Throttle requests (maybe at the IP level) for the prototypes, but resell hosting. Redirect your third-level domains transparently for the hosted sites - make it dead easy, just a matter of clicking a box to "add public hosting". Then keep the prototype there as the testbed for new development, with a button to "Click to publish".
That will take you, what, a whole day to set up? It's a no-brainer.
We're building a fully hosted CMS, with redundant, load balanced hosting, custom data types, generic non-branded client interface and so on.
But during our beta period we've discovered that it's also just about the easiest way we know to get a quick prototype or experiment online.
Now while deciding on our pricing, we get into a dilemma. If we go with the typical per project pricing model of most other hosted CMSs, aimed at agencies, we wouldn't really make it viable to use Webpop for quick experiments, prototypes or mostly static pages.
So now we're pondering if we can come up with a set of plans that supports both ways of using Webpop...
The suggestions from others here could work for you, too. Maybe you could offer "Webpop Jr." with severely limited bandwidth or a banner at the top. It might even be advantageous to make Webpop Jr. look like its own domain/application to differentiate it from the main product.
That is funny! Perhaps you should direct your customers looking for a prototyping product to us and we should direct our customers to you when they want hosting. Something to think about.
Have you considered a separate "prototyping" product, with different branding and pricing? It could potentially reach a very complementary audience ...
webpop minus the CMS and with a prototyping slant actually ends up sounding very similar to quplo. Not sure what that means. Perhaps 3 years from now we'll both look back and realise we were making the same thing...
I actually always assumed that you guys had consciously come up with a brilliant way of creating a hosted CMS through a lean/customer-driven approach by focusing on a rapid prototyping tool first! :)
Interesting article - just tweeted it. While i wouldnt say i am part of the "lean startup cult" - i must say i really like their stance on Pivoting.
The Flickr example is a great one, smart companies, the truly smart ones you find always had to pivot a few times. It has certainly happened at the company i work for (Grasshopper).
We used to be called GotVMail - we realized that being voicemail focused wasnt the right strategy, and that really we were a communications solution for entrepreneurs. We bit the bullet, as pivoting is always taking one step back to go forward. But its been amazing for our business ever since.
This was a great post. Situations like this are a big challenge for any startup: on the one hand it's a huge opportunity, on the other hand you want to stay focus, and on the third hand you want to listen to your customers ... I really like the way they're approaching it with an open mind, working with Q42 and asking broadly for feedback.
And it's an intriguing tool, too. One of the other impressive things about this post was how well they spotlighted their customer Q42 and their customers' client bransomcompany while also showing off their own product to good advantage. I signed up, and I bet I'm not the only one :-)
This is how not to write a product plug in a disguise of insightful blog post. Way way too much marketing content, no concise summary at the top of what happened and too long in general. Feels forced and disingenuous :|
Sorry you feel that way, eps. It's our product's blog and we wrote it because we're genuinely facing this problem and wanted to share. There was no intentional "marketing content", it wasn't forced, and it certainly isn't disingenuous.
Interesting product... Time to run with new features and new uses. I'd love to be in the position of having users figure out new uses for a product of mine. What better way to have user-guided product development?
quplo: Your workflow for managing projects is quite excellent, signed up to try it out!
plug: You all would probably also be interested in my Visual CSS3 Animation Designer, available for prototype preview @ http://editroom.splatcollision.com/
*It's hosted on a single dyno (free) heroku plan, so forgive any slow load times...
There's obviously a market for hosted services, from barebones shared hosting services like GoDaddy to VPS hosting like Linode and dedicated hosting.
There's also a market for specialized hosting like point-and click design with Squarespace and Ruby on Rails hosting with Heroku (which recently sold for >$200 million to Salesforce I might add).
My advice? Run with the idea. The barrier to entry is stupendously low. Resell a VPS or even EC2 or even get your own dedicated server. If it doesn't work out, no loss. If it does, it's relatively easy to scale up to more cost effective solutions.
You may want to make your preview sites not usable as free hosting however. Throttling requests, putting a banner at the top, limiting requests, something like that.
One simple differentiator would be to require free preview sites to be hosted under the quplo domain. That also lets previews get up more quickly, because the user doesn't have to get the domain, configure DNS, etc.
That will take you, what, a whole day to set up? It's a no-brainer.