50 ways to measure your analytics (with apologies to Paul Simon)
"The problem is all inside your figures," she said to me.
"The answer is easy if you think more than numerically.
I'd like to help you in your struggle to count your impact perfectly.
There must be (at least) 50 ways to measure success for a news article."
She said, "It's really not my habit to really think about the news.
Furthermore, I hope my meaning won't be lost or misconstrued.
But I'll repeat myself, at the risk of being crude:
There must be 50 ways to assess whether your piece is reaching the full potential audience it could.
Fifty ways to count your numbers.'"
The basics: It’s all about the views, Hugh
- Number of pageviews.
- Number of pageviews relative to the average number of pageviews for similar types of content. Is this piece performing about average relative to pieces like it?
- The number of visitors to a page.
- The percent or number of people who read more than X articles within a publication in a time period.
- The percent or number of people who click on X additional page through a specific page. Which articles lead people to more articles?
- Total time spent on page.
- Total uninterrupted time spent on page.
- Total time spent on page in last X number of minutes.
- Percentage of total page viewed.
- Which authors are performing the best? Which sections? Which authors and sections relative to how they typically perform?
- Number of referrers for a particular piece.
- Are visitors from a certain referral taking a certain type of action? Are viewers coming to your page from X publication more or less likely to subscribe or take another action?
Social media and sharing: It’s how you share, Claire
- Number of total shares.
- Number of shares by people who have never shared anything from your organization before.
- Number of total likes.
- Number of likes from people who don’t like your brand on a particular platform.
- Number of shares from other organizations or people outside your organization.
- Rate of growth of sharing. Did traffic pick up after it was shared by a particular individual or on a particular platform?
- Particular groups sharing a specific piece.
- Number of citations in academic articles.
- Number of citations in other news articles (or by general public).
- Percent of people who take an action (subscribe, follow, purchase) after seeing something on social media.
- Number of publications that pick up your piece (if licensed to be widely distributed).
- Number of organizations that fork or build upon on your project.
- Number of organizations that use your project (if it’s an open source project that can be used by other organizations).
- Number of laws passed as a result of the piece being published (or ways that the status quo is now changed because of the piece).
Engagement and action: Think about the action, Jackson
- Number of people who purchase something through an offer.
- Number of people who sign up for a newsletter.
- Number of people who open the newsletter and/or click on something in the newsletter.
- Percent of people who have read more than X number of newsletters. Did they get bonus material?
- Number of people who subscribe or sign up for a tiered level of access.
- Number of people who sign up for push notifications.
- Number of total comments on your publication.
- Number of total comments in other forums (for example, letters to their representatives).
- Percentage of comments that are from new commenters vs. previous commenters.
- Number of people who now know more about topic X than they did before engaging with the content.
- Number of people who now behave differently now that they engaged with the content.
- Number of people who now feel differently about a certain topic now that they’ve engaged with the topic or piece.
- Number of people who save the piece of content and/or clip it out.
For audio and video: It’s how you hear and see, Lee.
- Number of iTunes [and other platform] downloads.
- Number of people who actually then listened to an episode (not everyone who downloads ends up listening).
- Total listening time.
- Total uninterrupted listening time.
- How much time passes between someone listening to their first episode and their second? Is there a way to shorten this time?
- Average listening time per episode, which is a good way to find out if your episodes are too short or could perhaps be longer.
- Percentage of people who listen to at least X number of episodes.
- Percentage of people who only listen to one episode.
- Percentage of people who listen to a suggested podcast mentioned within an original podcast.
- Number of new listeners or viewers for a particular episode. Is your growth rate stagnant?
- Number of responses to an audio or video in that medium (or perhaps a different medium).
Thank you to WNYC’s Elaine Chen for brainstorming with me on this! Thanks to Paul Simon for his entire oeuvre.
Editor's note: You can take our course on digital metrics here.