The first thing most customers ask themselves is whether a product can solve their problem. The second thing most customers ask is what will it cost.
This second question poses a problem for freemium apps. It is fastest and easiest to get the free part of the product done first. It’s the part that people will try first, it’s the basis for the product. If people don’t use the free version, after all, they will never pay for the premium version. The problem though is that if we don’t set up the premium version right away then we never know if people will pay. So do we start with free and go freemium later, or do we start freemium from the beginning?
Assuming that time is our limiting factor , by focusing first on the free product it allows us to perfect and iterate on it specifically, making the best possible free product that attracts people to use it. With more users we can then query the free customers more easily to figure out which features are most important for the premium product.
It seems Glassboard took this approach. The company introduced its paid version long after having a free version. But this apparently caused customer confusion and concern. Brent Simmons, who helped found the company, said the following in a blog post:
It’s been hard to explain some of the text on the Glassboard home page:
Unlike Google, Facebook, and Twitter, we don’t mine your data to sell advertising. We don’t do ads.
You are our customer — not advertisers.
People would point out correctly that the app is free and we’re not asking for any money.
That kind of customer confusion is suboptimal and likely kept some customers from creating an account and trying the service.
On the other hand, we can release the free and paid version at the same time. In this case we can immediately start with a critical piece of data — will people pay? — in the very first group (cohort) of users. Evernote, I believe, took this approach. This, too, has problems. Again with time as the limiting factor, this means our development and testing attention is split between a longer and longer list of features. On top of that, figuring out where to draw the line between free and paid may take a number of conversations with users who have used the service, which we don’t have yet.
Maybe both are optimal positions and work best for different kinds of businesses. Clayton Christiansen, in The Innovator’s Solution (Amazon, Powell’s), talks about three different kinds of market disruption: new market, low-end and sustaining. New market disruption assumes we are competing against non-consumption; low-end assumes we are addressing a new market with a cheaper alternative; and sustaining assumes we are bringing a better product into an established market. 
The iPhone, as Horace Dediu would argue, is a new market disruption. At the time the iPhone shipped only 1% of all mobile phones were smartphones. The iPhone was intended to convert the 99%. Android, however, shipped as a low-end disruption. It was specifically intended to play the low-end counterpoint to Apple’s, RIM’s, Microsoft’s and Nokia’s higher priced offerings. I can’t think of a sustaining example in the mobile world , so I’ll go with the appeal of Netflix when it launched. Like Blockbuster but more convenient with a better selection.
For each of these business disruptors, it seems, maybe a different approach to freemium needs to be put into play. With a sustaining or low-end disruption, the basis for competition is well known. Customers inherently understand the product with a simple comparison. Android’s like iPhone but cheaper. Netflix is like Blockbuster but better. In these cases the customer inherently understands the product and just needs to be convinced to spend money on the alternative. Thus, having both free and paid at the beginning is optimal.
But new products require education, and when education is required the tricky part is getting anyone to even try the product. Getting any users, let alone paid ones, is the axis to optimize on. And thus, in this case, there is no need to worry about premium yet. Start with free first.
 This is from the old axiom about how you can’t optimize on all three axes — time, money and people — only on two. I believe almost all new products are limited by time. If you have funding then your time is limited by the amount of money you’ve raised and how long it can keep your team together. If the new product is developed out of revenues, then the time available to work on the product is limited since attention must be paid to the projects that pay the bills.
 Page 44
 Maybe the move from 3G to LTE but that’s pure technology.