In our, Selling software: Where have all the business models gone?, we looked at the problem faced by traditional PC indpendent software developers and explored why the new mobile app store model might not be enough to guarantee survival. In this article, we'll take a look at the cloud-based freemium model and explore why it might actually be a winning solution for struggling PC ISVs.
Let's look at the freemium software-as-a-service business model in some depth. The idea of freemium is that you get to use a product for free, but if you want to take advantage of additional features, you have to purchase the service.
I'm starting to believe the best replacement model for this type of developer is the SaaS freemium model.
In some ways, this is almost like shareware. Back in the day, shareware developers sold software that you'd install on your PC (remember when you ran installers?) and if you liked it, you send money to the developer.
Over time, pure shareware morphed into a model where developers would either time- or feature-limit the product, and unlock individual instances of the software with a serial number or other key mechanism. They essentially distributed demos that could be unlocked into full versions.
This shareware-derived download-with-unlocking-key model exists to this day and is the predominant software product model for both PCs and Macs. Yes, both Apple and Microsoft offer a half-hearted attempt at an app store for their desktop and laptop operating systems, but most serious developers still sell their products the way they've done so since the early days of the Internet.
The benefits of this model compared to the packaged model are obvious: there's no packaging cost. There is a piracy problem, but that's common to almost all software product business models (perhaps less so for the SaaS model I'm about to discuss).
Additionally, one serious problem that developers have with app stores is that it's very difficult to reach out to customers for additional sales, cross-selling, or upgrade sales. The app store owns the customer contact information, not necessarily the developer. This limitation, while not universal across all app stores, makes it difficult to build ongoing relationships with happy customers, nurture them, and sell them new offerings.
The download-with-unlocking-key model is still almost ideal compared to the mobile app store model and the packaged product model, in that prices are still at a per-unit price that comes with support but there's no physical cost of goods sold.
The only real downside to the download-with-unlocking-key model is this: the PC market is tanking. With a reduced size of market, the model may work, but there are a lot fewer customers.
I'm starting to believe the best replacement model for this type of developer is the SaaS freemium model.
Like with the old shareware/demo model, freemium cloud products provide a set of features that prospective customers can use for free. Freemium products are rarely time-limited (or if they are, it's the premium functions that are time-limited, not the entire offering). Instead, most freemium products give their users the chance to upgrade to a premium service that adds features and functionality to the product.
As an example, let's take a look at two freemium SaaS products I bought premium upgrades to, and another that I'm still considering.
Let's start with Evernote. Evernote is, at its most basic, a note-taking and database application. If you were running this thing on a PC in the 1990s, you'd have a UI and a backend database, along with a search capability.
Of course, because Evernote is online, your notes are available everywhere. Evernote starts as a free service. The basic service allows you to upload 60MB worth of notes or documents per month, while the premium service raises that to 1GB. The premium service also allows you to search inside PDFs, scans and OCRs images faster, provides a note history, allows you to share notes, lets you create offline notebooks, provides PIN locks for mobile, devices, and provides what Evernote calls "high quality" support.
Evernote Premium is billed at $5/month or $45/year. If you think about it, $45 is about what you'd pay for a basic packaged software product back in the day. I'll come back to this in a minute. It's an important point.
Let's move on to Todoist. I recently moved to Todoist because I like how it easily links to and assigns to-do items from Gmail. It's extremely smooth and it's had a positive impact on my productivity.
I signed up for the free version of Todoist and, within a day, was disappointed. I couldn't create notes in my to-dos; that was a premium feature. While I didn't mind buying the premium service, I really wanted to decide if the product would do what I wanted before taking the plunge.
While adding notes to my to-dos was the feature I noticed was missing, Todoist offers a wide range of additional premium features, including the ability to text you reminders for upcoming to-do items, the ability to share to-do items, and more colors and tags for categorization, file attachments, custom search filters, and a variety of fancy reminder mechanisms, including location-based reminders.
Todoist doesn't have a monthly plan, but the yearly premium plan is $29. I liked it enough to take the plunge, signed up a few weeks ago, and I don't regret it. On the next page, I'll tell you the gimmick Todoist used to help convince me to sign up.
Again, note that $29 is right in the range of what the old shareware and packaged products generally cost for this category of software.
The third product I'm using for the purpose of this discussion is Pocket. Pocket is a read-anywhere Web reader, where you send articles you want to read to Pocket and you can read them later from any device or browser connection.
I have found Pocket to be a wonderful way to keep up with my reading. I'm no longer chained to the pile of open browser tabs I used to create for my morning reading. Instead, I send the articles to Pocket and I can read them at any time -- although I still allocate morning reading time to keep up with the important stuff.
In any case, Pocket Premium is almost exactly the same price as Evernote Premium: $4.99 a month or $44.99 a year (a penny difference). I'm guessing this is no accident, because one of the premium services Pocket offers is to archive articles you've read in perpetuity -- just like Evernote. Pocket Premium also offers better searching than the free version of Pocket.
The reason I've hesitated to upgrade to Pocket Premium is that I've built my article archive in Evernote and haven't decided if I want to duplicate portions of it in Pocket. That said, I'll probably upgrade to the Premium version because searching for articles I've read is so important to my daily activity that Pocket can serve as a backup in case I didn't store something in Evernote.
Here we have three cloud-based products that replicate some of what used to be done in basic PC utility or productivity software, for just about the same price.
There are differences in the model, of course. First, all three products I've mentioned have excellent mobile apps, which means that (a) the developers had to fund development on more than one platform and (b) users can choose to use their products no matter what platform they're using.
For me, this is huge. I jump between a PC and Mac ten or more times a day, I use an iPhone, an iPad, a Nexus 7 and my Samsung Galaxy S4 pretty much interchangeably, and I want my data and tools on any device I use.
Can you imagine, these days, downloading your email via POP3 and just having it live on one computer? Stone Age, right?
I need everything on every device I use. My calendar (Google Calendar), my email (Gmail), my to-do list (Todoist), my document library (Evernote) and the articles I need to read (Pocket) are all available to me at all times, from anywhere, on any device.
Evernote, Todoist, and Pocket stand out because they support pretty much all the platforms that users might use. You'll note that I haven't mentioned cloud-based SaaS products that don't support all my platforms. I don't use any. Even QuickBooks online supports iOS and Android devices.
Next up: the business model for SaaS freemium...
That brings us back to the business model for SaaS freemium. Development costs are higher than either the traditional packaged or demo/shareware model. They're higher than the app store model, too. SaaS freemium vendors need to develop a cloud application as well as apps for mobile platforms and probably desktop platforms as well.
Instead of one or two binaries, they need to create a cloud infrastructure, and at least four unique binary distributable apps and applications. That's costly.
Free is a great way to beat the cost of promotion.
The second cost item is the infrastructure load serving the free users. Although the ratio varies wildly, consider yourself lucky if one percent of your free users convert to premium. That means (best case) each paid customer needs to support the infrastructure for 100 free users.
While I store a ton of data in Evernote and I might do the same if I choose to upgrade to Pocket, I store very little actual data in Todoist. Some users might use a lot of attachment data, but to-do list storage infrastructure for 100 users from my $29 is probably not very costly. By contrast, the Evernote and Pocket IT folks need to manage storage utilization very carefully, because free users are still using a lot of space.
That said, Evernote limits free users to uploading less than a gigabyte a year. Using Amazon Web Services as a pricing model, that means a free user costs Evernote two-to-three cents a year in storage capacity. My $45 premium payment allows Evernote to support something on the order of 1,500 free users.
Conversions are essential, and this is where many of the dot-com bubble companies went wrong. Evernote, Pocket, Todoist, and the other companies like them need to be sure that enough of a percentage of their free customers convert.
I complained earlier about the fact that the free version of Todoist didn't support notes in to-dos. About three days after I signed up for the free service, I got a note from the company that they were upgrading me to premium for free -- for a month. Using the drug dealer analogy, they gave me my first taste for free. I liked it, and I didn't want to lose the premium services, so I bought. Todoist got their conversion.
Free is a great way to beat the cost of promotion. People are often very willing to try out a free service. As a result, rather than incurring a lot of marketing expenses, companies can simply offer their free version, and they increase discoverability. This is why in-app purchases in mobile apps generate so much more revenue than paid-for apps. The free offering gets customers in the door and the upgrades are what get them to pony up the cash.
While there are millions of apps in the Google Play and iOS app stores, there are not nearly as many in the Chrome and Firefox Web store equivalents. As a result, SaaS freemium discoverability stands a much higher chance of success with the addition of an app or simple extension. There's also enough money in SaaS premium to support a sales team. Discoverability isn't the only way you can reach customers. You can actually afford to call on likely prospects.
Price point is interesting as well. The price points for Web app premium services are much more in line with old school PC software prices than apps. Take Todoist, for example. A to-do app in the iPhone app store might sell for a buck or, best case, three dollars. But as a Web app premium upgrade, Todoist is selling their service for 10 times that amount.
Pricing in the double-digit dollars allows for providing customer support as well. One of the benefits of the freemium model is you can crowdsource your support of the free version (which will generally also support some premium customers) to forums and set up traditional helpdesk support services for the smaller percentage of paying customers.
There's one other very tangible advantage to the premium service model over both the app store model and traditional PC sales: subscriptions are annuities. The old PC product business relied on cranking out upgrades, and convincing users to both pay for the upgrade and install the upgrade. Without a regular upgrade business, software companies were generally unsustainable.
But with SaaS, you're selling a monthly or yearly service that can renew automatically. Granted, you're likely to have a measurable number of cancellations, but you're likely to have a far larger set of automatic "what the heck" renewals than you would convincing people to place new orders for upgrades.
The SaaS freemium model makes even more sense now because of the availability of on-demand infrastructure like Amazon Web Services. Even five years ago, the availability of very cheap, very scalable, pay-as-you-need infrastructure wasn't around.
Developers back in the day would have had to build out their own infrastructure, and the load on their gear would never have matched the buying patterns of customers. This meant that big, chunky capital expenses often had to happen at the least convenient times.
AWS and services like it completely eliminate that problem. It's entirely possible to scale load in direct correlation to user demands, and as a result, the key is pricing your service to support the extra load of those free users.
Let's compare these models now that we're looking at all of them. We'll take the table we introduced in the first article and expand on it here
|Packaged Products||Apps||SaaS Freemium|
|Development cost||Yes||Yes||Higher, must support all platforms|
|Physical production costs||Yes||
|Physical shipping costs||Yes||
|Risk of over-or-under producing inventory||Yes||
|Internet infrastructure operational costs||Just a support and sales Web page||
Sometimes back-end services
|Extensive, but can scale smoothly on-demand|
|Per-unit-sale revenue||Much higher||
Comparable to packaged products
|Per-unit-sales revenue must support large number of free users||No||
|Unit sales revenue helps cover tech support cost||Generally||
Not even possible
|Percent of selling price received by software vendor||30-40%||
100% unless using VARs to do value-add distribution
|Ongoing sales opportunity||Must produce and sell upgrades||
Sometimes, but often free upgrades unless new product introduced
This is a repeat-billing annuity business model
|Time to payment||90+ days or longer||
30 days or so
Credit card billing or individual corporate contracts
|Ability to reach total prospective customer base with marketing||Expensive, but possible||
Ongoing as part of the use of the product
On the surface, it does look like the SaaS freemium model is the winner. But here's one key caution: the formula of exactly what you offer between free and premium has to be perfect or you may not generate conversions.
Finally, there's one additional benefit I've been seeing in the SaaS freemium business model: business-class plans. Both Evernote and Todoist offer free, premium, and business-class plans. The business plans add additional management and collaboration features, raise the revenue basis higher, and provide an additional source of scalable income.
My key premise, that freemium SaaS may be a more valid replacement business model for packaged software products than apps boils down to a few key items:
- There's a built-in annuity for repeat sales that doesn't easily exist in the other models
- The price-per-user supports support
- Discoverability is easier
- There can be an upgrade path to business-class services that scale
- You may have to develop apps for all the appropriate platforms, not just one
- You better define your free/premium mix perfectly or you may not get conversions
Beyond what I've discussed here, there are many other freemium SaaS business models with a wide array of prices, upsells, and options. What I chose to focus on here were those that were price-analogous to more traditional PC software pricing.
Keep in mind that the secret sauce isn’t prices similar to old PC software. The secret sauce is finding that exact balance of value and price that will encourage conversion from users-as-freeloaders to paying customers. Finding this sweet spot is not an easy task, but it must be central to your entire product pricing strategy.
Today, if I were starting a software company or building a product, I'd go the freemium SaaS model. I don't see any other model that's particularly startup or bootstrap-friendly that's not a complete shot in the dark in terms of predictability of success.
And no, I'm not going off and starting a software company. I ran a software company for years. Been there, have the T-shirts. But if you do start a freemium SaaS business, let me know how close my analysis is to your experience.