Heading home after a great weekend at #bcniphilly


Defining new metrics for journalism

Greg Linch led a session in the morning at BCNI Philadelphia about how we can define new metrics for journalism. It was prompted by the similar post he wrote earlier this year.

To help we need to look at the metrics other industries are using to define success in their work. Greg’s idea stems from Alexis Madrigal’s notion that you have to define your own metrics. You can find influence from others but don’t just adopt their method.

We have methods for measuring how many people see our work and how long they spend on a site. But, we don’t have solid metrics for measuring the impact of journalism. As Greg said, “What does a pageview mean?” The metrics we currently have mean things. They tell us how many, how long, and other aspects of our work’s impact. But they are largely one-dimensional and don’t tell the whole story.

Current approaches

Altmetrics.org was mentioned as one different approach. As their site says:

Altmetrics expand our view of what impact looks like, but also of what’s making the impact.

Sabermetrics was the focus of the Moneyball movie and is a similar approach in sports to redefining the metrics of an industry. Things like VORP and OPS bring new light to the performance of athletes who may appear successful by traditional metrics.

Propensity score matching is yet another approach. Tries to answer the questions of, “What’s the possibility of X to happen based on Y?”

Greg also metioned the way the Cleveland Orchestra measures performance. They strive toward superior performance and measure it by standing ovations, ticket demand, and more.

Someone in the audience mentioned the way Gawker rotates staff. Nieman recently wrote a bit about this. For a different way of rotating staff check out how The Economist does it.


Someone mentioned that one way to approach this is asking readers the question “Do you understand this issue better after reading this article?”

I brought up the way KISSmetrics approaches site stats from a user-specific perspective. For journalism that could be looking at how many people read the first story your publish on an issue and all of the followups. You could also measure how many jump from one report you’ve published to another one on a different topic.

Greg mentioned the net promoter score that many organizations use to measure the loyalty and reputation of a company or a product. One person in the room works at an organization which uses this on a per-brand basis. They separate the content creation from the content display. Each coverage team is given a net promoter score as well as each product or brand their content is flowing through.

The concern over not being able to measure impact in communities that are not online was also raised. Ultimately the way to fix this is to get out in the communities and talk with people about whether they understand issues and how they’d like to learn more.

Albert Sun said that part of why metrics are so confusing is because they are traditionally geared toward what advertisers want. The predominant approach is, “All metrics are good but we need to make money off of them.” The business side is failing, though, and what we need to do is create a new, better model. What we need is a way to measure the difference and change our journalism causes in the world.

If anyone has more to add feel free to leave your own notes as a comment.

Red eye flight to #bcniphilly via Houston


Looking forward to BCNI this year. Should be a great weekend.

Confessions of an Ivy League Frat Boy

Confessions of an Ivy League Frat Boy: Inside Dartmouth’s Hazing Abuses. Particularly this quote:

The college has also charged 27 other members of SAE, stemming from events in the 2011 pledge term. While the other students all categorically deny doing anything illegal, the information that Lohse provided to Dartmouth officials may directly implicate him in hazing. As a result, Lohse – the only student to come forward voluntarily – may be the only student who is ultimately punished.

Also, see “Allegations of hazing leveled against TKE initiation practices” and Daniel’s post. I have an idea for an enterprising reporter: take a deep look at fraternity abuse reports like this and answer:

  • What percentage are followed up on by news organizations, particularly college newspapers, after the initial report?
  • What percentage result in concrete action undertaken by college administrations?
  • In how many cases is the student who reported the offense the one who takes the brunt of post-publication attacks?
  • How frequently are reports the second, third, etc. time allegations have been made against a specific fraternity?

There are more questions that would be interesting but the above would be a start.

Install Happiness

Install Happiness

Someone put this sticker outside the door to PIE. I like it.

Fantastic last-minute seats to tonight’s Blazer game


Hard to turn down free last-minute tickets to the last Blazer home game of the season. Thanks Andrew Witherspoon!

Hollywood Theatre at night

Hollywood Theatre at night

I snapped this a couple weeks ago when I saw Jiro Dreams of Sushi at Portland’s Hollywood Theatre. I didn’t realize it actually turned out well until flipping through my photos this morning.

The speed of support

When we help people in support we want to quickly solve whatever problems they face. There is an urge to get the answer sent as soon as we can. The faster reply is the better reply. But, that is not always the best way to make someone happy. For those people who are frustrated, shy, or unclear in their original description, you may end up solving the problem as you constructed it, not the problem they had.

This problem is exacerbated when you face a backlog of requests. Someone has already been waiting for days so you really, really want to get that answer to them. Now.

Sometimes, the best thing you can do is ask a question. Patience and care trump raw speed.

Taking a step back gives you both the opportunity to make sure you are on the same page. You show that you are there to help. You want to do more than send a response and move on. You want to take time, understand things, and then solve them. When you do this you work toward honest help, not just efficiency.

The less we assume the better we will be at making people happy. Good support is about answering the questions someone has. Ensuring that we properly understand those questions is the most important step toward that goal.