09/14/2012 04:15 pm ET Updated Nov 13, 2012

Data Trumps Gut: Introducing Product Analytics

It's my experience that people who appear to be generating product ideas from thin air are actually great listeners who read widely, talk to people in their industry, consume their competitor's sites and apps, and dig into the data on site traffic, viral trends, comment numbers, and other metrics on a regular basis.

They have informed opinions about what new site features or changes to existing ones will have an impact upon those metrics, and they're keen to measure their success.

I'm a researcher and a consumer. In my first job out of college (2006), I confounded my bosses and co-workers by insisting that it was an important part of of my work day to read industry news.

I'm a networker. I have that quintessentially Midwest 'way' of talking to strangers. I enjoy hearing people's opinions about the sites and apps they use every day. I figure most people have an area of expertise and that gives them a different slant than me.

I'm an analyst too, though I didn't mean to be (and I think it still surprises my parents). I joined the Digital Analytics Association in 2006 to learn best practices and found myself on the ground floor of an effort to establish social media measurement standards across the industry.

The same techniques I learned then, I apply today. I'm now a data junkie. I go to the data to test my gut, to debunk my own assumptions as opposed to confirm them.

At the end of the day, it doesn't pay to build anything, no matter how cool it seems or how 'right' you feel, unless it's backed up by numbers.

If you do not keep goals and metrics in mind, development becomes an endless cycle of aimless iterations and your team's efforts are expended on trifles. What you launch will be disconnected from everything else. The goals of the business will not be met. You have to prioritize.

This is how we do it.

1. We look at the data during planning stages. Here, we'll look at historical performance of an existing feature, such as the right rail module(s), towards identifying ways of improving performance, and/or redesigning the module(s), and/or to inform best practices for use of a new or existing tool by our Editors.

2. Once we've decided we want to work on a product, we identify KPIs (Key Performance Indicators) so that we'll know whether we've been successful after launch. KPI's are basically goals with metrics attached, and they differ from product to product.

For example, our moderation tools are aimed at increasing our moderators' efficiency, making sure that the team's decisions are aligned with each other, and decreasing wait-time for comments posted to the site. Other products, like our 'right rail' are aimed at increasing time on site.

Understanding the goals for a product is a bit of an art, as is identifying the metrics that will 'prove' each goal we've identified has or has not been met.

3. At the launch of a new product, we monitor performance against our KPIs to ensure that the correct tracking methods are in place and documented correctly.

4. Post launch, we continue monitoring and recommend alterations to underlying logic and design based upon performance and how well we're meeting our goals.

As someone who has worked in this field for a while, I think it's important that best practices like this be shared. So often, it is not clear to users or even to people who work in media and tech as to how or why we're making the changes we do.

With that in mind, I want to introduce you to Jenna Green, one of the newest members of my team. She is our resident Products Analyst, and will be blogging about all things data at HuffPost.

A little about Jenna's background. She comes from the MA program in Statistical Analysis at Columbia University where she studied data mining methodologies, probability, and linear regression models among other things. She is kind of a rockstar, but she'll say she's "proficient" in SAS programming, R, Maple, MySQL, and C++.

In her first post, she's pulling back the curtain on a certain famous (infamous?) page. If you've ever wondered why sideboobs were worthy of their own section...well, I'll let her explain.