This is such a big one for me. At my previous workplace it was fairly low friction to try new tools. I'd often just sign up, start a trial, play around with it, and start trying to get people interested. Eventually we might realise it would solve some problems, we'd then start signing people up and getting data into it. But by that point the 15-30 day trial had inevitably expired.
I would often reach out to ask for trial extensions, and was always given these, but that's friction! One thing I loved about Linear was that the trial was pegged to the number of issues you created in it which meant we could take our time and work it into our processes without them dictating a schedule, then only when we were seeing value would we need to start paying.
More services, particularly B2B SaaS companies, need to figure out trials that aren't based on time. Some tools took >1 year to go from first person on the team signing up for a trial, to valuable use.
Free tiers are hugely impactful for the cases you mention where people onboard is the limiting factor for value not time.
I find the really interesting part about free tiers is balancing both how much of your user base is monetizable and how much value you give in the free tier vs save for your higher tiers.
If you have a product that is valuable to an individual or small team, but only monetize at 5 people, you will miss out on being able to monetize a large percentage of your users.
If your free tier is too valuable people won't be compelled to upgrade, not valuable enough and people won't be compelled to give the product a try or stick around long enough to see or want the value of a higher tier.
I think this could make a good topic for our next article. Would love to hear others experiences and thoughts on free tiers vs free trials!
> If your free tier is too valuable people won't be compelled to upgrade, not valuable enough and people won't be compelled to give the product a try or stick around long enough to see or want the value of a higher tier.
This is my experience when trying pick a read-it-later/bookmarking app. I’m currently using the free tier of raindrop.io and I don’t plan to upgrade at all. The only features I need is bookmark storage and highlights. I don’t think it’s too valuable, but it’s valuable enough to satisfy all I want from a bookmark manager.
When I used Instapaper and Pocket in the past, because of their limited number of highlights in the free tier, I found myself peeking at the pricing page constantly and pondering whether I should upgrade. I didn’t have enough money to pay for yet another subscription, but the pressure to upgrade is there. I think I even subscribed for a month or two.
My other thought is on the granularity of upgrades. Probably for simplicity, both for the users and developers, upgrades tend to limited to maybe 2-5 tiers with the base tier being free. Raindrop.io, Instapaper, and Pocket’s page on the premium tiers tend to list a lot of features but I find myself skipping through most of them. When I see that list, I know they’re charging me for the whole feature set when I only need one or two. As a result, I’m quite hesitant to subscribe to any of them. The ratio of need vs offered is important to me; if those features were compiled into a checklist and I only check one or two, then I’m out of the premium funnel.
An example of decent intersection between what I need and what is being offered, is Tutanota’s private paid tier. Even though, there are still some features I don’t need, but the price is low enough that I can tolerate the feeling of not utilising my subscription enough. They also offer some other mini-upgrades and I can see why they’re an extension to the paid tier instead of being a part of it.
I've learn't a lot about this for my product[0]. Often teams deploy web apps multiple times a week (or day!) so having a time based trial works. For web extensions, some teams release once a week, others once a month and some a few times a year. When you're at the mercy of a review cycle (Chrome Web Store can take hours to days), it really impacts how rapidly you can release your work.
If I had a 2 week trial - those that submit web extensions to stores rarely wouldn't see the benefit of the product before their trial was over. I made it simple and capped it at 5 successful submissions. Success is important here because you can go through the review cycle 5 times ('5 failed submissions' - this number could change, but it works for now) and then your trial would be over. At the end of the day, regardless of how fast teams work, I want everyone to see the value under the same conditions.
I've worked at places that are focused on time trials because there's no practical way to limit the product in another sense that would give users time to fit in their workflow but also limit their actions enough to convert. Perhaps that's a failure of the scope of the product etc. but it works well enough for them.
Did you test conversion or is it your untested hypothesis that threat of cutting off a trial before so many valuable interactions converts worse than waiting on further use. You have a logic but unclear if it matches reality
> The number of users that take part in trials doesn't matter to your bottom line. The percentage of people that convert does.
I stopped reading at this point. What actually matters is the absolute number of people that sign up and pay for your product. Making it harder for people to try out your product obviously lowers the absolute number of users who get to know your product, which has knock-on effects that you can not measure in A/B tests.
If your product requires me to provide credit card information, or even contact sales, it better be exceptional for me not to bail instantly. If it is exceptional, your marketing probably does not matter much anyway.
We weren't as clear as we should have been here. Apologies for that!
The total number of subscribers you get at the end or total ARR is the key metric as you say, trial to paid is an important metric that goes into that analysis and is a good leading indicator.
You are 100% right that more friction makes people more likely to bail, lower the number of people who start a trial.
Credit cards in particular are an interesting one, because you are effectively asking people to sign up twice if you don't require them for the trial. Once when they create an account and again 7 or 30 days later when the trial ends.
That friction being added for people to continuing to use the product has worse outcomes that getting the friction out of the way up front.
I'm sure it just depends on each product. Where it gets tricky is when customer service costs increase with trial users who aren't paying for the product. I mean, this isn't rocket science. Do you make more money with similar lead generation from credit card upfront or not?
"The only people willing to go through the trouble of creating a new email address or listing someone else's address just to get another trial, are people who love your product."
Completely wrong.
I've seen plenty of cases where initial setup went wrong and there wasn't a way to reset the account to the initial state, so then creating a 2nd one was the only way to actually do the trial.
Also, life sometimes gets in the way so that I sign up for a trial and then the 7 days are over before I ever had time to login and look around.
What kind of setup you've had? I guess the author is making assumption for consumer - easy to setup apps. Enterprise or team software is in software often completely different world
Story time. When I worked on ESPN.com's Fantasy Football rebuild in 2003, showcasing an upgrade was by far the largest converter of free users to paid. You could play with your N buddies where everybody was playing for free, but you could also individually upgrade.
Some context: during NFL games on Sunday your team of players would be playing someone else's in your league. As players on your team did things in the real-world NFL games, your fantasy football team would accrue points. There were two ways to see how many points your team had scored: you could wait until Monday morning when the weekly job ran, or you could pay to see your team's live scoring updates as they happened in real time on Sunday afternoons/evening.
Here's what drove the most individual upgrades. Someone on my team got the bright idea to show some players' live updates on your roster as a "teaser". You might have 20 players, and two would show live scores. There'd be a big CTA to upgrade to see the other 18 players' scores. It was admittedly pretty cool - you'd watch an NFL game, see one of your players score a touchdown, and then seconds later see his score change in your fantasy league.
I don't remember the exact conversion rate but IIRC it was 15%+ (2% is incredible conversion for free to paid).
That's one reason why, in my SaaS trial to paid consulting[1], I recommend scattering in-app CTAs to upgrade where contextually relevant. It still converts like crazy (which makes sense - you're trying to do a job, the path forward is right in front of you, and you can pay a few $$ to get the job done).
There are lots of instances where I find a tool I want to try, sign up for the trial, see the inside of the tool and think, "I don't have time for this, I'll try it tomorrow". Inevitably, the 2 weeks or whatever run out and I've barely touched it.
It seems like it would be better to base the trial on usage rather than time and/or offering a generous free tier with no-brainer upgrades once you're committed.
Yes! Credit to Trailforks for doing this, btw - 10 day free trial which I assumed was the usual nonsense but they're actually 10 days of non-contiguous use.
If anything, it seems like putting a time restriction in that case is actively good for the business, as you want to create a sense of urgency for the person to experiment before the time runs out. If it's on usage, the person trialing has no incentive to engage immediately?
In my case, that often means that I simply won't engage at all.
I usually find myself wondering if now is really the best time to start the trial, or whether I'd be better off saving it for another time. I might need it more later, or have a project that's a good fit for the trial, or have more free time later, so I put it off. Then that cycle repeats until I eventually forget about the product entirely and move on to something else.
It's a bit like the "Too Awesome to Use" trope [0].
I am much more likely to engage with feature-limited trials without an expiration, or at least usage-based trials.
For our own SaaS[0], we provide a timed trial. But we regularly provide trial extensions because reality of business is that it takes time to get everybody on board and onboarded.
Reading this post, I suspect '30 days of use' would result in less 'please extend the trial' emails and would mean less friction during the trial.
However, there is a tradeoff: when somebody reaches out for a trial extension, it may be their first contact, where having a fast and effective response gives them a sense of the level of service we provide. Sometimes it also results in further conversation about how we can help them make a stronger pitch to whoever internally is approving the purchase. So a looser trial would mean less of these conversations.
22 comments
[ 0.23 ms ] story [ 57.3 ms ] threadThis is such a big one for me. At my previous workplace it was fairly low friction to try new tools. I'd often just sign up, start a trial, play around with it, and start trying to get people interested. Eventually we might realise it would solve some problems, we'd then start signing people up and getting data into it. But by that point the 15-30 day trial had inevitably expired.
I would often reach out to ask for trial extensions, and was always given these, but that's friction! One thing I loved about Linear was that the trial was pegged to the number of issues you created in it which meant we could take our time and work it into our processes without them dictating a schedule, then only when we were seeing value would we need to start paying.
More services, particularly B2B SaaS companies, need to figure out trials that aren't based on time. Some tools took >1 year to go from first person on the team signing up for a trial, to valuable use.
Free tiers are hugely impactful for the cases you mention where people onboard is the limiting factor for value not time.
I find the really interesting part about free tiers is balancing both how much of your user base is monetizable and how much value you give in the free tier vs save for your higher tiers.
If you have a product that is valuable to an individual or small team, but only monetize at 5 people, you will miss out on being able to monetize a large percentage of your users.
If your free tier is too valuable people won't be compelled to upgrade, not valuable enough and people won't be compelled to give the product a try or stick around long enough to see or want the value of a higher tier.
I think this could make a good topic for our next article. Would love to hear others experiences and thoughts on free tiers vs free trials!
This is my experience when trying pick a read-it-later/bookmarking app. I’m currently using the free tier of raindrop.io and I don’t plan to upgrade at all. The only features I need is bookmark storage and highlights. I don’t think it’s too valuable, but it’s valuable enough to satisfy all I want from a bookmark manager.
When I used Instapaper and Pocket in the past, because of their limited number of highlights in the free tier, I found myself peeking at the pricing page constantly and pondering whether I should upgrade. I didn’t have enough money to pay for yet another subscription, but the pressure to upgrade is there. I think I even subscribed for a month or two.
My other thought is on the granularity of upgrades. Probably for simplicity, both for the users and developers, upgrades tend to limited to maybe 2-5 tiers with the base tier being free. Raindrop.io, Instapaper, and Pocket’s page on the premium tiers tend to list a lot of features but I find myself skipping through most of them. When I see that list, I know they’re charging me for the whole feature set when I only need one or two. As a result, I’m quite hesitant to subscribe to any of them. The ratio of need vs offered is important to me; if those features were compiled into a checklist and I only check one or two, then I’m out of the premium funnel.
An example of decent intersection between what I need and what is being offered, is Tutanota’s private paid tier. Even though, there are still some features I don’t need, but the price is low enough that I can tolerate the feeling of not utilising my subscription enough. They also offer some other mini-upgrades and I can see why they’re an extension to the paid tier instead of being a part of it.
If I had a 2 week trial - those that submit web extensions to stores rarely wouldn't see the benefit of the product before their trial was over. I made it simple and capped it at 5 successful submissions. Success is important here because you can go through the review cycle 5 times ('5 failed submissions' - this number could change, but it works for now) and then your trial would be over. At the end of the day, regardless of how fast teams work, I want everyone to see the value under the same conditions.
I've worked at places that are focused on time trials because there's no practical way to limit the product in another sense that would give users time to fit in their workflow but also limit their actions enough to convert. Perhaps that's a failure of the scope of the product etc. but it works well enough for them.
0 - https://atriumapp.dev
I stopped reading at this point. What actually matters is the absolute number of people that sign up and pay for your product. Making it harder for people to try out your product obviously lowers the absolute number of users who get to know your product, which has knock-on effects that you can not measure in A/B tests.
If your product requires me to provide credit card information, or even contact sales, it better be exceptional for me not to bail instantly. If it is exceptional, your marketing probably does not matter much anyway.
The total number of subscribers you get at the end or total ARR is the key metric as you say, trial to paid is an important metric that goes into that analysis and is a good leading indicator.
You are 100% right that more friction makes people more likely to bail, lower the number of people who start a trial.
Credit cards in particular are an interesting one, because you are effectively asking people to sign up twice if you don't require them for the trial. Once when they create an account and again 7 or 30 days later when the trial ends.
That friction being added for people to continuing to use the product has worse outcomes that getting the friction out of the way up front.
Completely wrong.
I've seen plenty of cases where initial setup went wrong and there wasn't a way to reset the account to the initial state, so then creating a 2nd one was the only way to actually do the trial.
Also, life sometimes gets in the way so that I sign up for a trial and then the 7 days are over before I ever had time to login and look around.
Story time. When I worked on ESPN.com's Fantasy Football rebuild in 2003, showcasing an upgrade was by far the largest converter of free users to paid. You could play with your N buddies where everybody was playing for free, but you could also individually upgrade.
Some context: during NFL games on Sunday your team of players would be playing someone else's in your league. As players on your team did things in the real-world NFL games, your fantasy football team would accrue points. There were two ways to see how many points your team had scored: you could wait until Monday morning when the weekly job ran, or you could pay to see your team's live scoring updates as they happened in real time on Sunday afternoons/evening.
Here's what drove the most individual upgrades. Someone on my team got the bright idea to show some players' live updates on your roster as a "teaser". You might have 20 players, and two would show live scores. There'd be a big CTA to upgrade to see the other 18 players' scores. It was admittedly pretty cool - you'd watch an NFL game, see one of your players score a touchdown, and then seconds later see his score change in your fantasy league.
I don't remember the exact conversion rate but IIRC it was 15%+ (2% is incredible conversion for free to paid).
That's one reason why, in my SaaS trial to paid consulting[1], I recommend scattering in-app CTAs to upgrade where contextually relevant. It still converts like crazy (which makes sense - you're trying to do a job, the path forward is right in front of you, and you can pay a few $$ to get the job done).
1- http://www.trialtopaid.com
It seems like it would be better to base the trial on usage rather than time and/or offering a generous free tier with no-brainer upgrades once you're committed.
I usually find myself wondering if now is really the best time to start the trial, or whether I'd be better off saving it for another time. I might need it more later, or have a project that's a good fit for the trial, or have more free time later, so I put it off. Then that cycle repeats until I eventually forget about the product entirely and move on to something else.
It's a bit like the "Too Awesome to Use" trope [0].
I am much more likely to engage with feature-limited trials without an expiration, or at least usage-based trials.
[0] https://tvtropes.org/pmwiki/pmwiki.php/Main/TooAwesomeToUse
I could see a few scenarios where this might be a useful approach.
Reading this post, I suspect '30 days of use' would result in less 'please extend the trial' emails and would mean less friction during the trial.
However, there is a tradeoff: when somebody reaches out for a trial extension, it may be their first contact, where having a fast and effective response gives them a sense of the level of service we provide. Sometimes it also results in further conversation about how we can help them make a stronger pitch to whoever internally is approving the purchase. So a looser trial would mean less of these conversations.
Would have to test :)
[0] - https://enchant.com
We are writing up some content about free tiers as a follow up based on the feedback we got from this thread. Looking forward to sharing it.
We just shared a follow up article on how to deal with repeated trials that expands up on what we included in this article
https://news.ycombinator.com/item?id=32696631 https://upollo.ai/blog/when-users-repeat-trials