Archive for April, 2013

Since my last blog post, things have been moving quickly over here. Hack Week was invigorating (i.e., hectic) as usual, and while I didn’t get to demo my hack because I was out sick, I did make a ton of progress on building a reader that’s based on what I’m now calling the Shaun Appreciation Score –  a measure of a story’s Average Engaged Time, average scroll depth, and content height. 

Measuring appreciation

The Shaun Appreciation Score is my attempt to measure how much time users are spending within the actual content – the story itself, and not just on the page. I calculated my score by sampling our existing data from thousands of articles across our publishers sites*, as well as writing a system to get new data that we currently don’t have available to us. This includes scraping content from these pages, determining where the content starts and stops, and then figuring out how much time it takes to consume that content. 

Once I collected this massive set of data, I then chatted with Josh, one of our awesome Chartbeat data scientists. He suggested that I start plotting some of this data to get a feel for what the distribution might look like. So I calculated the mean and standard deviation of all the key data points I wanted to measure: average scroll depth, Average Engaged Time, and content height (the physical length of a story on a page).

The beauty of a bell curve

After these calculations, I plotted this data and seeing if there was any correlation between Engaged Time and scroll depth. As it turns out, there is a strong correlation(!). Content that has above-average Engaged Time generally has higher average scroll depth as well, but that’s not the whole picture. I wanted to see how much time was spent within the content, not in the comments section at the bottom, or in the slideshow at the top of the page. There wasn’t an easy way to determine this, so I decided taking the Engaged Time relative to the content length would help weed out articles with fewer words, or users who spend a ton of time in the comments section.

Plotting the data in a histogram looked something like this:


This bell curve is good news, I got something that resembles a normal distribution, which tells me that my data quality is good, and that there is in fact a difference between what I’m calling “good” content appreciation and “bad” content appreciation.

More to do next Hack Week

While I didn’t finish building my reader within this past Hack Week, I did get to a point where I’m feeling pretty good about my results so far. I built an API endpoint that will spit out the top 5 and bottom 5 pages ranked by my appreciation score and this is enough to build a rough prototypal UI on top of it. Hopefully next Hack Week I can revisit this application and finish up my ambitious goal of building what’s effectively a quality-based reader.

What do you think my reader’s UI should look like?


*While we experiment with the data available from our API, our clients’ data is kept private and never shared publicly.

Last week was Hack Week at Chartbeat, and while Shaun’s excited to update you soon on his reader project, there were so many awesome hacks this time around, I had to share a few.

In case you didn’t know, The Chartteam has a Hack Week every cycle where people can work on a project of their own choosing. The only rule is that you have to have something to show at the end of the week. This past week, we saw some pretty fantastic (and useful) hacks. Enjoy the three projects below – two of which will soon be available for you to use.

Tracking coffee consumption at Chartbeat

Coffee is a pretty big deal at Chartbeat – as in we recently built a specific countertop to fit a larger coffee making set up because we’re intaking that much caffeine every day. Josh, Justin, Allan, and Tom got their hands dirty with a system involving hardware hacking and face recognition that tracks coffee consumption at Chartbeat by individual drinker. Crazy right? Here’s how it works: 

  1. Justin did the hardware. A pressure sensor connected to an Arduino was placed under the coffee pot. Data from the Arduino was downloaded to a Raspberry Pi, which tracked the pressure (i.e. weight of the pot) over time

  2. When people took cups of coffee or refilled the pot, the system detected the change in pressure, recorded the event, and snapped a picture of the coffee consumer’s face

  3. Josh wrote an facial recognition system that labeled each photo with its presumed Chartteam member

  4. Tom wrote a web interface for looking at photos. If photos were mislabled by the classifier, people could relabel them with the correct people and the new labels were fed back into the classifier.

  5. In a few weeks we’ll know who’s drinking the most coffee at Chartbeat (so we can blame our Joyride invoices on them).


Building Big Board 2.0

Matt decided the very popular Big Board needed a new look for 2013. So he built a new Big Board on our Sharkbeat code with a much cleaner interface. His Big Board offers both light and dark background choices, several different metrics you can filter the board with, and the option to hide numbers. Check out this sweet video of the new Big Board in action.

New Chartbeat Big Board from Matt Bango on Vimeo.

Creating a domain-based Total Total

For a while we’ve had a Chartbeat Total Total that aggregates concurrents across all of our sites. We’re big fans of displaying our Total Total around the office, and David figured some of you guys might want a domain-specific Total Total – so be built one. Total Total tells you in real time what platforms, operating systems, and browsers you current visitors are using to access your site, which might be new data for some of you. It’s an awesome get-your-data-real-quick visualization to put up on a monitor in your newsroom.

These projects will be posted on our Labs page soon – with the new Big Board and Total Total soon available for you to use. Get pumped for some awesome new visualization bling for your newsroom monitors.


What do you think we should build next Hack Week?

We figure you might not have your eyes on your Chartbeat Dashboard all day. We know that you’re busy.

But even if you’re paying close attention, there is some information that can’t be fully realized or understood until the dust settles – once the day is over, all the pings have been counted and the final results are in.

Our motivation for creating Chartbeat Perspectives was simple: For a long time, Chartbeat customers have been asking us for an end-of-day summary that measured the totality of what happened on their site the previous day. Basically, you guys requested some more context around your site’s data, and we heard you loud and clear.

But we wanted to take things a step further – in addition to providing a general overview of your must-know information (traffic volumes, top stories and sections, top referrers) we wanted to uncover the more significant insights that help newsrooms really piece together what happened. The insights that contribute to a larger understanding the value of your content as it pertains to the quality of your audiences’ experiences.

Starting with the basics, the Daily Perspective surfaces that simple but slightly elusive information, for e.g.:

  • Who was your biggest referrer?
  • What was your top social page? Search page?

  • Which page attracted the most new visitors?

It also gets at more deeply hidden information:

  • How many stories were published site wide, or in a given section?

  • Was that number up or down and does it explain traffic volumes?

And finally the traffic anomalies worth knowing – for e.g. if search or social traffic were substantially up or down. We’ll also tell you if you succeeded in getting your sideways traffic to stick around and read more.

Where do you find these insights? Let’s take a quick tour….

Top Stories by Audience Engagement

The Daily Perspective starts with a list of your top stories, as measured by total engaged minutes.

What the blazes is an engaged minute?” you ask.

An engaged minute is a minute spent by a visitor actually reading or watching or writing – a minute he/she spends investing in your content. When a visitor goes ‘idle’ (no keyboard or mouse events in an active tab), Chartbeat stops counting Engaged Time.

Engaged Time is important because it tells us something not only about the volume of traffic, but also the quality of the content – did people fully consume the piece or did they chase a fun-sounding headline only to leave once the story fell short of their expectations?

That’s Engaged Time. Now back to the tour. Here’s the top stories view.

top stories

For each of the top stories we show you a graph of concurrent visitors and what the peak number was, as well as the Average Engaged Time on the story (how much Engaged Time on average a visitor spent on the story).

You’ll be surprised to find that often the story that had the most concurrent visitors didn’t have the most Total Engaged Time. In fact, there will be stories in this list that were only briefly broke the top ten in your dashboard – or maybe they didn’t reach the top ten at all. Or the story was big in the evening or early morning when you weren’t at your dashboard (we’ll see an example of that later).

Click on a story to get additional insights:

additional insights

The top referrer is listed for every story. Other vital takeaways are also listed, if there are any. In the above instance it’s noted that this story was successful in driving visitors further into the site.

Here’s an article that was popular (lots of concurrents) and had great Average Engaged Time (over 3 minutes). From what we know about Engaged Time and audience growth and what others have likewise figured out, these exceptional pieces of content are what really build a loyal audience.

loyal audience

Top Sections by Audience Engagement

The sections summary is organized similarly to the top stories. The ten sections with the most Engaged Time are listed down the main column and compared to the same day from the week before, with most important takeaways displayed in the left column.

crushing sections

It’s easy to see which sections totally crushed it. In this case Environment obviously had a huge day, and Local was off-pace.

And the takeaways on the left provide those below-the-surface insights. In this case it mentions that Politics had a substantial amount of traffic from links.

To dig deeper into what happened within that section, use the section filter to generate a report exclusively for that section – complete with all the top pages and top referrers for those pages.

top pages

A peak at the Daily Perspective for Politics will help us understand why it did so well that day. What were the big stories? Big referrers? How much content went out?


Every report, including the ones for individual sections also has a third part – the summary. What does this summary tell us about Monday, March 25th on

27 percent

Total Engaged Time was down 27% compared to the previous Monday. Meanwhile the total number of stories published was way up – double what it was the previous week. And if you look at the peak concurrents you can see that in general there just wasn’t as much traffic. But then there’s that big spike at the end of the day on the previous Monday (denoted by the grey line). What was that?

To look at a Daily Perspective for any day in the pastjust use the date selector control.


And here’s the answer – at least regarding that late night spike.

Coming soon: more access to historical data

Look forward to much, much, more in terms of access to your historical data.

Very shortly you’ll see a Weekly Perspective that we hope will really rock your world. The Weekly Perspectives aim straight at understanding your audience – where they come from and their propensity to return. Not all visitors are a good fit for what you do – some come, consume, and move on. Others stick around, read more, and drop subtle hints that they’ll be back. The Weekly Perspectives will help you identify these visitors so you can concentrate on acquiring and retaining more of them – we want to help you figure out what’s bringing your audience in today and what will keep them coming back tomorrow and long after that.

Your feedback is very welcome regarding our Daily Perspectives –  so let us know what you think and how we can help out more with your audience loyalty goals.


Our CEO Tony Haile had the pleasure of sharing his thoughts with Alley Watch about Made in NY™ this morning. Made in NY™ is a favorite initiative at Chartbeat since we love all things digital, we’re lucky enough to partner with a whole bunch of New York media folks, and heck, we love New York City. A lot.

Tony discussed things like getting a larger funnel of engineers into New York City, the perks of a thriving tech community, and how Chartbeat’s values align with Made in NY™’s push for collaboration and open dialogue in the local digital and media industries.

Check out the full interview at Alley Watch, and enjoy the excerpt below:

What does “Made in NY™” mean to you and your company?

For Chartbeat, it means unparalleled opportunities to build the best products for the most successful industries.  New York is a media town.  This city was built by creative doers- writers, artists, publishers, advertisers, and Chartbeat is lucky enough to partner with them to make their jobs a little easier, a little better.

What can be done to maximize the “Made in NY™” campaign?

The New York digital community is interconnected, but sometimes tech communities can seem insular in communicating its value and significance to those without a CS degree or startups on their resume.  What’s so great about NY over other tech communities is that we have an incredibly welcoming community; we understand that it’s not about building something for a newsworthy exit, but to really help others accomplish something they couldn’t before.

This is a big month for Chartbeat. The company just hit its fourth birthday, we’re debuting our new office digs to the public, and most importantly, we’re releasing several awesome new features and products in the coming weeks.

As you know we’ve been busy – we’ve had our hands full in the data science lab studying audience development and our metric Engaged Time, we pushed out some sweet changes to the Heads Up Display, and AdAge’s article about our new quality-focused advertising product got us some love too. Plus we put together an Engaged Time infographic to help people understand this key metric a bit more. And that’s all just in the past few weeks!

Since we’re pretty pumped about the work we’ve been doing and these crazy-big upcoming milestones, we thought this month would be a great time to reflect on where we’ve been and talk about where we’re going. And we’ll discuss the amazing things our clients have achieved despite the online publishing industry’s ups and downs.

So stay tuned for some storytelling, some humblebrags about our way-too-cool clients, and even a few thoughts from our CEO, Tony. And possibly some shameless shots of our office dogs because they’ve been along for the journey too.

Thanks for celebrating with us all month long!

photo (17)