Fail fast: ✓

Experimenting and learning from failure

Posted by Claudio Sabia on

On new year’s eve we launched our first experiment, trying to gauge interest for a new javascript UI library targeted specifically at backoffice web applications. After a week, a few runs on Hacker News, Reddit and LinkedIn, and a brief review of the website analytics, we now have a fairly clear idea of what failure looks like. With an handful of upvotes and hardly any expressions of interest out of ~600 visits, it is fair to say that our proposal is not gaining traction.

###Experimentation and learning from failure

We’ve come up with three potential reasons for the lack of interest we witnessed:

  • the way we are presenting it is not effective in showing Popsicle value
  • we are not reaching the right people
  • there is no interest for such a product

###Ineffective presentation? Our initial blog post introducing Popsicle was very high level and did not go very deep in trying to clarify the issues that we have with what is already out there; we also did not go into any detail regarding the design choices we have in mind. The website was trying to sell a product rather than an idea, and we did not put much emphasis on the engineering strengths of our idea. This might have been ineffective in explaining the value of our initiative and at gathering feedback. Luckily for us, this is easy and inexpensive to verify: we’ll try again with a less snazzy and more technical blog post, and see if this works better, worse, or as poorly as the previous one.

###Not reaching the right people? It might be that we’re not addressing the right people. This too should be easy to verify, diversifying our targets and comparing the results we get for different audiences. We plan to do that later on.

###Not an interesting product? It might just be that whatever libraries people are using are already “good enough” and there just is not much interest for something amazing as the one we have in mind. We have no direct way to verify that, and it probably isn’t so important. Ultimately we’re trying to be lean and this is one of many experiments we are free to conduct here at OptionFactory. What we can do then is fork out a new experiment, try a completely different idea and then choose which one to pursue based on the results we get back. We’ll try and build ourselves another option.

###What now? We’ve come up with a new blog post detailing the rationale behind Popsicle. You can read it here.

In the meantime, we’ve also released our Google Drive - Slack integration tool and service. It is open for good and free (as in beer) for the time being.

Finally we’ve begun visually tracking our path in the option building space, soon we’ll make our work publicly available. Let us know what you think.