Thanks for the insight. We're looking at pricing options for our new SaaS tool, and we have come to similar conclusions. It's important to have a free version (and not just a trial) so you establish a solid user base, but to make sure that there's enough premium content/features worth paying for down the line. Another focus could be enterprise-sales, where free accounts generate some attention but focus on enterprise-sales generate the revenue.
One thing that stood out was "Free users eat up support". I would have thought the first thing to offer to paid users but not free users would be support! By all means provide a forum for community-lead free support, but talking to an actual person should come at a cost.
Or, at least, such support needs to be best-effort and only insofar as it benefits the product. If a free user reports a bug, by all means put it in the queue to fix, prioritized appropriately for its impact to all users and not just the user who reported it.
>I would have thought the first thing to offer to paid users but not free users would be support!
Unfortunately, this makes the free plan less viable as a demo for the paid plan. People who sign up for a free plan to test out the service before getting a free plan may not convert if they don't get support.
Got tiered pricing? Get tiered support too. Not a difficult problem to solve, by any means. Support can go all the way from prioritized ticket systems (cheaper -> lower) to having the cell phone number of the guy in charge of your account.
Pretty sure that back in the day 37 Signals said that one of the mistakes they made was giving away a free plan for their apps, and that if they could undo that they would have.
Pricing is HARD - properly hard. It's so easy to get an imbalance between value offered and amount paid. It's equally as easy to get distracted by competitors undercutting you. Solution to this? Never compete on price, it's much better (and more sustainable) to offer value that sets you apart. At least this is what we've done with my SaaS B2B app http://www.staffsquared.com and so far so good :)
We regularly debate whether to continue offering a free tier. Usually one of us will read a blog post like this and we re-start the conversation where it left off last time. In our case, the free plan has survived because, with real analysis, they cost us almost nothing, ask for almost no support, and help us create mindshare of a service that many of our potential customers haven’t discovered yet. I’m not certain we’re correct to keep it, but so far we’ve not found anything that convinced us to stop offering it.
the impression I got was that they are just no longer offering the free plan. If you signed up when it was free, you are good. Otherwise, you have to pay now
If anybody wanted to bother they probably could hold them to it. They didn't really mention whether the existing accounts were grandfathered in. If not then it was rather dishonest of them to promote "free forever." They could have just said "free" instead.
If your software targets businesses, just charge. Many assume businesses won't pay because their initial target audience are tech savvy individuals who are used to freemium. I work for a small business. They are willing to pay.
One thing not discussed is the difference between $0 and $24. It's likely that some users that actually valued the service looked else where because they didn't offer a $10 or $15 option, not because they just wanted a free service. Of course, this is just speculation.
Yeah, we now have a 15$ option. You are correct, this is something we realized after getting started. The 15$ plan is in fact now our most popular plan.
One lesson I've learned, and that's hidden in this post, is to never use language like "free forever." It feels awesome and badass to write something like that, but I think for most of us it's hubris, we can't actually know that "forever" will work out.
Later, when the choice is between "iterate on your pricing model" or "go out of business", I also think most of us would iterate. But the feeling of making that "free forever" box disappear is much shittier than the little high you got from writing it.
One reason Freemium works for Pandora is because, according to an employee I've talked to there, the advertising revenue from a free user can be nearly as much or even greater than the 4.95/month from a paying user. Granted, I don't have a source other than the person's word and I doubt Pandora reports on their revenue split that way.
Pandora properly embraces the philosophy, "If you aren't paying, you're the product." If you can't monetize free users, then they are a liability, not an asset. Especially when they are eating up resources like support time. I have no idea why free users would even have access to the same support as paid users, as seems to have been the case here.
I can believe this. I have a free Pandora account and usually listen to it for 1-2 hours almost every day. In that time period I hear between 5-10 ads on average. I am not sure how much Pandora makes on each of those ads, but that is potentially around 200+ ads I am exposed to monthly. Plus with Pandora an advertiser should have a better than average chance their ad is heard/seen since the user is forced to hear/see the ad in order to continue listening in most cases.
Not sure why companies are still obsessed with obtaining millions of cheap/non-paying users? Fantastic, you've acquired an audience of people so cheap they refuse to buy your entry level tier product.. Sounds like an awful problem.
At the very least, if you're going to offer a free tier, make it a trial and capture their credit card up front. Charge it after 30 days. Make the user get serious about whether this product is for them. And in turn, it makes your company get serious as to whether this is a viable business model or not.
make it a trial and capture their credit card up front. Charge it after 30 days.
FWIW, I personally refuse to do business with companies that utilize this kind of flow - it feels like a scummy dark pattern, where they're banking on people to forget they've signed up and they can get at least one month's worth of charge out of it. That goes double when you can sign up online but it requires calling in (and talking to "retention" people) to cancel.
"But we don't want to interrupt your service.." goes the salesman.. yeah right.
If you're going the free trial route, fine, but it should require affirmative action after the trial to begin billing. You don't need my credit card until I've actually bought something from you.
I agree. Ticking time bomb on the account is fine, and a series of triggered reminder emails should be sufficient to prevent the dead switch surprise. Asking for payment information before I even begin a free trial leads me to assume that you don't believe your product is good enough to withstand the trial period on its own merits, and you are hoping to catch me just forgetting about it. Makes me wonder what other questionable actions might be taken, especially regarding hard things like security. Does not make me feel safe at all.
FWIW, I personally refuse to do business with companies that utilize this kind of flow
Well, I think you know the river runs both ways, they probably want to filter out people who share that mentality, not judging your views on morality or whatever but it does not support your argument at all. If you are not willing to provide a card up front it means you don't have money to spend right there and then, and those companies probably are more interested in talking to those who have shown they are able to pay if required.
I understand the want to filter out people mentality, but I completely disagree with thinking that just because a person does not want to provide a credit card upfront that they are not a valid customer or don't have the money to spend.
I look at it this way. Let's say I come across Acme Corporation's website. They are selling what appears to be a really cool product. I figure I will give it a try and signup for the free trial and at the same time enter my credit card info. I give it whirl, but in the middle of testing it out I get sidetracked with a phone call or whatever and forget to finish testing. I wasn't impressed enough with the little testing I had completed to go back immediately and after a couple days I completely forget I even signed up. Now 29 days later my credit card gets whacked for a $20 monthly fee. I have two choices at this point. I can file a chargeback, which isn't fair to the company. Or I can eat the cost and cancel immediately, which isn't fair to me.
In my opinion, there is absolutely no reason to demand a credit card upfront for a trial.
Which as recent FDA actions can tell us, are often worth the paper they're printed on.
Sign up now, give my credit card, and to cancel I have to call a phone number which may or may not work and talk to a person who's entire reason for employment is to make it difficult for me to cancel with no assurance that anything has actually been cancelled until the next billing cycle.
When I take over the world, it will be a law on pain of summary execution that any recurring billing which can be signed up for online must be cancellable online :/
Well if you can pay for a trial even better, the credit card requirement is just another step below that,the next which is a free trial with no card. Whichever they offer is valid because businesses differ. To claim requiring a card for a trial is shady is just one small irrelevant aspect for the vendor.
The card requirement is a great way to filter out the less serious ones, if we miss one exception like yourself, it's no biggie, there's always someone who need the solution and willing to pay for it.
The only reason where I'd be reluctant to do a card trial is if it was difficult to cancel or they didn't offer a refund if I accidentally forget.
Plenty of shady people using fake cards to sign up for trial as well.
It means I want to try before I buy and those who use dark patterns probably cut in other areas like security. Online it is too easy to lie about your product and too hard to get refunds in case of false promises.
If someone was selling you a phone at a flea market, would you give them your CC number before you knew the phone turned on?
Paying with a debit card isn't much different from using only cash. I know who to file a report against if the product isn't delivered, but the amount that I expect to recover compared to the effort makes it comparable to having paid in cash with no recovery.
I totally understand your pain but I don't agree. We got rid of the CC on our signup (www.nusii.com). We got many more users because of that (3 to 4 times more) but we've seen that the conversion rate (trial to paid) was WAY lower. 4% instead of 40%. We don't exactly know why this is, but I think it is because the CC makes you want to try the product much more. You end up with a happier user.
Our biggest mistake was getting rid of the CC on the signup. I could write a very similar blog post. Free trial period with a 60 days money back guaranty works the best for us and our users.
Not that I'm any expert in this field, but how does your position rhyme with the observation that almost every big-time saas / paas / *aas platform out there uses that model?
- Github
- Heroku
- Cloudflare
- MS Azure
AWS perhaps being the biggest one to charge immediately, although that's so incredibly low, I wonder if it still even counts. And even they have a free tier...
The free tier is for 1 year, so technically, it's a limited time trial and does require your credit card up front (which is their main method of converting free accounts -- upselling on that credit card).
I think AWS is the type of free the GP supports: limited time, credit card up front, clear upsell channels for the limited service.
GitHub have clearly benefited from the network effects of being the de-facto platform for open source. Their free tier doesn't include any private repos, so it won't cannibalise sales of paid tiers. I think that their free tier is effectively a trial on steroids - you're free to contribute to open source projects, engage with the user community and learn the platform, but you have to start paying from day one if you intend to use GitHub for commercial projects.
Heroku's free tier is again a sort of supercharged free trial, in that the free tier is too limited to use in production. The free tier offers very little value to users who have no intention of ever upgrading to a paid tier.
AWS's free tier is just a very generous free trial - after twelve months, you have to start paying.
I think in all cases, the key factor is whether there's a concrete plan for how the free tier feeds into your business. If your product has high switching costs and requires a big investment in learning to use effectively, then it may make sense to have a free tier specifically aimed at letting people learn the platform. Network effects are the obvious case where free users have real business value. I think that the mistake comes from the presumption that there ought to be a free tier, regardless of whether there's a compelling business case for it.
As regards you contention that almost every big SAAS product offers a free tier, I think there are at least as many counterexamples - Salesforce, SAP, Citrix, Blackboard, Basecamp, Moz, Kissmetrics.
Not sure why companies are still obsessed with obtaining millions of cheap/non-paying users? Fantastic, you've acquired an audience of people so cheap they refuse to buy your entry level tier product.. Sounds like an awful problem.
THIS. Basically, most people are just following the herd. "Oh my competitor has it I better do the same", except you might not be as big as 37signals or Salesforce to be able to support such a huge ratio of freeloaders vs paying users.
With a credit card trial, you get less number of freeloaders but they are still there.
Free user can promote service with his friends, can blog about it, etc. Basically you are paying (usually) negligible cost of maintaining free account for some possible marketing. Whether it works or not is questionable, but obviously it does work in some areas.
I'm the author of this post and yes, this is exactly what we found. All I can do is speak from experience in selling a B2B app that is priced at $5 a month / user (I can see how this would be different for different pricing models), but what we found was that these were not the users that were talking about the app and spreading the word for us.
RE: collect credit card up front... This is one of the next tests we are going to set up in our on-boarding process. Right now we don't bill the cc (or ask for it) up front.
>>> RE: collect credit card up front... This is one of the next tests we are going to set up in our on-boarding process. Right now we don't bill the cc (or ask for it) up front.
Please, pretty please, don't do that. I am really not sure if a few extra bucks are worth having a bunch of frustrated users.
Apologies for a off-topic question : isn't HubStaff from the same guys as Hubspot, which is very well funded marketing platform? if so, isn't Hubstaff a distraction from Hubspot? Or is this the proverbial pivot ?
They are not affiliated, and are in quite different stages.
Hubstaff currently nets some $40k MRR (you can see their live metrics here [0], pretty cool), whereas Hubspot has $115 million dollars in annualized revenue [1].
I was contemplating to pick a freemium model for an app I'm building but after reading this (and few more) post(s) from this blog I changed completely my mind. Premium + Free Trial is actually the best choice for many SaaS projects. Read it carefully...
A long time ago Mailchimp had a great post on the free pricing plan. One of the most important observations they made was "Wait until you have a full blown support team in place before offering a free plan" So they were basically saying wait 2 - 3 years in at least until you could layer on the additional free support for these plans before offering them.
I once worked at a non-tech startup. It was a startup, it was less than 50 people, it catered to film industry rather than tech. What I learned was this -- giving your software away for free will probably give you bad rep with businesses.
In that company, the head of the digital team "understood" tech, obviously. But it took us insane amount of work to convince management to use a free product. They thought of it as a toy, which would cost company time (and hence money) to implement and get used to, and it won't work. And because it's free, probably nobody would come fix it. They were always worried of a 2-person ramen hogging company running away with their data.
I also learnt that they paid in name of support, even if we did not need the product. We bought highly expensive Ooyala licenses just to upload videos to Youtube (Ooyala offers a lot more, but we didn't care for that). Yes, I know we can directly upload them, but as Youtube was free, they didn't trust the upload system in case of crisis. With Ooyala, they had someone to call if shit hits the fan during critical and timely upload.
Free plans are weird. I would rather try offering companies a demo or pilot. And offer a reasonable money back guarantee. At least with making the customer contact you first for demo, you're in conversation with them. You can gauge if they will ever buy the product. If yes, you have a shot at selling.
Automation of demo is awfully lazy for a SaaS startup that's just starting out. Outreach, talk, connect. The lazy free automated trial only helps in case you have a me-too product where the customer is jumping from product to product trying to figure out which one to use. If that's the case, you either don't have a USP or don't know how to use it. Either way, you end up in a spiral to the bottom.
I believe the simplistic division of product offerings was the problem. Pay to get more users? To easy to circumvent. Divide the product along feature set boundaries, where the user can self-select their way past pain points, or between spending their time or their money to save time, then I think you get more conversions to a paid subscription.
I think it depends on what it takes to find out if your service is right for me, which is dependent on the type of product you have.
For Zencoder, for instance, I wanted to know if I can interface with their API, and do so effectively and easily. So Zencoder offers a free tier, but the encoding is capped at ~5 seconds. I can't actually use their service as a product, but I can use it as a test and development bed. There were competitors I did not evaluate because I could not evaluate them as easily.
So the question becomes, what does it take for your service to offer a trial, so that I can determine it's value to me?
52 comments
[ 3.1 ms ] story [ 116 ms ] threadOne thing that stood out was "Free users eat up support". I would have thought the first thing to offer to paid users but not free users would be support! By all means provide a forum for community-lead free support, but talking to an actual person should come at a cost.
Unfortunately, this makes the free plan less viable as a demo for the paid plan. People who sign up for a free plan to test out the service before getting a free plan may not convert if they don't get support.
Pricing is HARD - properly hard. It's so easy to get an imbalance between value offered and amount paid. It's equally as easy to get distracted by competitors undercutting you. Solution to this? Never compete on price, it's much better (and more sustainable) to offer value that sets you apart. At least this is what we've done with my SaaS B2B app http://www.staffsquared.com and so far so good :)
Later, when the choice is between "iterate on your pricing model" or "go out of business", I also think most of us would iterate. But the feeling of making that "free forever" box disappear is much shittier than the little high you got from writing it.
At the very least, if you're going to offer a free tier, make it a trial and capture their credit card up front. Charge it after 30 days. Make the user get serious about whether this product is for them. And in turn, it makes your company get serious as to whether this is a viable business model or not.
FWIW, I personally refuse to do business with companies that utilize this kind of flow - it feels like a scummy dark pattern, where they're banking on people to forget they've signed up and they can get at least one month's worth of charge out of it. That goes double when you can sign up online but it requires calling in (and talking to "retention" people) to cancel.
"But we don't want to interrupt your service.." goes the salesman.. yeah right.
If you're going the free trial route, fine, but it should require affirmative action after the trial to begin billing. You don't need my credit card until I've actually bought something from you.
Well, I think you know the river runs both ways, they probably want to filter out people who share that mentality, not judging your views on morality or whatever but it does not support your argument at all. If you are not willing to provide a card up front it means you don't have money to spend right there and then, and those companies probably are more interested in talking to those who have shown they are able to pay if required.
I look at it this way. Let's say I come across Acme Corporation's website. They are selling what appears to be a really cool product. I figure I will give it a try and signup for the free trial and at the same time enter my credit card info. I give it whirl, but in the middle of testing it out I get sidetracked with a phone call or whatever and forget to finish testing. I wasn't impressed enough with the little testing I had completed to go back immediately and after a couple days I completely forget I even signed up. Now 29 days later my credit card gets whacked for a $20 monthly fee. I have two choices at this point. I can file a chargeback, which isn't fair to the company. Or I can eat the cost and cancel immediately, which isn't fair to me.
In my opinion, there is absolutely no reason to demand a credit card upfront for a trial.
*Edited to fix a typo.
Sign up now, give my credit card, and to cancel I have to call a phone number which may or may not work and talk to a person who's entire reason for employment is to make it difficult for me to cancel with no assurance that anything has actually been cancelled until the next billing cycle.
When I take over the world, it will be a law on pain of summary execution that any recurring billing which can be signed up for online must be cancellable online :/
The card requirement is a great way to filter out the less serious ones, if we miss one exception like yourself, it's no biggie, there's always someone who need the solution and willing to pay for it.
The only reason where I'd be reluctant to do a card trial is if it was difficult to cancel or they didn't offer a refund if I accidentally forget.
Plenty of shady people using fake cards to sign up for trial as well.
If someone was selling you a phone at a flea market, would you give them your CC number before you knew the phone turned on?
Our biggest mistake was getting rid of the CC on the signup. I could write a very similar blog post. Free trial period with a 60 days money back guaranty works the best for us and our users.
Are you judging a conversion as someone that's paid for at least one month off service?
- Github
- Heroku
- Cloudflare
- MS Azure
AWS perhaps being the biggest one to charge immediately, although that's so incredibly low, I wonder if it still even counts. And even they have a free tier...
I think AWS is the type of free the GP supports: limited time, credit card up front, clear upsell channels for the limited service.
Well, it worked out for GitHub because:
1) they have a great product and everyone can see how great it is, without signing up;
2) because of the nature, the word spreads exponentially. One free user can spread the word to literally tens of thousands of potential paying users.
So GitHub is in a rather unique position.
GitHub have clearly benefited from the network effects of being the de-facto platform for open source. Their free tier doesn't include any private repos, so it won't cannibalise sales of paid tiers. I think that their free tier is effectively a trial on steroids - you're free to contribute to open source projects, engage with the user community and learn the platform, but you have to start paying from day one if you intend to use GitHub for commercial projects.
Heroku's free tier is again a sort of supercharged free trial, in that the free tier is too limited to use in production. The free tier offers very little value to users who have no intention of ever upgrading to a paid tier.
AWS's free tier is just a very generous free trial - after twelve months, you have to start paying.
I think in all cases, the key factor is whether there's a concrete plan for how the free tier feeds into your business. If your product has high switching costs and requires a big investment in learning to use effectively, then it may make sense to have a free tier specifically aimed at letting people learn the platform. Network effects are the obvious case where free users have real business value. I think that the mistake comes from the presumption that there ought to be a free tier, regardless of whether there's a compelling business case for it.
As regards you contention that almost every big SAAS product offers a free tier, I think there are at least as many counterexamples - Salesforce, SAP, Citrix, Blackboard, Basecamp, Moz, Kissmetrics.
THIS. Basically, most people are just following the herd. "Oh my competitor has it I better do the same", except you might not be as big as 37signals or Salesforce to be able to support such a huge ratio of freeloaders vs paying users.
With a credit card trial, you get less number of freeloaders but they are still there.
RE: collect credit card up front... This is one of the next tests we are going to set up in our on-boarding process. Right now we don't bill the cc (or ask for it) up front.
Please, pretty please, don't do that. I am really not sure if a few extra bucks are worth having a bunch of frustrated users.
Hubstaff currently nets some $40k MRR (you can see their live metrics here [0], pretty cool), whereas Hubspot has $115 million dollars in annualized revenue [1].
[0] https://hubstaff.baremetrics.com/ [1] http://marketingland.com/hubspot-2014-earnings-report-revenu...
http://sixteenventures.com/freemium-or-free-trial
In that company, the head of the digital team "understood" tech, obviously. But it took us insane amount of work to convince management to use a free product. They thought of it as a toy, which would cost company time (and hence money) to implement and get used to, and it won't work. And because it's free, probably nobody would come fix it. They were always worried of a 2-person ramen hogging company running away with their data.
I also learnt that they paid in name of support, even if we did not need the product. We bought highly expensive Ooyala licenses just to upload videos to Youtube (Ooyala offers a lot more, but we didn't care for that). Yes, I know we can directly upload them, but as Youtube was free, they didn't trust the upload system in case of crisis. With Ooyala, they had someone to call if shit hits the fan during critical and timely upload.
Free plans are weird. I would rather try offering companies a demo or pilot. And offer a reasonable money back guarantee. At least with making the customer contact you first for demo, you're in conversation with them. You can gauge if they will ever buy the product. If yes, you have a shot at selling.
Automation of demo is awfully lazy for a SaaS startup that's just starting out. Outreach, talk, connect. The lazy free automated trial only helps in case you have a me-too product where the customer is jumping from product to product trying to figure out which one to use. If that's the case, you either don't have a USP or don't know how to use it. Either way, you end up in a spiral to the bottom.
For Zencoder, for instance, I wanted to know if I can interface with their API, and do so effectively and easily. So Zencoder offers a free tier, but the encoding is capped at ~5 seconds. I can't actually use their service as a product, but I can use it as a test and development bed. There were competitors I did not evaluate because I could not evaluate them as easily.
So the question becomes, what does it take for your service to offer a trial, so that I can determine it's value to me?