Sachin Kamdar discusses three ways content analytics can drive revenue: building loyal visitors through engagement metrics like loyalty rates; increasing brand recognition through sponsored content and measuring outcomes; and testing different content strategies like headline variations or distribution channels to understand audiences and maximize offsite engagement. The presentation provides examples from Parse.ly's network of 600+ sites and ideas for testing different content approaches.
2. • 600+ premier content
websites
• 50 Billion page views /
events tracked monthly
• 600 million de-duped
visitors each month
• 200 million social shares
each month
@SachinKamdar
10. Online Media: Recommended
Content
• 14% of traffic now leads to
scrolled pageviews
• Up to 3.36 pageviews per visitor
since launching infinite scroll
• Time on site is up 8%
@SachinKamdar
17. Sponsored Content
Testing Idea:
What are ways you can get more
people interested in sponsored
content?
Are your clients getting the
branding/audience recognition they
are looking for?
@SachinKamdar
19. Setup.
Keep it simple.
Share.
The quickest way to appease clients.
Show off.
Make sure you’re capturing the breadth of the
readers.
Sponsored Content: Making Post-Campaign
Reporting Easier
@SachinKamdar
20. Direct Lead Generation
Testing Idea:
Can lead generation be an opportunity
for an additional revenue stream?
How can content provide a lead gen
channel for your organization?
@SachinKamdar
21. Are readers spending less time, on average, with
long-form content than shorter pieces?
@SachinKamdar
Lead Generation: How Parse.ly Does It
22. Are readers spending less time, on average, with
long-form content than shorter pieces?
Lead Generation: How Parse.ly Does It
@SachinKamdar
Q2 2014 Q3 2014 Q4 2014 Q1 2015 Q2 2015 Q3 2015 Q4 2015 Q1 2016
25. @SachinKamdar
Where is your audience
finding your content?
Discover patterns!Produce content for
this community.
Analyze results.
Rinse and repeat.
How to Build and Monetize Communities
26. • Trying to attract a new audience
or launch a new line? Test
strategies for different distribution
channels.
• Are you staying data driven even
when it’s not your site that’s
hosting the content?
Testing Ideas
@SachinKamda
27. Parse.ly + Optimizely:
Enabling A/B headline testing
directly in the editorial
workflow
Online Media: Headline Testing
@SachinKamda
28. • Homepage
• Audience development
• Social media
Headline Testing: Use Cases
@SachinKamdar
29. • Keep the tests short.
• Create variations.
• Implement the “winner.”
• Move on.
Headline Testing: Tips
@SachinKamdar
Parse.ly is an analytics platform built specifically for content publishers.
Parse.ly serves hundreds customers including the logos you see here. There’s three high-level use cases for Parse.ly:
We democratize data insights and analytics for entire organizations to use through our analytics dashboard.
We empower product teams through our API to build a wide-range of personalized and dynamic capabilities in their website.
And we fuel data science teams through our pipeline infrastructure so they can worry less about how to collect data, and focus more on analyzing it.
The easiest way to learn about what we do at Parse.ly, is probably just to take a look at our dashboard. We collect a ton of data on how your content and audience intersect, and then we visualize it to make it easy to understand. We do a good job of walking the fine line between…
(AGENDA) As a content creator, it’s becoming increasingly important to think about how user site experiences can tie back to revenue. This differs by the goals of your particular position and type of organization, but this webinar will attempt to address this question from a few different angles:
Loyal Visitors
Loyalty: Authority Report
Recommended content: Slate Case Study
Long-form content: Pew Research / Lehigh University Studies
Brand Recognition
Sponsored content (general)
Lead generation (Parse.ly use case)
Testing!
Distributed content
Headline testing
GOAL (desired outcome): For online media, loyalty of visitors can help build a fast growing business through opening the doors to multiple revenue streams.
KPI (key metrics indicating success): You’ll likely be interested in engagement.
According to our data (from a recent Authority Report), the average site in our network has 11% of its visitors returning at least once in a 3-day period. There is huge room for growth here. So, if you’re not focused on loyalty as part of your job, you’re missing the boat. It takes a lot of work to get people to come to the site once - but if you can “convert” them, then you don’t have to spend the effort to do that all over again.
Plus - loyal engagement can be a source of potential revenue outside of the traditional display ad model - which is huge, especially as the adblocking conversation continues. It means more opportunities to play around and succeed with using sponsored content, membership models and subscription models.
One question regularly asked is how do I know when I can start to test out these additional revenue streams. Well, taking a look at the right chart we can actually see when it might make sense for you.
There’s a few other interesting things to consider here though:
Though loyal readers make up a small percentage of visitors, they can make up the majority of pageview traffic.
Who should you be programming your site for?
Does this mean you have to give up on growth? Huge growth opportunity. If not focused on driving loyalty, you should be. Requires effort but even incremental gains are huge because of the frequency of visits.
No! In fact, the sites on our network with the most loyal visitors were also the largest sites. Success is a combination of factors – and growing a loyal fanbase is one of them.
If you are a smaller site, focusing on loyalty might allow you to get to that next tier in terms of audience size.
Remember – every site and site section can have different goals.
The audience lifecycle plays a huge role here, as you have to hit all the other steps: discovery, engagement and action (i.e. sharing, etc).
Remember – every site and site section can have different goals
So, what data should customers look at (or are they looking at) to help them gauge loyalty?
There are actually two: engaged time and then returning visitors. You can think about each as a part of the overall funnel. Get someone to visit your site, get them to value your content, get them to come back.
Here are some testing ideas for online media looking to get engagement through recommended content (see slide).
Other ideas: Look at top posts sorted by engaged time and note the commonalities among top posts:
Do the most engaging posts include pictures or videos?
How long are the most engaging posts?
Slate is another Parse.ly customer in online media looking to increase engagement by rethinking how it shows recommended content to readers.
Slate believes that eliminating clutter on its pages focuses attention on its content, that page-load time is a key element of the user experience, and that reducing the number of revenue-generating modules ensures that those remaining offer the best visual and advertiser impact.
Prioritizing time spent on the site over drive-by traffic will enable Slate to balance editorial goals with revenue maximization.
The company’s infinite scroll experience, powered by Parse.ly, is a product completely rooted in these philosophies. (Talk about how Slate is using Parse.ly.)
Here are the initial results:
14% of traffic on article pages is now scrolled pageviews
Slate went from 2.96 pageviews per visitor YTD on desktop to 3.36 pageviews per visitor since launching infinite scroll
Time on site is up 8% since the launch of infinite scroll
Like Slate, online media can use the Parse.ly API to build out the Parse.ly Recommendation Engine on itssite pages. This Pandora-style way of recommending content and presenting readers with articles that cater to their interests is a great way for publishers to keep readers on the site rather than bouncing.
Using Parse.ly Analytics, they monitor if this is working by looking at the "activity after viewing post" section at the bottom of post detail pages and seeing what % of people stay on the site and view other posts and exactly what those posts are.
Here are some testing ideas for online media looking to get engagement through recommended content (see slide).
Other ideas: Look at top posts sorted by engaged time and note the commonalities among top posts:
Do the most engaging posts include pictures or videos?
How long are the most engaging posts?
It turns out that long-form content is another way to keep readers engaged. Parse.ly recently worked with Pew Research Center, in association with the John S. and James L. Knight Foundation, to figure out of long-form content is becoming obsolete on mobile devices.
- What did we learn? Long-form journalism does have a place in today’s mobile-centric society.
- Long-form stories attract visitors at nearly the same rate as short-form
- Across all five distinct parts of the day, readers spend about twice the time with long-form news content on their cellphones as with short-form.
Lehigh University used Parse.ly to determine if the length of its content was affecting the average engagement time with different sections, topics, formats, etc. Wanted to know: Are readers spending less time, on average, with long-form content than shorter pieces?
Found that: Readers spend about the same amount of time on all pieces of content, relative to the length of the article.
The Brown and White has a very short window to capture its audience and convey its message before readers jump off the page.
Long-form content is not dead… The editorial team at the paper renewed its focus on investigative pieces that often require higher word counts.
This is greatly helpful in determining how reporters structure articles, what kind of content we produce (e.g. infographics, illustrations, etc.), and what that content is about.
Does your site have lots of pageviews and low engagement? Poor user experience? It is likely you are losing loyalty opportunities.
Does your site have lots of engagement and low pageviews? What’s your distribution strategy?
GOAL (desired outcome): Readership will likely be important to you.
KPI (key metrics indicating success): You’ll want to know if your clients are getting the branding/audience recognition they desire.
Sponsored content is here to stay. It’s a evolution of the industry. Marketers are disappointed with the results from display ads, and publishers want to provide a path to unique value to their advertisers. Sponsored or customer content is here to stay.
But every potential client is going to be focused on brand, audience recognition, and messaging when it comes to sponsored content. That’s where data and analytics can help out.
Digital media wants readers to consider sponsored content as highly as editorial content.
Setup. Work with what you have already! We recommend making use of one of the most underutilized features of your existing CMS: tags. Since you can see analytics for any tags in the Parse.ly dashboard, simply using the tag “sponsored” and the brand name, i.e. “microsoft-sponsored,” means you have your analytics all set up already. As long as you can access tags, you can access your sponsored content data in your dashboard.
Share. Give clients access to their campaign. They want to be treated like they’re part of the process, and understanding their audience is key here. At Parse.ly, you have the ability to “share” specific analytics (like a campaign) really easily. Clients now can see all readers of their campaigns historically and in realtime. It now gives them the ability to be both reactive and proactive. They can take immediate actions, like engaging on social media or in the comments, and historical, which gives the big picture over the course of a campaign. Whether you’re being asking for a brief recap or a longer-term comparisons across multiple campaigns, you get to make a client happy without spending any time in Powerpoint.
Show off. Make sure you’re capturing the breadth of the readers. Brands invest in native advertising to appeal to a wider audience – yours! The more information you can give them about those readers, the more they’ll understand why it’s so important to work with you. Show them which social networks sent readers, and how much of an increase or “lift” that created over what the article would have normally seen. This can be done at an article or campaign level, if you’re using Parse.ly reports. In addition to showing off your audience, this information can help the brands understand their own target market better. Be more than a distribution platform for advertisers; become a partner that can help them understand how to craft a strategy that works in the long term.
Lead gen can a great opportunity to increase revenue. In fact we use content specifically for this purpose at Parse.ly!
At Parse.ly, we create content for our blog and in our Resource Center as forms of lead generation.
Our blog includes a variety of content, such as tips, trends, team-related news, product news, data pieces, and more. Using data, we look at how each topic resonates with our audience and create future content (and our content marketing strategy) accordingly.
In the Resource Center, we post all other content we’ve created – Authority Reports (original data and research), case studies, product information, videos, whitepapers, infographics, and more. By tracking where leads come from, and where they are going, we are able to make informed decisions about what types of content we should be creating to generate interest and revenue.
In using audience data to track what topics and types of content resonate with our audience, Parse.ly has been able to improve lead growth over time.
Here’s a chart depicting leads over time – from Q2 2014 through Q1 2016. Of course, this changes based on the events we attend, the amount of content created, and other factors. However, you can see from the black trend line that leads are increasing for Parse.ly over time.
GOAL (desired outcome): Interested in understanding your audience.
KPI (key metrics indicating success): Offsite engagement, and whether or not people share.
One way to test this: Look at the average engaged time for new readers. Maybe there's a gem that is engaging a new audience that surprises you.
Having brand managers ready to engage directly with an audience can require a cultural shift internally. This is where community engagement becomes key. There’s been some discussion lately around what's going on with the bottom half of the internet page; many sites have turned off comments altogether, or won't engage with a thread of comments on their social channel--it's just too volatile and unproductive.
We recommend starting a conversation! This applies directly to the social KPIs that Parse.ly surfaces. A community editor or social media team will know that a piece that's getting shared wildly is great for brand awareness, and that a post that's getting clickthrough from those shares is the more engaged audience. Both serve a purpose, and that's two different types of engagement, with two different insights for the publisher.
What type of content is our highly shared content?
What type of content gets our social (and likely, newer) audience to come back and engage?
Asking these questions can help you build out your overall core competencies as a brand, and also within each audience channel. And there's a few ways to find these answers: users can bolster your editorial intuition after looking at top shared content, or you can dive into content topics sorted by social referrers.
You probably already realize that users are reading and finding your content off your site. But did you know that this is actually a great way to build and monetize communities? Digital publishers will soon be able to see traffic broken out by distribution channels — like Facebook Instant Articles or Google AMP — and here’s how will look within Parse.ly’s dashboard.
But regardless of what tools you have to track distributed content, using analytics data you can start to understand the seed of how these audiences are growing. If you’re seceding control to these platforms, take some back through data. Foster these audiences and you may actually build a large enough audience to monetize a considerable amount of money. For example, Cosmo last year claimed it was getting 3M viewers a day on Snapchat and that it can monetize.
They key thing here is using testing as your approach to building audiences in these communities. Audience development should be proactive not reactive, and looking at analytics is critical to taking this approach.
Steps to build audiences in distributed communities:
Look at your current analytics and see where your audience is finding your content. Do you see a trend in a specific community that’s growing over time?
Dive deeper into the data to figure out if there is pattern behind that trend (video, longform, shortform, specific topic).
Start producing content specifically geared towards this community.
Go back to your analytics and see if it sticks. If so continue. If not, go back to step 1.
Optimizely powers headline testing in Parse.ly to help you optimize story headlines and increase article views.
Create tests easily in post detail pages
See and compare the results in one place
Get real-time feedback on what headlines bring in the most readers, and how different headlines affect engagement
Why is headline testing important? “A/B Testing” is new to newsrooms. Using the headline testing feature that Parse.ly developed in conjunction with Optimizely may be the first time editorial teams ever try A/B testing.
Homepage editors can test headlines to revive an under-performing post
Audience development editors can test headlines’ effects on increased engagement or loyalty
Social media managers can try to boost shares by testing a headline before posting on social
Don't run headline tests for too long if you're running them for 100% of your audience.
One way we've seen many of our customers approach headline testing:
Create variations. The more dramatic you make the changes the more likely you'll find a meaningful winner
If your find a winner in ~1 hour (in this example), implement the winner
If the experiment hasn't found a meaningful winner in an hour, pause the experiment and move on, trying more dramatic variations next time
(RECAP of PRESENTATION – Questions?)
Loyal Visitors
Loyalty: Authority Report
Recommended content: Slate Case Study
Long-form content: Pew Research / Lehigh University Studies
Brand Recognition
Sponsored content (general)
Lead generation (Parse.ly use case)
Testing!
Distributed content
Headline testing