Two days ago our beta request site went live
Late Friday afternoon we soft-launched our teaser website and have been blown away by the interest we’ve seen so far. The teaser site was put together in just six days (from zero to live) to have a placeholder while in stealth mode. I’m proud of the passion and experience the team brought to deliver such quality in so litte time. It’s still very early but we thought we’d share some quick stats and lessons learned so far.
We’re super happy with the results so far. The big surprise is having more than 10% conversion from visitor to sign up, given the fact people had extra obstacles (we ask for Linkedin authorization). Also, folks have spent an average of about 2 minutes per visit, so they’re spending time with the story.
Productivity tools we used:
The teaser site started while we were literally around the world from each other (I was in Bali for my wedding). Being 15 hours and +1 day away from each other made for around-the-clock work. Not surprisingly, we used these same tools even when we ended up in the same office.
We’re fans (and friends) with the guys at Skitch, which is the ultimate visual grok machine– a super rapid way to asynchronously collaborate on the stuff on your screen. Just snap a pic of a part of your screen or drag a url into the app, annotate on top of it, and click “share.” Skitch then gives you a url you can share with anyone who can see just what you mean in seconds. Here’s a pic of one day of my Skitch history.
I had never used folders or even shared folders before in Evernote. My personal Evernote is just one massive dump of inspiration or forget-me-nots, so this was the first time I had used Evernote collaboratively. We had a number of shared folders set up to grab inspiration, work on the script, and pull a-lot of our formative assets.
File sharing and iteration really hit hard the last 3 days (out of 6) as the web assets were getting assembled and needed to be in one place. As we gave feedback or changes to Jason (our designer), he’d simply replace the file in Dropbox so we always had the most up-to-date stuff. Matt & team then used a script to automatically pull the latest and publish to the sandbox environment. From new graphic update to sandbox typically took < 1 minute. This allowed us to review and approve new updates, within the live site structure, instantaneously.
Matt & team had over 230 commits to GitHub over the 6 days it took to build the teaser site and integrate with our back-end engine over REST. Being this agile and leveraging a slew of new technologies required constant code review, merges, and sandbox pushes. As we aggressively hire engineers (ping Matt if you got what it takes), we will continue to leverage GitHub’s distributed collaborative infrastructure to organize, store, share, and iterate on our fast evolving code base.
We wanted a way to tell the story of two ways to land business, so we imagined two different downtowns: Goosechase (the old way) and New Business City (the new way). Jason’s original visual inspiration for the city came from Grand Theft Auto for iPad (I attached his first screenshot). He then found some awesome scrolling interaction sites like Ben the Bodyguard, We Bleed Design, and Bullet PR.
Vehicle-wise, we looked through Google at a number of cool steampunk-styled contraptions. Originally we were going to use more of a parade float for the “new business” side vehicle. For the final design, Jason fused a float with an iPhone.
Other sites with cool interaction design that inspired us:
As soon as we went live, a number of people reached out to help make our beta site better. One of them was our friend, Justin Kistner (who’s way smarter than we are when it comes to social media marketing).
He recommended a number of things:
- If we added the ability to comment on the FB like button we’ll get full posts with a picture and text (and folks see these posts more than others)
- We hadn’t thought through how to track twitter referrals from folks clicking links via the Twitter client.
- If we use campaign IDs on links from the Like button, our integration with LinkedIn could separate out how much traffic came from those sites because of links published from people seeing posts from those buttons vs. organic visits from those sites from people sharing without using our share buttons.
Reporting from here
We’ve made most of the recommended changes and will continue to respond to advice and ideas from those who want to help. You can bet that we’ll be sharing more official results once we’re further along. Thanks a ton to everyone for your feedback and support.