Should We Release Now?
A common question with software companies is when to release the next version of their software. Should it be as soon as enough new features and benefits are available that people will be interested in? Should it be when there’s enough value that virtually everyone will want to purchase an upgrade rather than only a percentage? Should it be every month? Should it be once every year, two years, etc.? This is a very hard question to answer and every software company handles it differently.
Not only does every company handle it differently, but sometimes different releases of the same product/project are handled differently! And this happens to be the case with this latest version of LandlordMax. Normally I only like to release a major version of LandlordMax when we’ve added enough new features that it will excite a large percentage of our customers! If it’s only a few features, we sometimes will just release that as a patch, but I generally like to push features to major releases.
This particular upcoming release will be different than usual in that it’s only been a short time since the last release. Although version 2.12 states that it was released in December, it was actually released in April. That means it’s been just 6 months, which is very quick for a new major release. Is it too soon is the better question! And this is where I’m facing a very big delimna which I’ve finally resolved over the last few weeks.
I’m sure you’re wondering why it was a difficult choice, and I’m going to explain why right now. In the upcoming version we’ve added several features that have a major impact on how people perform their daily tasks. One of these features is the “Late” button shown in this animated tutorial. Although this is just a simple button, the amount of effort it will save some of our customers is significant. Another small feature that has large implications is that when you now create a new lease, a scheduled accounting entry will also automatically be created for you. The more tenants (ie. leases) you have, the more beneficial this quickly becomes. For some of our larger customers with hundreds of tenants, this is a huge benefit!
On top of this we’ve added many other features. I don’t want to divulge everything yet, but another major new feature is the ability to create, store, and print receipts. This is a whole new section that many people requested. We’ve also added date formatting, some preference settings, etc.
As well, we’ve done a lot of very significant performance enhancements. Between version 2.12b and version 2.12c we initially did some amazing performance enhancements, as much as an order of magnitude faster for many screens (that’s 10x faster performance!). Well in this version we did another order of magnitude in performance for most data and list screens! We ran a test database with over 2000 tenants, over 2000 units, over 5000 workorders, over 5000 receipts and invoices, over 50,000 accounting entries, and all the related data. I can assure you that the larger your database, the more significant the performance enhancements! Some screens have improvements where the display is virtually instantenous for larger tables (under 1/2 second)! We’ve improved the speed of the UI (User Interface) and the database calls (which I’ll probably write about the later very shortly). All in all, there are some very significant performance enhancements in this upcoming version!
Now you’re probably still asking yourself where’s the dilemma? The dilemma is that we wanted to add two other new and very powerful features for this upcoming version. The problem is that neither of these features is going to be ready for at least 2 more months. So do we wait 2 months for these two features or do we go ahead and release now with all the great features and benefits we’ve already implemented?
You still don’t get the dilemma? The dilemma is that if we release now, we might now also want to release the other two great features in 2 months. Releasing major versions too often will annoy your customers! Nobody wants to be upgrading every other day. So if we release now, then when do we release the two other major features? In case you’re wondering, these are very highly requested features, so I believe they will have a significant impact on sales and customer happiness.
So we can postpone the release for 2-3 months, which means we lose revenue (opportunity cost) as with every new version, every new features, we increase our sales. If we release in 2-3 months, I personally believe that these two new features will generate additional significant sales by themselves! So we want those in sooner than later too. So if we release now, do we push those features to the next major release to avoid having too many upgrades and annoy our customers or do we release them in 2 months? Which do we do?
As well, each release has technical support costs. Although our upgrades are fairly easy to do, all you need to do is re-install the software overtop the old one (the database is automatically converted for you), many of our customers still require technical support. What you have to understand is that many of our customers aren’t all technically literate (we do offer the easiest property management software after all!), which means it will convert to some extra technical support costs for us. This adds to the equation in that more releases is more expensive.
So just to recap as this is fairly complex, here are our three choices:
1. Release today missing two major features. Re-release again in 2-3 months as another major upgrade (knowing that people don’t like too many new major releases, and that there will be additional technical support costs).
2. Release today missing two major features. Push the 2 major features into the next major release several months away, even possibly as long as a year (and lose the additional sales for that time).
3. Release in 2-3 months and withold all the great features we’ve already built. This also means we’ll lose additional sales during 2-3 months.
As an extra factor, September to April is our busiest time (they’re multiples of the rest of the year). So these 2-3 months are right in the middle of where we make most of our yearly revenue! This means that 2-3 months is actually like 6 months for most businesses!
Which would you choose? I personally opted for option 1. I’m willing to absorb the extra support costs that come with each upgrade because I think the additional revenue during our busy season will outweigh them. I also believe that we shouldn’t hold back our current features for these two other features, no matter how great they are! I don’t care if they’re pure gold, we already have so many highly requested features that I’d like to get them out there now rather than later. The other two features are important enough to me that I don’t want to wait until another major version, I’d rather release them as another upgrade in 2-3 months. They’re big enough to warrant an upgrade rather than a patch, but I’m not willing to wait for the next major release.
Which option would you take with these assumptions?
· November 9th, 2006 · 11:07 am · Permalink
I’d go with 1. If customer’s are “annoyed” by another major upgrade/release so soon, they themselves can wait (or are you “forcing” customers onto each major release?).
· November 9th, 2006 · 11:47 am · Permalink
If you make the upgrade a “one-click” process, where it will download remotely and then upgrade without unnecessary user intervention, users will be happy to upgrade as often as there’s a feature or bugfix they are interested in.
BTW, it’s spelled “dilemma”, and Firefox 2 has a built-in web spell checker, if you post from your browser.
· November 9th, 2006 · 12:24 pm · Permalink
Hi,
Actually, no offence, but I believe the dilemma is false. 😉
We’ve struggled with the very questions for years and I believe we have found a solution; we are now shifting to subscription-based model. It’s definitely a matter of what you like and what you don’t like but it *is* a solution that lets you avoid having to answer the kind of questions that you’ve stated.
To illustrate it, the users of our FollowUpXpert (https://www.xtreeme.com/followupxpert) buy a license along with a subscription for 1 year of free upgrades and support. After the year they can extend it by paying a percentage of the retail price ($39 instead of $125 for the Professional Edition). Of course, they can keep the software for however long they want but won’t be able to upgrade (with the exception of critical bug fixes).
It enables us to make releases as fast as possible without decreasing the revenue. The most important thing is it works!
We still have “major” and “minor” releases but “major” is a word reserved for releases that have extensive and “horizontal” impact (for instance, version 4.0 we now work on will let the user use a SQL database as a backend instead of the current proprietary forma) and “minor” – for smaller, “vertical” changes (e.g. adding better support for web-based sign-up forms etc.).
Just my $0.02.
· November 9th, 2006 · 3:31 pm · Permalink
Hello Steph,
Firstly, if your customers subscribed to your blog they would know what to expect and what concern you have for the value of your product and your client’s satisfaction.
You have solved part of your dilemma by having a great product that:
-Includes 1 year of upgrades/support.
-An easy one step upgrade function.
I would go with option 1, as you have decided.
You are getting your added benefits out now; increasing sales, customer satisfaction.
If you release it with a well worded primer about the next update (released before tax time, and free with your 1 year upgrade), the client can chose to install the new upgrade now or wait for the release.
It might also increase immediate sales, as the client might have been waiting for the “coming” features, but will buy now, as they know it is coming, will be free and in the meantime have the opportunity to familiarize themselves with the product, convert their systems etc..
Great subject and insight on the inner workings of your business…
Cheers,
Eric
· November 9th, 2006 · 10:34 pm · Permalink
Hi Jo,
That’s a good point, and many people actually do!
As for forcing the customer to update, no we absolutely don’t! If someone wants to stay on the last version their entitled to, we have absolutely no problems with that.
· November 9th, 2006 · 10:40 pm · Permalink
Hi pdq,
That’s actually one of the features we had to push back on for this version again. We’ve been wanting to add it since version 1.08, but we just haven’t had the time. It’s all based on priorities of what our customers want and need. I’m really hoping to push this into the next release with the other two major features…
In the past it wasn’t so much an issue since we only released 3 patches over a little over a year. But with version 2.12 we had 5 patches in 6 months. At least 1-2 of those patches could almost have been considered new versions! Anyways, assuming that we again want to release another major shortly, I think this is becoming a higher and higher priority feature.
And thank you for pointing out the spelling error. It’s appreciated. I’ll fix it in a moment. And yes, that’s a great feature in FireFox 2! I’m really glad they added it. I’m just downloaded this week, I’m still using 1.5 on the box I write though.
· November 9th, 2006 · 10:42 pm · Permalink
Fixed the “dilemma” spelling mistakes 🙂
· November 9th, 2006 · 10:51 pm · Permalink
Hi Martin,
Actually, that’s pretty much the same model we use, but it’s not officially called a “subscription model” (although in essence it is). When you purchase the software you receive a year of free upgrades, including all versions.
The problem with this model is that not everyone upgrades with each version. As well, not everyone wants to deal with upgrades. Upgrades can be scary!
For example, you talked about changing the backend from a proprietary format to an SQL database engine. Depending on the type of customers you have, and how critical their data is, such an upgrade can cause a lot of anxiety. I don’t know much about your application, so please excuse my ignorance if it’s not applicable, but with LandlordMax, some of customers are each dealing with up to ten’s of millions of dollars of accounting and management information. Some of our customers want to be absolutely sure everything works, and therefore are willing to push back upgrades, even if it has new amazing beneficial features…
For new customers though, this model is great! You don’t need to wait until a major cycle to release new features. And this is what I especially like about it.
As well, it takes a lot of the risk of purchasing out of the equation. If you don’t need to worry about when the next version is going to come out, it’s a lot easier to make that purchase now than wait until just after a major version becomes available! I know I personally prefer to purchase software with these types of licenses.
· November 9th, 2006 · 11:02 pm · Permalink
Hi Eric,
Thank you. And yes you’re right, but again, no matter how great your product is, some people will always be adverse to change, and I can’t blame them! I’d be a little risk adverse too in some of their cases.
And you’re absolutely right about increasing sales. Each new version we’ve ever released has increased our sales! I don’t think it’s so much that people are waiting for that one feature to buy, but that the total value of LandlordMax for them just keeps getting higher and higher. In other words, the more it does and help them, the more they want to buy it!
Something to note, more of a personal preference really, but I really don’t like to sell LandlordMax based on upcoming features. I know a lot of companies do this, but it’s not for us. I’ll only mention a feature is coming if it’s already built in our development environment. Otherwise it’s too easy to disapoint someone later. For example, if you tell a potential customer that feature X will be in the next version, and it has to be pushed back for whatever reason, then you’ve just annoyed them! Why do this? The most we’ll ever do is say a feature is “anticipated”, which means we’d like to have it, but we’re not promising.
I think Joel Spolsky says it best in his article: Mouth Wide Shut.