"Features must align with business goals, because they create the theoretical framework in which the application operates; the alternative is an aimless application with a lot of code, a lot of options, and a lot of usability problems."

I recently done two tech talks about using MongoDB for Rails projects, one at Coimbra.rb, and another one at the LinkedCare offices. Here are the slides:

"As you build your startup, you are digging a hole. If it doesn’t get easier to get new users, take the time to get back to the surface and dig a brand new hole instead of simply trying to correct course. The right path might not be far. Once you find it, the entire universe will conspire in helping you to succeed and you will feel it."
Get out of the rabbit hole by Xavier Damman
"[A MVP] Should validate or invalidate key hypothesis: Rarely is the most critical hypothesis - can we build X. Remember, most startups fail because no one wants what they built not because they couldn’t build it."
"They did not want to hear that, but it made things clear: we had to build a better product. There was no other way out. No window, no hole, no escape hatch, no backdoor."

A recurrent story, when you start reading about retention in, is the story of how Twitter manages to keep new users around:

Elman, whose goal is to make Twitter users stick around, said his team has identified an “aha moment” when a casual user turns into an “active user.” That moment happens when users follow 30 accounts, and when one-third of the people they follow also follow them back.

from Making Sign-Ups More Complicated Is a Good Thing, and Other Lessons From Twitter’s User Retention Efforts

That inspired me to find out what would be Bundlr's retention threshold. We already knew that, for new users, clipping something had the best correlation with coming back to the service (compared with creating a bundle or following a user or a bundle). But how many clips does it take? Is there a correlation with the number of clips created?

So, I made a quick unscientific study with data from new users in the last three months, which resulted in the following graph:


Disregard the last spikes, we don’t have that many users with more than 40 clips on their first day.

It’s noticeable a sharp climb until the 10 clips (~55%) and then it just hovers randomly around 60%. So, the best we can do to keep our users around is help them make 10 clips!

Of course, it’s not that straightforward. There could be a lot of other variables in play. But it’s a good number to keep in our heads while tuning Bundlr’s new user experience.

Do you know more retention thresholds? Share them in the comments.

A more recent personal project has been a new blog called Coimbriefing. I use it to share quick updates about the Coimbra startup scene (in Portuguese only), to keep the community and my close network informed of what is happening.

Fell free to follow it and send new tips my way.

In the last months, we’ve done some projects outside Bundlr. One of those was just released: Androplr, an Android app for the popular sharing service Droplr.

We worked together with Michael Hart and it was a great project to brush up our Android skills. If you’re a Droplr user with an Android, give it a try. If you’re not using Droplr, you’re missing out on a really useful tool.

My first Barcamp, back in 2006, was sort of life changing. So, I’m very excited to help organize another edition, here in Coimbra.



It’s already this weekend. Join us.

This weekend, me and some fine gentleman/women will be teaching Rails, on an event called Rails Girls. It’s a great initiative. Know more at: