Measuring Quality
How Chartbeat aims to increase good content online
Chartbeat is an analytics company that makes tools based around attention versus clicks. The company launched out of Betaworks in 2009. Internet Week sat down with CEO Tony Halle to learn about the company’s plan to make engagement metrics more valueable than page views.
Internet Week: How did Chartbeat begin?
We came at a time when the receipt was that real-time analytics as was a bad dangerous thing. No-one did real-time analytics at the time. We didn’t find any of the metrics people were used to seeing. The metrics being used were pageviews, but we focused on concurrents and engaged time, which were bizarre to people. Were were a small service doing real-time data when real-time data was bad. But we felt pretty strongly there were new use cases to be found in real-time web analytics.
We had been part of Betaworks and seen how the real-time web had changed search, and were curious to see the same experiment done in analytics.
The first year was slow. We were going out speaking to people until there was a certain moment when we got an email from Time.com that said, “hey, all our front-end people are using this.”
It took off. Now 80% of top publishers use Chartbeat.
Internet Week: What did you discover about online metrics?
What’s really interesting for us is the idea of measuring feedback. The problem with the metrics the world was using was they were largely focused on things like page views which is the measure of clicks to a piece of content, not the content itself. The metrics just encouraged clicks, and no engagement with the content, and that was plain wrong.
It’s why we wanted to avoid page views in the first place — it’s not a nice metric to use. You can have content not many people click on but really love when they get there.
One of the things we found was ad sales people started using these metrics to to sell ads indirectly. When we asked why, they said, “if we’re fighting an impressions game, where there’s an infinite number of inventory, we will always lose.”
The thing we care about a lot is the mission of Chartbeat, that 10–15 years
from now, good content on the web is making money.
We started exploring how to make the case that quality actually matters and has an impact on the bottom line.
There was a clear correlation between brand recall recognition and the amount of time spent capturing their attention. We started asking how to build the tools necessary for publishers to make the case for quality content and advertisers can get trust and transparency not by getting an ad that’s loaded for one second, but for 20 seconds because the great quality content keeps people there.
Internet Week: What were things like in the early days?
There were just two of us at one desk, and we didn’t have much of a clue as to what we were doing.
Our biggest challenge was that there were already a lot of players in the analytics space, and then Google started offering analytics for free. We knew we wouldn’t be faster than Google, so made a list of all the things they did with analytics and then made a list of the opposites of those things and just did the entire opposite. We thought if we’re like Google we’ll lose, but if we’re the opposite of Google, we may lose as well, but at least we’d be different.
The other thing we found was that analytics tools used to be for people who were good with data. But we found organizations had to become more adaptive and people were democratizing decision making. The people who were making decisions weren’t data scientists nor people who could look at an Excel spreadsheet like it was poetry.
Internet Week: What advice do you have for someone starting a company?
It’s enormous fun. It’s terrifying, but don’t be afraid is the biggest thing. You will be surprised by how many people will give you advice if you just ask. You start realizing there are a lot of people out there who were in your place once and they all want to to see you win.
Internet Week: What do you tell people who ask how to grow their readership?
People tend to think there’s a trick to this. There are tricks to
growing traffic, but there are very few tricks to growing audience.
Traffic is a one-off person who comes from Argentina or Latvia but never comes back. An audience knows you who you are and what you’re doing and keeps coming back. You have to write something genuinely interesting that people want to read.
Internet Week: So how do you build an audience?
First, make sure your site is readable and that you have enough people creating enough content. There’s a certain amount of organic growth around that.
Then, set up your site for search, and at that point, start hitting social strategically.
It’s interesting what you can do with paid promotion. If you’re spending money to get people to the site, see if those people are converting. If you’re paying for one-off visits, then it’s not valuable, but if that person will come back once a week for the next year, it’s worth it.
Once you have people coming, focus on the Property of Return: someone who
stays three minutes is twice as likely to stay as someone who stays for less time.
Your job is to get them onto the next story or keep them reading long enough to know who you are. 55% of all page views on Web get less than 15 seconds of engagement. Find out what works for you and focus on that source until it taps out.
