he used Google as an example of an existing company that does default to Freemium in their “apps” business.
Too often people cite Google as an example of a company that does Freemium but this is wrong. Google is in the advertising business and everything they do is geared toward acquiring new eyeballs for their ad inventory; Google Apps is just another piece of their core advertising business. So, according to my point above, Google does not do “true” Freemium since they monetize with Ads. Users are just a cost of doing business in order to attract their true customers; advertisers. In fact, to get out of seeing Ads in Google Apps, you pay them money for the “premium” product, and then manually turn them off (they are on by default!)
JP- I guess people confuse the “shoot for the moon” approach of those companies where the user landgrab is more important than early monetisation versus the realities of the segment they’re in- you used the example of time management in the construction sector earlier- that’s never going to be or need the scale of a Google.
LM- Yes, Google is a bad example to use since most companies don’t have the resources, especially bootstrapped startups, to build products without any real go to market strategy and just roll them out for free and see what happens. Many have the resources but cannot risk shareholder value on wild chances. Google can; as long as they are creating new ways to deliver ads or extract usage patterns, then nothing is a waste. Even if the product ultimately flops, or the acquisition doesn’t pan out, if they served ads and collected usage patterns for a while, it was probably worth it. And in the case of Google Apps, it turns out its actually a cool way to get people in front of the Google brand all day while being productive and not just searching so it worked.
When I said “existing companies” in the paper, I didn’t have Google in mind. I was talking about legacy software vendors, small ISVs, or those not in the software business at all, like consultancies, professional services organizations, or the like. These existing companies don’t want to give stuff away for free. They understand the value they bring to their market, and the need to return share holder value. They recognize an opportunity to leverage the SaaS Business Architecture and the multiple revenue streams to generate maximum stakeholder value.
JP- Indeed- I really like your work on the 7 revenue streams which is on your blog. I was wondering why you say are free users not hot prospects?
LM- Simple; The numbers don’t lie. Publicly disclosed and anecdotal numbers, along with those obtained confidentially, indicate that overall conversion rates are quite low. It is pretty clear that when the vast majority of users never convert, even those who are active in the system, they are not hot prospects.
But I think the real question shouldn’t be “are the users hot prospects,” instead it should be “why did they become users in the first place?” Are you bringing in users that are not hot prospects? Is it something you are doing that is causing the wrong type of user to sign-up for your product? Too often any sort of activity or traction is considered progress. This isn’t exactly true.
JP- I guess you’re right- if they’re not converting then they can’t possibly be hot prospects. The only other reason to keep them is for other reasons, like network effect, advertising etc and then you’re not pure play freemium….
LM- To follow-on to my mention of Techcrunch from earlier, I’ll bring back the notion of the “Techcrunch effect” which was popular a couple of years ago. For those not familiar with it, for any startup profiled on the Techcrunch site, the first 275k users (the number of subscribers to their feed at the time) didn’t matter because they were just there to try out the site or app. Startups were either going to investors or the industry and saying “look at us, a quarter million users in the first day!” It was when people called shenanigans on these numbers that the TC effect was coined and the first 275k users were deemed unimportant because they would likely never really use the system or become paying customers. Fast forward to today and the TC effect has been dampened with the popular invite system, but the effect is still being felt in other ways. Are you creating your own TC effect by bringing in large amounts of users that will never pay to use the system and mistaking that for progress?
Another possibility is that the users actually were hot prospects when they came in and the free product doesn’t work as advertised so they either stop using it or put up with it but would never pay a dime for it. If a poorly executed free product is any indication of what the premium product is like, few will convert.
JP- Indeed- the penny gap as mentioned before. That’s why product market fit is so important. This might seem obvious now, but why should free users not be considered customers?
LM- Demian Entrekin already took issue with this on his blog in a post titled “Users or Abusers“.
He took issue with my attempt to differentiate between those people using the product for free and those paying money; users vs. customers. Its funny because the only reason I broke it out like I did was some of the reviewers wanted me to explore it further. As Demian said, its not very clear. I get that. Its difficult because in software we’ve always had “users” and we’ve always had customers, but for the most part, they were one in the same. The user of an ERP system was still generating revenue for the software company even if the customer was the Enterprise. In other words, the ERP company might have charged the Enterprise customer for each user of the system. In SaaS, and in particular with the Freemium model, we have the notion of non-revenue generating users. They aren’t customers. Period.
JP- Ah, OK- I get it now. Customers in pure play freemium are only users who pay you money- Users who don’t pay aren’t customers.
LM- To adequately understand the impact of your Freemium strategy, you must differentiate between those using your product for free and those paying for access. For Freemium to be successful, or for a SaaS vendor to know when its time to change course, requires management and monitoring of critical metrics. These metrics can be skewed when the vendor doesn’t really understand the difference between the the users of the system, application customers, or even stand-alone customers (network effect data, advertisers, etc.) and how all of this comes into play.
JP- Yes- sometimes people call them “Vanity Metrics”- metrics that make you feel good but don’t help your business at all. A bit like you’re TechCrunch example above.
LM- That said, I don’t care how you refer to the people that consume your resources, just know that those who pay are different from those who don’t and that there is a cost for users who do not generate revenue. I’m just glad, like the Freemium topic overall, this is being discussed and I thank Demian for talking about this topic; even if he didn’t like the title of the paper or the tone!
JP- I can understand that- I sympathise with his views but what I’ve come to realise I’ve never advocated Freemium in it’s purest sense, only as a hybrid and as a marketing tactic. Perhaps I didn’t realise that, but I do now. Why did you call freemium a virus?
LM- As Josh Kopelman of First Round Capital says in a great blog post of his, many startup marketing plans call for a product “going viral” for user growth.
Quite often for startups looking at Freemium, you will often find a large amount of reliance on “going viral” to get to market. The reality is, no matter how well you build in the viral loops to your application, going viral is not really something you can make happen. You can do things to help it along, even pay to grow a viral campaign (many campaigns that seem organic are actually quite contrived), but you can do little to guarantee its virality. Because that is often the plan on how to make the “numbers game” of Freemium work, and it is often little more than hope, I refer to it as the Freemium Virus.
As with many viruses, little good comes of them. Freemium seems to itself be viral and has spread to startups around the world, causing it to be perceived as more popular than it really is. An interesting look by Bernard Lunn on ReadWriteWeb at 103 SaaS companies found only 6% have a Freemium model, showing it is much less popular than the buzz might have you believe.
And with vendors like 37 Signals and Survey Monkey, among others, starting to de-emphasize their free versions based on pricing page re-designs, it seems this marketing tactic might be losing favor with some of the mass market, but still B2B SaaS companies who have traditionally leveraged the model. I think in B2C (entertainment, games, etc.) its likely here to stay and in those areas it makes a lot more sense.
JP- But I guess we need to focus on our sector- B2B SaaS and what actually works for us, not gaming or entertainment etc. Why do you say that Free = Zero Value? Surely if someone is using it they are trading their time and attention for your software?
LM- True or not, like it or not, some people think free means “no value”. Its a perception issue. While these types of blanket statements are great at getting people’s attention, you need to know your market. How does your market, and specifically those making the decision to bring your product into their company, feel about “free”? Its up to you to know this before you dive in with Freemium. It might be fine, or it could sink you before you leave dock.
JP- I guess a good way to do this would be to compare yourself against your competitors. If you’re competitors are free and gaining traction then it might be fine, but you’ve got to ask- why would you want to enter a sector where the competition gives away their product for free…. Do you think business customers have issues with free?
LM- I also said in the paper that if a startup offers their flagship product for free that this will throw all sorts of red flags up with potential customers. Startups, even well-funded ones, are a risk for business customers of all sizes. How does a startup giving away their core product look to the target market? Does it matter that you have multiple pricing tiers, SLAs, etc… nope, you lost because the end-customer couldn’t get past “free”. Its not fair, but that doesn’t matter. You didn’t know your market and you lost out because of it. Know your market is the overall lesson here.
A quick note, however, on how startups often deal with this lack of trust; they eliminate barriers to exit. To appease the customer, or potential customer, they often build tools to expedite the exporting of data (and in some cases application code) for the users when they want to leave. This effectively eliminates any need for the end-user to stick around, which means they likely won’t convert. This type of knee-jerk reaction is tied directly to the notion that their product is so great that once the users are in, they’d be crazy not to start paying.
JP – Thanks Lincoln- it’s been a pleasure and I can’t encourage people enough to check out the Freemium paper- it will open your eyes!
LM – Absolutley! Thank you for the opportunity to set the record straight and to help spread the word about this popular, touchy and potentially confusing subject.
JP- Thanks again!
Briefly, in other news:
Ash Maurya wrote a brilliant post on Continuous Deployment in practice- something Eric’s been blogging about for ages:
Of all the Lean Startup techniques, Continuous Deployment is by far the most controversial. Continuous Deployment is a process by which software is released several times throughout the day – in minutes versus days, weeks, or months.
The biggest waste in software is created from waiting for software as it moves from one state to another: Waiting to code, waiting to test, waiting to deploy. Reducing or eliminating these waits leads to faster iterations which is the key to success.
Ash then goes on to outline his journey and the techniques he used to achieve nirvana. Inspirational stuff.
Sachin Rekhi wrote a really nice post on what he’s learnt from being an Entrepreneur in Residence at a VC firm. If you’re in the process of creating a new business- these are some great points to think about:
I thought I would take a moment to share some of the most compelling lessons I learned during my Entrepreneurship-in-Residence.
Why Hasn’t it Existed Before. One of the most valuable thought exercises for innovative opportunities is to think about why your solution hasn’t existed in the market before. If there are structural changes that have occurred in say the last 5 years that make this opportunity possible today, then you may be on to something. If there are no such changes and you simply think that “no one has thought of this before” then I would urge you to dig deeper, as you may simply not have enough domain expertise to know why it is a bad idea.
Seek Leverage. For every startup opportunity, think about how to get the most leverage in the ecosystem to command the lion’s share of the value created, as opposed to being limited to taking a small sliver. This was one of the most important perspectives VCs offered. For example, Trinity helped me think through how to take an end-user product idea and expand my thinking to eventually become a platform for an entire ecosystem.
Human Behavior is Fundamentally Consistent. When looking at introducing new consumer behavior, look for analogies that exist in the offline world or using previous technology that may be applicable to this new behavior. For example, while Twitter is definitely a novel service, it can be fundamentally thought of as an expansion of the the traditional town hall concept.
Disrupt Incumbents. In certain spaces, after the market leaders gain significant market share, they stop innovating and simply focus on servicing and maintaining their existing customers. This creates opportunities to disrupt these spaces with cutting edge technology. For these spaces, think “if I were to re-invent this space from the ground up using present day technology, what would it look like?”
Original Thinking. When you look at some of the biggest innovations in the past decade (Facebook, Twitter, YouTube), they often come from original thinking. The founders of these startups are often young and unlike their older brethren, haven’t yet fossilized their world view into the existing paradigms. This thinking creates opportunities for innovation.
I particularly like #4- Disrupt incumbents with the key question- what would that service look like if it was invented today and unless the competitor has a brilliant innovation culture, there’s an opportunity. Steve Blank wrote a brilliant post today on firing founders- because of the mismatch of their early stage skills with the skills needed as the company develops. But he does point out that this often causes the company to lose it’s innovation and risk taking culture in favour of execution- which is no bad thing- but creates opportunities for startups to disrupt!
I wrote about two things this week- which you would have seen if you subscribe to the RSS- Is Cloud Computing doomed or EC2 over optimised and Cloud Consultancies view Customer acquisition as key to selection. The first post was interesting because since Amazon introduced spot pricing, latency has risen significantly- coincidence? Secondly I picked up an interesting comment in the Appregatta blog- that they actually look at a SaaS vendors Customer acquisition strategy as a key component to their selection process when they’re looking for SaaS products for their customers. They think if you’ve got a bad customer acquisition strategy you’re more likely to fail. SaaS vendors take note!
I couldn’t resist sharing this Dilbert on Cloud Computing with you:
Have a great weekend!
p.s. I’m still available to hire!