Diego Scataglini - Latest Comments

Re: Ode to procrastination – part 1

Diego Scataglini — Mon, 29 Nov 2010 15:24:51 -0000

A definition contributed by @deadprogram: getting ready to start thinking about thinking about starting to get ready.

Re: About

juanchaparro — Thu, 18 Nov 2010 00:41:08 -0000

Hey great meetup tonight in boca!...

Re: Freemium for you doesn’t mean looseium for me

Mike Deutsch — Thu, 23 Sep 2010 22:12:51 -0000

Ah, right. You have to test a/b simultaneously in order to control for external events.

I may be getting ahead of myself a bit with the pricing stuff, actually. I haven't a/b tested to optimize my ad keywords, landing page pitch, etc. I imagine it's important to get that down first, right? If you don't have the entry path totally solid, you're testing pricing on unstable ground.

This is all an adjustment for me; I'm trying to lean-ize what was initially a fully planned-out app. Thanks for the advice. :)

Re: Freemium for you doesn’t mean looseium for me

Diego Scataglini — Thu, 23 Sep 2010 09:34:35 -0000

One more thing when doing your analytics analysis try to disprove your theory versus confirming it.

We are all too clever in finding random datapoints that confirms our theories, we'll create narratives and then look for confirmation.

The right approach is negative empiricism (http://bit.ly/9I5Lkr). Trying to disprove your theory, while painful, will get you closer to the truth faster.

Re: Freemium for you doesn’t mean looseium for me

Diego Scataglini — Thu, 23 Sep 2010 09:14:14 -0000

Mike,
You can actually do both. If you have a website that offer a free trial, and you have enough traffic, I would suggest to a/b test the different options.

Show 50% of the traffic the current site, that will be your baseline, and to the other 50% show the site without a free signup and purchase only option and let the experiment run for a month.
If you have enough traffic/conversion, one month should be enough, otherwise it might take longer to get a real idea of what's going on.

If you don't a/b test it, it's very hard to tell if the increase in sales it's because one options is working actually better than the other. It could be that you finally crossed the tipping point or maybe there was some 3rd party influence. For Ex.:
Let's say you switch marketing model and somebody writes a kickass review of your site, let's say on stumble upon. You might think that your increase in sales is because of the switch in marketing while it wasn't.

So I'd go with a/b test the 2 or 3 models and look at the results keeping always an eye on the source of your traffic. You want to be aware of 3rd party influences on your experiment. I recently ran a 3-way a/b test on a site where one of the landing pages was shared through a social network and received twice as much traffic. I had to make adjustments to my analysis of the results.

Re: Freemium for you doesn’t mean looseium for me

Diego Scataglini — Thu, 23 Sep 2010 09:12:53 -0000

Putting it in your thread

Re: Freemium for you doesn’t mean looseium for me

Mike Deutsch — Wed, 22 Sep 2010 23:35:55 -0000

Diego, I'm really glad you cross-posted this to Ruben's comments. Very well put together post.

Reading Ruben and some of his commenters I had reached their conclusion that Free just does not work as a marketing on-ramp to Paid -- the conversions just aren't there historically; better to just get people to sign up as paid right off the bat. That's more or less in line with others like Dharmesh Shah (charge early, charge often). It's awfully tempting to see Ruben's conversion numbers jump and think "Whoah, I should do that."

Your counterpoint seems to be that Free *is* in fact an effective on-ramp; you just have to be mindful of what you're doing, consider which angle to take (there are many) and how to accurately evaluate the support load, and study the results closely enough to actually learn from them.

So after this back-and-forth, I'm not 100% sure whether I'm going to put my free option up front or tucked away on the side. Your approach seems to be: Free up front for starters, when the name of the game is eyeballs and finding "fit", and then evaluate what I've got. Am I getting that right?

Mike Deutsch

Re: What can you say about him?

oldfartdeveloper — Mon, 23 Aug 2010 18:56:06 -0000

All programmers aspire to be like him in every way.

Re: What can you say about him?

oldfartdeveloper — Mon, 23 Aug 2010 18:55:42 -0000

Cars automatically open their driver side doors and start their engines whenever he walks by them.

Re: What can you say about him?

oldfartdeveloper — Mon, 23 Aug 2010 18:54:54 -0000

Airlines adjust their itineraries to provide him the shortest trip whenever he thinks he might want to take a trip.