* The Business of Software

A former community discussing the business of software, from the smallest shareware operation to Microsoft. A part of Joel on Software.

We're closed, folks!

Links:

» Business of Software FAQ
» The Business of Software Conference (held every fall, usually in Boston)
» Forum guidelines (Please read before posting!)

Moderators:

Andy Brice
Successful Software

Doug Nebeker ("Doug")

Jonathan Matthews
Creator of DeepTrawl, CloudTrawl, and LeapDoc

Nicholas Hebb
BreezeTree Software

Bob Walsh
host, Startup Success Podcast author of The Web Startup Success Guide and Micro-ISV: From Vision To Reality

Patrick McKenzie
Bingo Card Creator

Limiting features in the trial

I know this gets talked about a bit on here, and the general consensus seems to be that a completely feature unlimited trial is the right way to do things, but this is my situation now:

We use a third party licensing system with a trial period feature. It turns out that this is rather easy to reset, and knowledge on how to do this has now made it onto a couple of forums frequented by my users. They talk rather openly about it, actually. Although I don't aspire to be 100% piracy free (I know its not an achievable target), I do want piracy to be at least a little bit of trouble for potential users. Now we have a situation where every new release can be used immediately by a large (and growing) body of users, with none of the risk inherent in running a cracked copy or reliance on someone to nobble the latest version. Just to illustrate the extent of the problem currently - one of these forums is now the single largest driver of traffic to my site.

The simplest solution seems to me to be to feature limit the trial, and the simplest way to do that would be to prevent the use of the final product. This is B2C and the result of our program is a file that can be used on other hardware/software. The process of creating the file would be unhindered, and work could be saved, used in the full version etc etc.

We've got a good "natural" chance to implement this now, as we have a major release coming up within a few weeks.

Even though I know it is unlikely that many of these users would buy, it is still the case that our sales have definitely dropped since this workaround became widely available, even more so in the country where this forum is based. It seems that it is just too easy and too tempting.

Is this a crap idea or should I try it? If not, then what? My thinking is to try it and see how bad the reaction is from potential real users. If there is a lot of blowback, I'll re-think things.

Our software retails for $150.

cheers...
Anon123 Send private email
Tuesday, October 21, 2014
 
 
My trial doesn't allow that to save, print or export if they have >30 guests. No limit on duration. Seems to work pretty well. See also:

http://successfulsoftware.net/2011/09/19/types-of-free-trial-for-software/

If you have the resources/inclination, you could try 2 approaches and A/B test them. If you do, let us know the results!
Andy Brice Send private email
Tuesday, October 21, 2014
 
 
Thanks Andy,
Perhaps feature-limiting is not as evil as I thought. The truth is, that our major competitor doesn't offer a trial at all, and a previous competitor offered a very limited trial that could only use preinstalled demo data.
What I am mulling is allowing everything within the program, including importing data, saving workspaces etc, and just preventing the final export. I realise that testing how this export works for a user might be a key element of the trial, but I can perhaps work around that by offering a range of finished examples for download. Who knows - my conversion might even go up like one of the cases quoted on your link..
Anon123 Send private email
Tuesday, October 21, 2014
 
 
Or watermark the export? or only allow export below a certain database size? Or only allow some types of export? There are lots of options. Depends very much on the product+market.
Andy Brice Send private email
Tuesday, October 21, 2014
 
 
About limited feature of time based trial.  This video should clear everything up: 
https://www.youtube.com/watch?v=nkhkPYXtX0k 

For us Limited feature trials always converted much better at least twice as good, but they do piss off the users.
C. Stark Send private email
Tuesday, October 21, 2014
 
 
I get very few complaints about my feature limited trial. Occasionally people will add 200 guests and then get annoyed because they can't save without buying a licence. But I do make this clear on the website and in a modal dialog when they start the trial. Its not my fault they didn't read it. I'm not trying to trick anyone.
Andy Brice Send private email
Tuesday, October 21, 2014
 
 
"For us Limited feature trials always converted much better at least twice as good, but they do piss off the users."

I love my users, but if pissing them off means that they buy at twice the rate - count me in :)

I've been thinking about this a little more over the last few hours, and I think that since we are now more established and known about in the field, limiting features probably won't hurt as much as it might have done in the past. The user still gets a very good look at our app, and can continue to use (and export) any data they have worked on in the trial once they buy - so fingers crossed this won't work against us and might even result in an increase in conversions. It is also true to some extent that users could have got all they needed out of the trial before, and even though people certainly expect to try before they buy, I also think that people are quite used to some sort of feature limiting during this process.
Anon123 Send private email
Tuesday, October 21, 2014
 
 
"Occasionally people will add 200 guests and then get annoyed because they can't save without buying a licence."

They might get annoyed, but they don't have any right to.

I imagine I might get the same, something along the lines of "I spent 2 days farting about making a project and now I can't export it so I can't do anything with it." To which I guess the response is that this is not freeware - if you use it and you like it, then feel free to buy it.
Anon123 Send private email
Tuesday, October 21, 2014
 
 
If you are going to change your trial model, why don't you A/B test the old and new trials models? Have 2 versions of the trial download page (1 with the old software, 1 with the new) and measure which one converts better. It shouldn't be much extra work if you are planning to code the new trial model any way.
Andy Brice Send private email
Tuesday, October 21, 2014
 
 
@Andy: Would not it be a better idea to show that dialog when the user crosses the guest number threshold, with a Buy Now button and an Unlock Code field? This way, you place the user in the driving seat - they can decide to see how it works for a smaller number of guests, buy your product right away, or proceed knowing they won't be able to save their work.

Another option I can think of is to let them save, but encrypt the file with a public key, and include (a piece of) the respective private key in the license.
Dmitry Leskov Send private email
Wednesday, October 22, 2014
 
 
"If you are going to change your trial model, why don't you A/B test the old and new trials models?"

I think the problem with this (for my specific situation) is that this would mean that both versions of the program are out there. All it would take is for the feature unlimited version to make it onto the forum in question, and the purpose of the whole exercise is lost. Our program is just an msi installer so would be very easy to distribute..

Now is a good moment to make the change because we have some major new features that a lot of people will want in the next release. If we make it later, my guess is that many users would stick the old version (they may, anyway). I don't mind that as much, I just want things to be a bit more hassle for them than they are now.
Anon123 Send private email
Wednesday, October 22, 2014
 
 
The question is how many of those forum visitors would pay $150 for your product vs how many people who would have paid, but only if you gave them a feature-unlimited trial.

Our trial packages have a "hard" expiration date in addition to the "soft" 90-days countdown set at install time. Our case is a bit special, though, as the files produced by our product are executables, so they also expire if created by a trial version. But if your product touches anything that has a timestamp, it can refuse to process data with timestamps beyond a certain date.
Dmitry Leskov Send private email
Wednesday, October 22, 2014
 
 
>All it would take is for the feature unlimited version to make it onto the forum in question, and the purpose of the whole exercise is lost. Our program is just an msi installer so would be very easy to distribute..

That wouldn't affect the A/B test, as you would only be measuring the difference in conversion rate between people who come to your download page.
Andy Brice Send private email
Wednesday, October 22, 2014
 
 
I'm not sure I get what you are saying Andy?
Track how many buy after downloading either the feature-limited or full trial version?
The issue really is that the motivation behind the change to feature-limiting is to stop a wide-spread abuse of the trial system, not so much to increase conversions. If conversions go up, great, if they go down or I get a lot of flack, not so great - but then again I think people are fairly well conditioned to expect some sort of limitation in a trial.
If I do an A/B test and track purchases from the app "buy" button (really, the only way I could do it at the moment), I would still leave the un-limited version in the wild, and I have no doubt that this installer would then be shared..
Anon123 Send private email
Wednesday, October 22, 2014
 
 
I am guessing the change in conversion rate from the website will dwarf the change in sales due to more/less abuses of the trial. I suggest concentrating on the majority (of mostly honest customers) and not fixating on a few less scrupulous ones.
Andy Brice Send private email
Wednesday, October 22, 2014
 
 
I agree with you Andy and I don't want to inconvenience users. *But* as mentioned in my first post this forum is now the single biggest driver of traffic to my forum, and my concern is that this is spreading beyond a few unscrupulous users - the workaround is too easy.

Having also floated my idea on a private forum I run for beta testers, most (all) of them seem to think its reasonable and I've got a bit more feedback from them suggesting that they know of quite a few people who *don't* circumvent the trial, but nonetheless have achieved all that they need within the 14 day trial period - so that's another group I hadn't thought about.

So I guess the question just boils down to how much I lose from limiting the trial (maybe nothing, maybe even a net gain) versus how many people buy that otherwise would not (also maybe nothing, of course).

I think I'll give it a couple of months of the new release with export disabled in the trial, and hopefully things will become clear after that and I can reverse or rethink.
Anon123 Send private email
Wednesday, October 22, 2014
 
 
I'm surprised no one has asked this yet, but why are you using a third party licensing system (and presumably paying for it) if it's so easy to crack?

Have you brought this up with the vendor?  Have you evaluated other licensing systems?
Jason Send private email
Wednesday, October 22, 2014
 
 
I have contacted the vendor, and the response was along the lines of "trials are inherently easy to circumvent, so don't use them".

Disappointing, as if the trial is easy to work around then it really doesn't matter how secure the licensing system itself is. The only proposed solution was to issue time limited keys instead of using the trial feature, but I really can't see how this would work. It would have to involve some sort of automatic generate and send of the trial key. and I don't see how I would stop someone requesting a new key every time a trial expires (you certainly couldn't do it based on the email address).

I must admit that I haven't spent a huge amount of time looking into how the current system works, but I would have hoped the trial was more of a "phone home with machine details" type thing than simply creating a hidden file on the computer, but it appears not.

I am actively looking into other options for our V2.0 - has anyone got any suggestions?
Anon123 Send private email
Wednesday, October 22, 2014
 
 
@Dmitry
Great idea to timestamp the output files. It is possible an output file was created and timestamped by an older version of the trial software after which the user downloaded and installed the current version of the trial. In my case, the EULA states that the trial is for 30 days or until the major version is updated at which time they can get the new version and start a new trial. Otherwise, using a newer trial violates the EULA which justifies blocking (or reminding) use of the dated output file. I think if  the software blocks/reminds use of the dated output file,  it should  explain why because the user may not have read the EULA and their perspective might be that the trial has not expired and they should not be blocked in any way.
Bill Anonomist Send private email
Wednesday, October 22, 2014
 
 
>> "trials are inherently easy to circumvent, so don't use them".

That's wrong. *Unverified* trials are easy to circumvent, *verified* trials are nearly impossible to circumvent without cracking (and depending on how they're implemented).

We talk about the upsides and downsides of verified vs. unverified trials here: http://wyday.com/limelm/help/trials/

We also plan on making verified trials much easier in LimeLM in the next 2 months. (Right now they're kind of a hassle).



>> "I must admit that I haven't spent a huge amount of time looking into how the current system works, but I would have hoped the trial was more of a "phone home with machine details" type thing than simply creating a hidden file on the computer, but it appears not.
"

Right, most licensing systems only support *unverified* trials (the hidden file type of trial).
Wyatt O'Day Send private email
Wednesday, October 22, 2014
 
 
@Wyatt

From your website on verified trials:
"4.If the key is a trial key, then use the result from GetFeatureValue("trial_expires") to compare the date against the current date to see if the trial has expired."

Can't the user get around this by just changing the system date on their computer or use Sandboxie for same purpose?
Bill Anonomist Send private email
Wednesday, October 22, 2014
 
 
>> "Can't the user get around this by just changing the system date on their computer ...?"

Not if they use our IsDateValid() function which detects date/time fraud.


>> "or use Sandboxie for same purpose?"

TurboActivate already detects Virtual Machines and lets you decide whether you want to allow the trial to run or not. In the version coming out in 2 months we're simplifying the entire interface and adding support for blocking the many "sandbox" and "time-stop" type applications (Sandboxie included).

That's the hassle of our current interface I was describing.
Wyatt O'Day Send private email
Wednesday, October 22, 2014
 
 
"the general consensus seems to be that a completely feature unlimited trial is the right way to do things"

I think that disabling saving and loading in a trial enables people to try it out, without making it so useful that they simply don't have to pay.
Scott Send private email
Wednesday, October 22, 2014
 
 
Limiting save, load  function  and watermarking the exported result in the trial version is a good solution.
MatrixFailure Send private email
Thursday, October 23, 2014
 
 

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics
 
Powered by FogBugz