How the Parse.ly team didn’t let implementation obstacles get in the way of delivering elegant, beautiful, and timely data digests to our customers via e-mail.

I frequently say that Parse.ly is lucky to have very smart and savvy individuals as customers. We work with editors, writers, and analysts at top media companies.

Let’s reflect on journalists for a moment. Their primary job is to collect voluminous and disparate raw data from a variety of sources. They then exercise analytical skills and editorial judgment to render that data into insights for readers and viewers. They need to do this quickly, as relevance and freshness are closely connected.

As a company whose flagship product provides web analytics to online publishers, this description of a journalist’s job sounds very familiar. In our case, our “audience” is the journalist, as well as the editors and analysts who support their work. This puts an enormous burden on us, but in a good way.

Our customers can tell the difference between real insights and noisy raw data. If we are providing noise with little signal, our customers will call us out on it. It means that every feature we release needs to be held to a high standard.

We also recognize that editors and writers are under enough pressure as it is. The last thing we want to do is waste their time when our software could do the work for them.

It is with this understanding that we embarked on a mighty project over the last couple of months. We built a powerful visual summarization of site traffic and publication trends called the Parse.ly Weekly Digest.

The goals: 

• summarize millions of traffic and content signals into a weekly email that is personalized to each author or editor’s content
• provide context by benchmarking performance against the average of the past four weeks
• deliver all the information in a personalized, rich e-mail, so that it can be viewed on the go (from mobile devices) or easily forwarded to other colleagues

This was a tall order. 

It required weeks of rapid prototyping with existing customers providing us direct feedback on the metrics that mattered to them. And it involved several design drafts: from wireframe, to mockup, to data visualization prototype, to completed implementation.

It also required a Herculean engineering effort: to acquire and aggregate the right data; to implement the beautiful data visualizations; and to make those visuals work in the restricted technology platform of e-mail.

The original wireframe for the weekly digest was developed by Mike Sukmanowsky, Parse.ly’s Product Lead. Mike came to Parse.ly from the Globe & Mail, Canada’s premier daily newspaper and a top online news destination. There, he was Product Manager of Analytics. He developed the wireframe from his experience working with writers and editors at the Globe, and via interviews with Parse.ly customers over the course of several weeks.

The wireframe was developed further through discussion on the Parse.ly team. A surgical team of three talented Parse.ly developers tackled the wireframe head-on: Kemper Smith, our visual designer; Toms Baugis, our UI/UX lead; and Vincent Driessen, a full-stack engineer.

Vincent started by prototyping some data analyses atop existing customers’ data. He had just finished an internal project where we analyzed customer post data to detect “Peaks” and “Troughs” — exercising our new real-time data store, which we call “Pulse”. This project gave him an opportunity to exercise our historical data store, which we call “PTrack”.

The result was a code prototype within Parse.ly Dash that did some weekly aggregation of customer data based on the wireframe.

It was at this stage that we could prioritize pie-in-the-sky ideas against the capabilities and limitations of our existing data stores, while also establishing whether the chosen metrics “looked right” given the brutal reality of production data.

We also realized that the wireframe packed too much information across horizontal space, and that this would be too dense for e-mail (and even for the web). So, we sought to simplify the presentation from the wireframe a bit at this prototyping stage, going for more of a vertical-oriented layout.

Vincent’s work was then handed off to Kemper. Since Kemper knew exactly what metrics we could support and had real data to support a design, he could approach the problem directly. He sought to make an elegant and compact presentation, as well as a visual design that matched Parse.ly’s overall brand. The result was a stunning mockup of the Weekly Digest that was instantly loved by the team.

We now had enough material to get some early feedback from customers. We began circulating the mockup to customers who had provided feedback in the earlier stage, only weeks before. Through their feedback, we were able to make some small adjustments. More importantly, we received resounding confirmation that we were on the right track with the Weekly Digest.

Yet, we still had a problem.

We had a visually stunning mockup based on data that we knew we could aggregate easily. But there was still a gap between Vincent’s experiment and the mockup developed by Kemper. Could we cross this chasm? The engineering team was healthily skeptical.

Our UI/UX lead, Toms, suggested we just charge ahead with implementing Kemper’s vision using our workhorse data visualization library, d3. Toms was confident that we could reproduce the visualizations that Kemper had designed using d3. Kemper agreed, and even started work on doing so. Kemper is that rare breed — not only a visual designer, but also a JavaScript programmer.

As this collaboration between Toms and Kemper went on, they knew there was a roadblock over the horizon: e-mail. d3 is a framework which only runs within a web browser, and for the visualizations we wanted to create, we’d need to use a powerful browser graphics subsystem called SVG. But e-mail does not support SVG.

Some background on the challenge of rich e-mail. Up until a few years ago, the best you could do in e-mail was plain text. But most e-mail clients over the last few years have adopted HTML e-mail, which allows use of a limited amount of HTML, CSS, and in-line images — though, no JavaScript or SVG. E-mails have to be very conservative in the markup they do use, since they tend to be viewed in, shall we say, “hostile” environments. Consider that the same e-mail can be rendered in a slew of desktop clients (Apple Mail, Microsoft Outlook), web clients (GMail, Hotmail), as well as  mobile environments (iPhone Mail, Android GMail, or Blackberry Mail). What’s worse, the e-mail in webmail clients renders differently in different browsers. This is even worse than the cross-browser problems that exist on the web.

How could we get from a rich, dynamically-generated visualization that only works in modern web browsers into this limited HTML e-mail environment?

I had an idea.

I had recently been testing browser automation technologies for doing better automated testing of Parse.ly Dash. One of the technologies I came across during my work on this was PhantomJS. It is a full-fledged modern browser (based on WebKit, same engine that runs in Chrome and Safari) that you can run on a server and script it via JavaScript code.

I wondered: could we perhaps get PhantomJS to run our data visualization on the server, capture the rendered output, and then assemble an email with light HTML, CSS, and rendered images? By moving the complexity from the e-mail to server, we could guarantee they work in the cornucopia of fragmented environments that define HTML e-mail.

I discovered that PhantomJS’s browser not only supported SVG, but also that it already had an interface for capturing rendered output from the browser engine into a variety of image formats. This seemed promising.

We had a hopeful path forward: a possible way to turn Kemper’s beautiful design into a pixel-perfect rendition in HTML e-mail.

Vincent got to work on a new internal library called domshot. The idea behind domshot: allow our team to assemble rendered web views — using the Python web technologies we use and the d3 visualization library — and capture parts of the screen as images. He then refactored his code prototype to use domshot.

The next couple weeks of iterative development saw Vincent getting closer and closer to having domshot working perfectly with the experimental PhantomJS project and our live data. Meanwhile, Toms and Kemper got closer to getting a web view of the Weekly Digest to work fully across all our customers’ data using live d3 renderings on the web.

The Parse.ly team was firing on all cylinders.

Once domshot and the renderings were working well together, it was time to get everything into the email itself.

Toms crossed some t’s and dotted some i’s. He leveraged Amazon Cloudfront to host the dynamically-generated, personalized images for each e-mail, and linked those from the rendered HTML mail templates. He also whipped together a Dash admin interface that let our team schedule mailings to individual authors and editors, specifying filters on things like section and author. Then, one day, I received this e-mail in my GMail inbox:

My jaw dropped. We had done it.

I once attended a one-day course by the renowned Information Visualization educator, Edward Tufte. Something Tufte said that day struck me:

In good information visualization, there are no rules, no guidelines, no templates, no standard technologies, no stylebooks. Instead, to convey the right information to your audience, your approach is both obvious and difficult. You must simply do whatever it takes.

In my work at Parse.ly over the last few years, I’ve developed and nurtured an enormous respect for the work done by our customers: editors, writers, analysts, journalists.

Day in, day out, they do whatever it takes to make real, sustainable journalism happen on the web.

The least we could do on the Parse.ly team is return the favor.

Written by Andrew Montalenti, CTO: @amontalenti

Follow us on Twitter: @Parsely

(Discuss on Hacker News)

—

Interested in Weekly Digest for your site? Visit http://dash.to/try to sign up for a 30-day trial of Parse.ly Dash, the analytics tool built specifically for online publishers. Parse.ly’s technology is trusted by editors and writers at The Atlantic, US News, Dallas Morning News, ArsTechnica, Mashable, and many others.

• Tags
• analytics
• dash

Traditional publishing is changing. We’ve seen for a while that print publications are figuring out the proper transition to digital and we see another case today with Newsweek abandoning print to go fully digital in 2013. At Parse.ly, we’ve had the privilege of experiencing this transition period first hand and have even helped publishers work toward a successful digital future. One such example is Press Enterprise, a daily newspaper in California that was actually the inspiration for this post. 

Andrew McFadden, Manager of Innovation & Business Development at PE.com, recently published an article on INMA.org (International Newsmedia Marketing Association) leading with the idea that traditional newspapers must, in a sense, become startups if they are to take advantage of the opportunity digital publishing presents. But this means a complete culture change which is not exactly easy. Regardless, it’s a challenge that PE and many others have embraced. 

McFadden speaks of the difference between reporting and analysis which hits quite close to home with us here at Parse.ly. In the days of print, reporters would research a story, write it, submit it to their editor, and their story would show up at the doorstep of millions of readers the next day. There was no real data to analyze. The only significant metrics to be aware of is total circulation of the paper and revenue. But this does nothing to inform them of how their audience responded to their content. Reporting on digital distribution goes deeper, but the flaw with traditional “reporting” is that it tends to be an end of month look back at “how we did.” What’s missing from this is the analysis to answer the what, why, and how of it all.

So this is where news organizations need to innovate. Lucky for PE.com, they happen to have a Manager of Innovation to take a top down approach in implementing new technologies, processes, etc. Most other organizations are not this lucky. In these cases, where does the innovation come from? Well, it has to come from everyone. Journalists need to understand that the distribution model has changed for digital and there is work to do after the story it written. There are innovative ways to research, publish, market, and optimize. This is where analysis over reporting is truly powerful. To illustrate, here’s a quote from Andrew McFadden’s article that appeared on INMA.org on October 1, 2012.

So, about a year ago, we began looking at analysis level solutions to help our news editors create journalist-level dashboards. We joined ArsTechnica.com, The Atlantic Monthly, and Mashable in a Parse.ly private beta. During that year, our newsroom has shifted from looking at pages viewed per month to assessing referral traffic, and then timing posts based on what we know will maximise our success.

By bringing the data to an incremental level where action can be taken, we’ve been able to take advantage of what users are currently doing on-site so that we can build what we call a “leading indicator advantage.” The result is that we’re guessing less, working from real data more, and yes, seeing better overall monthly results on those monthly reports. 

The key here is that anyone in the organization can access Parse.ly Dash analytics. We were quite careful not to create another analytics product that requires an analyst to make sense of the data. From our perspective, we think of each journalist as building their own startup / brand. The publishing process has evolved from simply researching and writing an article that goes to print to now add in basic performance analysis, marketing, social media, and general content optimization based on historical and realtime metrics. With analytics technologies like Parse.ly hard data is easily transformed to actionable insights that ultimately allow media companies to have a more integrated approach to the entire publishing process. 

Integrated approach to publishing = Success in the digital age

What type of innovation is coming next you ask?

• Deeper integration between staffs and less segmentation
• More forecasting and predictive analytics
• Constant innovations around consumption, delivery, & measurement
• Audience building at the author level, not just corporate level

We’re starting to see a lot of great changes as technologies improve and cultures change. But to get back to why newspapers should operate with a startup mentality…startups are fun. The opportunity to take full ownership, learn new ways of doing things, succeeding, failing, persevering, leading, and everything else that comes with the territory is absolutely an amazing experience and fully gratifying.

- John Levitt, Director of Sales & Marketing

• Tags
• Newspapers
• analytics
• startups
• Parse.ly

Having spent much of the past three years writing about restaurants, it is only natural that my friends, when told about my job at Parse.ly, respond, “does that have something to do with food?” 

True, I clicked on Parse.ly’s job posting because of the name—I, too, wondered what Parse.ly might have to do with food. 

But having spent much of the past three years working for newspapers, magazines, and digital publications, it is only natural that I spent this summer working at Parse.ly, a startup that provides analytics for the web’s best publishers. 

When I started working at Parse.ly in June, I walked through the door with a few preconceived notions about digital publishing. Such generalizations are not even close to plausible. I thought that digital publishers were clinging to an analog era. I thought that digital publishing was lagging behind the evolving web-mobile interface. And I thought that most digital publishers wouldn’t know how to use analytics even if they were given a detailed instruction booklet. After all, as a recovering digital publisher myself, I was clinging to an analog era. I was lagging behind the web-mobile interface. And I personally believed analytics to be little more than a diversion and a sales tool for booking advertising space. 

Over the past three months, I have made up for most of the past three years’s mis-instruction. I learned that most publications have one foot in the analog, one foot in the digital pool. I learned that web-mobile platforms are being innovated by publishers, not by social developers. And I learned that given the best analytics—Parse.ly’s analytics—publishers are able to improve their editorial strategies and produce better content for all of us.

There is a reason why I needed to be reeducated: there is much obfuscation and misinformation on the part of digital publishers themselves. The media scrum around, well, digital media is insane. Every day one opens the ‘ol browser only to find conflicting stories and opinions on everything from the New York Times’s paywall to a pseudo-publisher like Google. The paywall is working! The paywall isn’t! Google is a publisher! Google isn’t! Although the big questions surrounding digital publishing don’t have easy answers, readers, like myself, tend to draw the wrong conclusions from such contradictory speculations. It is easier to see the battlefield of future publishing and conclude, digital publishing is doomed, than to see the battlefield of future publishing and conclude, digital publishing is spiraling in the right direction.

Debates over digital publishing are, fortunately, dialectic. Analog and digital perspectives meet: out of their grand collision, a fertile fireball emerges, a synthetic, creative compromise between the two perspectives. Thus, publishing innovation always proceeds as two steps forward, one step backward, or a halting, staggering wander forwards. Progress is not linear, but spiral. As we circle back with longing, as we linger and indulge in nostalgia, as we try to reclaim what is outmoded and atavistic, we are drifting forwards. The circle rotates around a progressive axis. Though we perceive its motion as cyclical or static, digital publishing is, indeed, making progress towards more viable economic models. The process is painful and slow.

For example, the great growing pain of our moment is plagiarism. In a web-based publishing world, “stealing” and “copying” might better be referred to as “recycling.” Is there a line between a reblog and a rip-off? What is the protocol for Internet attribution? To what extent can you rewrite your own ideas? We should not forget that plagiarism, of the self or of others, is not a modern phenomenon; it is merely the case that the Internet makes plagiarism all the more visible and traceable. Therein, digital publishing is forced to confront an issue that was submerged or avoided for the past two millennia. When we have drafted some half-way satisfactory answers, the rotation about the circle will be complete, and we will have made a few inches of “progress.”

TL;DR: Everyone needs to calm down. 

This summer, I learned that there is such a thing as a “fitness office.” I had the opportunity to talk with a ton of startup CEOs. I went to some good places for lunch, like Alidoro and the Calexico Cart. I lived in Crown Heights and explored neighborhoods beyond my usual reach, including Sunset Park, Hasidic Williamsburg, and Bay Ridge. I saw at least two movies a week, more art than I could safely swallow, and a concert or two. The pieces of my summer puzzle don’t fit together in a coherent picture, nor would I want them to. I trust, however, that my miscellany of experiences, along with my work at Parse.ly, condenses into a few productive and ambiguous configurations.

Having spent much of the past three months writing about startups, it is only natural that my friends, when told about my job at Parse.ly, respond, “what did you do?” I could answer with a straightforward explanation of my duties, obligations, and responsibilities. Instead, I usually say, “I learned as much as I could.” That implies an unequal transaction of value between myself and my employer, but I think it’s an honest and fair description of what happened in Herald Square. 

I’ll be taking a few weeks off from blogging here; look for content from some other Parse.ly employers. In September, I’ll be back on a weekly basis. Enjoy the dog days. ~Jason

• Tags
• Parse.ly
• Dash
• analytics
• digital
• publishing
• technology
• startups

Read Part 7, Army of One. There’s a party on the Parse.ly homepage.

The key to content promotion is to hide all traces of marketing and advertising. The digital consumer’s nose is quite sensitive to the scent of desperation. Failing to disguise a promotion as such limits the potential reach of a content distribution channel. Desperation signals that your product isn’t cool—isn’t intrinsically good, isn’t good on its own merits. Desperation signals the necessity of a marketing campaign in the first place to paint a veneer of dazzle on an otherwise bland canvas. Although building relationships with readers is the most effective way to build a loyal audience, building relationships does not mean overpromoting and overbearing enthusiasm. Turn it down from 11. To about 7.

What are some signs that you might be desperate?

1. Unsolicited daily e-mail digests. Lately, I’ve noticed an increasing frequency of daily e-mail digests. A digital publication will, after obtaining an e-mail address from an affiliate or more illicit source, send a reader an unsolicited e-mail digest. In effect, the publication has subscribed a reader to a digest without his or her prior consent. While this kind of spam is generally legal, it is annoying. Massive exposure to content might trap some less than savvy consumers. It is likely, however, to turn-off valuable and otherwise potential subscribers.

2. Repetitive promotional outreaches within a constrained time period. Interactions between a publisher and a reader need to be transactional and reciprocal. That means one-sided interactions have a low ROI. If a publisher reaches out to readers continuously without reciprocal feedback, a structure of desire for the publisher’s product never emerges. Sufficient time must be allowed to expire so that recipients can take action on promotions or initiate reciprocal interactions.

3. Too frequent Tweeting. Clogging up the Twitter streams of your followers, again, may trap some user-types. Over the long term, though, too frequent Tweeting will limit the efficiency of your social engagement and content distribution. Use optimization models to determine the best times and frequency for distributing content. 

4. Using social media “lingo” inappropriately or unironically. In high school, there are two types of uncool kids. The first, those who naturally fail to fit in with the “cool” kids. The second, those who try too hard to fit in with the “cool” kids. It’s fairly obvious which late-adopters are failing to implement the vocabulary of social media; it’s more obvious which late-adopters don’t know how to use that vocabulary judiciously.

5. Unnecessarily conspicuous share buttons. There is an optimal size and position of “share” and “subscribe” buttons in any web site layout. Exceeding that optimal size or cluttering the page deters repeat visits and direct traffic. When readers actively notice the share button, it’s not doing its job: readers should be aware of the opportunity to share, but not overwhelmed with requests to share.

Don’t be desperate for reader love. The following advice gets bandied about too much on the Internet, but it’s true: if you build it, they will come. Sometimes, publishers with good, marketable content get unlucky. Usually though, a strong product speaks for itself.             

• Tags
• marketing
• social media
• Twitter
• audience
• analytics
• digital
• publishing
• content

Read Part 6, A Little Less Conversation. Learn more about what we do at Parse.ly.

When building an audience, never use smoke signals. Assume that you’re a publisher, and like most publishers, you think your content is pretty cool. You want other people to get a piece of your cool content—but you need to let them know that your content even exists. So you light a fire on some social media platform and start sending generic messages: links with pithy headlines, quotes from the content, provocative questions, etc.: hoping that a relatively low percentage of your extant audience will see the link, an even lower percentage will click on the link, and an even lower percentage will light their own fire and share the link. The logic of this content distribution strategy is simple and based on an efficiency proposition. Generic social media messaging requires a small investment of resources compared to the payoff, at least if you have a large enough audience. Low conversion rates from pre-packaged links necessitate a substantial crowd effect. Without a sufficiently massive user base, social media traffic and subsequent audience growth will remain inconsequential. 

For small and medium-sized publishers that lack intrinsic credibility and brand celebrity, building an audience requires a different strategy. Content distribution and engagement with readers needs to be customized and personalized. New readers must feel like valued members of a community where valuable content is circulated. Generic signaling does not generate instant value. Of course, startup publishers often lack the resources to conduct custom or personalized content distribution campaigns. Such campaigns demand dedicated personnel and an additional diversion of editorial resources from content generation into content distribution. Yet, building relationships with new readers is worth a considerable investment of resources. Exactly how much remains a question of business strategy, but the opportunity-cost of launching better content distribution channels is attractive regardless of individual publisher parameters.

What does custom content distribution look like? 

1. Targeted by demographic. New followers and readers should be grouped according to demographics: gender-identification, race-identification, age-identification, class-identification, etc. Content should be pushed out to targeted demographics using @ cc’s, direct messages, email digests, or recommendations surfaced on the site. 

2. Targeted by interest. New followers and readers should be grouped according to interests. For example, a tech publisher might classify readers on interest in hardware versus software, mobile versus desktop, etc. Classifications can be based on data flowing through cookies or social media profiles. Again, content should be distributed through @s, dm’s, emails, and recommendations.

3. Human, not hollow. The secret to social media interaction is to avoid the affect of automation, even where automation is unavoidable. Twitter responses from official, anonymous accounts should still betray the touch of a human team. That’s why NASA’s Mars Rover Twitter account has been so successful. Use first-person, avoid PR speech and marketing filtration, and rely on cultural references to communicate human intelligence (or the lack thereof). Obviously automated and generic “mass mailings” come across as hollow. As a publishing operation scales, accomplishing #1 and #2 while maintaining #3 becomes more difficult. Developing a thorough social media style guide that rejects “hollow,” mechanical, and computerized affect can help ease the resource burden.

4. Don’t play defense. Don’t wait for followers and readers to query you. Reach out and ask for opinions from individual readers based on #1 and #2. Use the search function on social media platforms to find small-time influencers outside of your core audience, and based on their demographic profile and interests, bring them into the conversation around your content. Although mentions, retweets, and comments should receive customized responses, there’s no reason to wait for inbounding engagement. Start personal engagements before your readers.

5. Reactive, not responsive. Never just “respond” to conversations on social media platforms, reader inquiries, or inbounding messages. Be reactive. When a publisher is responsive, he concludes an engagement series. “It’s a wrap.” Instead, react to readers. Think about starting new engagements from otherwise limited interactions.

Building an audience means building an army of one. Or rather, an army of many ones. Over time, the resource intensive production of custom content distribution channels can be transitioned into permanent pseudo-custom structures. Therein, individual readers, though incorporated into selected demographics, feel like distinct, independent consumers.

• Tags
• analytics
• social media
• audience
• publishing
• digital
• media

Read Part 5, In Search of Traffic. And sidle on over to our homepage, if you will.

The headline of this post is misleading, unfortunately, but I needed to capture your attention on whatever social media saloon you frequent. And if you’re reading this, it worked; you clicked-through. Congratulations; another victim of false headline advertising. Instead of arguing, counterintuitively, that building an audience for a digital publication requires a minimization of “conversation” between content creators and readers, I advocate for a maximization of interaction between producers and consumers. Digital publications do need a little less conversation and a little more action. But I mean that editorial teams need to think about conversation as a valuable “action,” not a diversionary side project or an annoying obligation.

Thanking readers for retweets and rewarding frequent engagement on social media are fine for a start but unlikely to forge enduring relationships. Before the rise of social media, there was an asymmetry between writers and readers. The former made content, the latter consumed it and occasionally commented on it via letters, or later, emails. Social media empowers the reading public to directly and publicly comment on content and comment to content creators. The divide between journalists and their audiences is rapidly shrinking; the power imbalance is approaching a kind of equilibrium; the fundamental binary of media is becoming more symmetrical.

As readers feel more confident in their own voices, and as social media platforms commoditize user-generated content, the condescending position of professional journalists towards readers will become a less effective strategy for building an audience. Rather, readers are expecting a more equitable relationship and a continuous exchange of commentary. The incorporation of forums and aggregation platforms like Reddit into the journalist’s toolkit has increased the visibility of readers as content creators. In effect, the convergence of journalists and readers has given formerly silent audiences a new entitlement. Readers feel deserving of special attention.

How can journalists make conversation more active?

1. Search social media for your content. Find out who’s talking about your content on social media platforms. Talk back—which means more than just thanking readers, for, well, reading. Ask readers their opinions about your article. From Twitter bios, identify the individual interests of individual readers and target your inquiries. For example, if you write an article about gymnastics, and a Twitter user who identifies as a former gymnastics coach retweets the article, an obvious opportunity emerges for a personalized conversation.

2. Search social media for related content. Brainstorm a list of related and peripheral topics to your recent content. Insert yourself into on-going conversations. Redistribute your relevant content to those parties. Talk back, rinse, repeat.

3. Let your readers provide you with value. Take advantage of your readers’s expertise. Ask questions, ask for quotes, query the crowd. USA Today has a well-developed campaign for extracting the aggregate value of their readers. USA Today’s official social media accounts regularly ask their thousands of followers for help with articles. Acknowledging the power of readers—their intrinsic value as information sources—makes them feel like valued members of a community. 

4. Share outside your product. Once you engage in orbital-level conversations about related topics, don’t be afraid to share content from other media outlets. Behaving in non-self-promotional ways demonstrates your sincere interest in the reader and your altruistic commitment to the conversation.

5. Don’t be anonymous. Employees of large, corporate media outlets should engage their readers from personal Twitter accounts. But official, anonymous social media channels should exhibit more personalized, non-anonymous behavior, too. Don’t just use the official channel for distributing your own content; share cool stuff with your readers and contribute to related conversations. Become a many tentacled beast with a face. (Slate does a great job with an officially faceless channel.)

Active conversations are the best way to build a resilient and loyal audience base. Quantity of interaction is meaningless minus quality. Don’t be afraid of treating readers like…real people.

• Tags
• analytics
• technology
• digital publishing
• publishing
• media
• social media

At Parse.ly, we’re keeping a big secret: Dash isn’t just an analytics tool for web publishers. We also offer our clients access to our API. Ars Technica, a tech news site that caters to the “alpha geek,” has been putting the Dash API to work. As Conde Nast’s only completely digital publication, Ars Technica is a natural fit for our innovative and open development structure. With the help of our API, Ars Technica has translated their analytics data into a better reader experience and better marketing strategies.

Ars Technica has implemented the Dash API in two areas, a site overlay and a recommendations engine. The site overlay displays real-time page view data on top of the Ars Technica homepage, allowing editors and writers to respond quickly to trends and to adjust article distribution pathways. According to Jason Marlin, the Director of Technology at Ars Technica, the response-time of the Parse.ly system made the overlay possible. “You start to look at what’s available to you, and Parse.ly’s already got this page view data available. We can just cache that for ten minutes and show that rather than building from or extending our existing crazy pixel system,” Marlin said.

Using the overlay, Ars Technica has already seen measurable results. “We just had a really good month in the month of June for Ars, it was actually the best month we ever had in terms of page views and visits,” Marlin said. “I’d like to say that some of that is probably attributable to other editors logging into the site and seeing all these numbers at a glance, and recognizing that other writers are going to see those numbers as well…I think having it right there at a glance has helped people tweak how they make their headlines and the types of content that they write about.”

Besides the site overlay, Ars Technica uses the Dash API to power a recommendations engine. The recommendations appear in two places, in a “My Stories” space tailored to specific users and in a grid underneath posts. “We’re using the profile capabilities of the API where we’re storing a user profile, so that we can build a base of information and browsing habits on specific users whether they’re anonymous or not,” Marlin said. Since switching from a commercial recommendations system to their homegrown engine, Ars Technica has even seen a slight increase in the percentage of users clicking-through the recommended stories.   

Over the long-term, Ars Technica has a specific goal for the Dash API—to increase the number of pages per visit. “That’s something we’ve had trouble growing,” Marlin said, “because naturally as you grow the number of unique visitors to the site, the amount of overall and the breadth of the traffic, the type of different users you get when you’re reaching out to Google News versus the same two hundred users, those stats are going to go down.” Using Parse.ly, the Ars Technica team can “make the individual user experience ever more tailored to the needs of that user.”

For example, evergreen content that hits it big on Reddit can escape the notice of new visitors. “Sometimes there’s some pretty cool things that happen where you have a story that was a year old that suddenly spikes through the roof because it hits Reddit or people are passing it around Twitter,” Marlin said. “It’d be cool to surface some of that stuff for users, so the user coming to the site can take part in the social discovery that’s going on even though they were not exposed to it through Facebook or Twitter.”

As Ars Technica continues to evolve, the Dash API will along with it. For Ars Technica’s team, “the best part of working with Parse.ly is just how good you guys have been about implementation versus, ‘oh well we’ll put this in our future request list and it might make version 12 or something.’ It looks like you guys have a more agile development environment,” Marlin said.

At Parse.ly, we’re committed to going where other analytics providers can’t or won’t. “I guess real time stats are a pretty crazy idea,” Marlin said. But when we think about Dash, “crazy,” never crosses our minds—because with our API, we can make crazy possible.

Interested in learning more about Dash and our API? Give it a try at dash.to/try, contact us at hello@parsely.com or take a look at our API docs at parse.ly/api.

• Tags
• Dash
• Parse.ly
• API
• analytics
• Ars Technica
• Conde Nast
• digital
• publishing
• technology

Read Part 4, Contact Buzz.

Managing search engine traffic is a job best left to the SEO specialists. Site architecture and engineering parameters determine performance on search engines, within limits; that is, increasing visibility on search engines is only half the battle. Potential readers first need to be searching for terms relevant to your content. So if the technical intricacies of SEO escape you, never fear: with the help of an analytics platform like Dash, you can take action to direct more traffic at your content through search engines.

It is difficult to make generalizations about search engine strategy, because each site has its own technical idiosyncrasies. Nevertheless, there are two universal editorial habits that can help improve search engine traffic:

1. Be Trend-Forward: Although news reportage happens in retrospect, entering into an already crowded topic area sets you up for an SEO battle. Instead of trend-following, try to predict the niche angles from which peripheral competitors will approach stories. Easier said than done. But over time, and through trial-and-error, it is possible to formulate a comprehensive iteration sequence for covering news from different SEO friendly positions. Thinking “trend-forward” also means thinking outside the news story: producing non-news content that takes advantage of memes, click-bait, and other buzzy material. Monitoring aggregators and message boards is not enough. You need to move off the front page, where content is already peaking, and identify which content, topic-areas, and themes are gaining traction.

2. Use Search Analytics As A Tool: Search engine traffic is a powerful indicator for trending content. Dash gives you a lot of control over data filtering. You can look at the performance of posts published within a certain date range and from a certain referral source. By selecting search engine referral, you can compare the total number of views in a time period to search engine referred views. If old content starts resurfacing with most of its traffic from search engines, it’s a good sign that there’s a resurgence of interest in that particular subject.

From that original insight, you can begin generating related and supplementary content to build a new buzz. “Evergreen” content is at its most valuable when it spins off new, ancillary content. In aggregate, the sustained trickle of traffic from evergreen content expands into a substantial boost when you satisfy the demand for additional content.

Search engine traffic is not an end in and of itself. Rather, it is a means to an end, either as a technique for building 360 degree content or as an indicator for hot topics.

• Tags
• Dash
• Parse.ly
• analytics
• building an audience
• traffic
• search
• SEO
• search engine optimization

via Steve Burns

November 14^th, 2005 marks the beginning of an era. On that date, Google first offered a free analytics tool to the public.

The Google Analytics Era

In April 2005, Google had acquired Urchin Software Corporation, a leading web analytics platform. Urchin calculated traffic by analyzing log data, a rather primitive method of collection. As early as 1993, programmers had realized that HTTP recorded the interactions of users on websites in a special archive called a log. When a visitor lands on a page, he makes “requests” to a server to access certain files. Those request are stored in a log file. A crafty analytics program can parse the requests stored in a log file to determine the number of “hits” on a web page.

According to the Oxford English Dictionary, the first recorded appearance of “hits”—in a computer science context, of course—was in Charles Sippl’s Computer Dictionary and Handbook, published in 1967: “In file maintenance, the finding of a match between a detail record and a master record.” The OED formally defines a computer “hit” as a “match” or “the percentage of records in a file which are accessed in the course of a processing task.; also used analogously in other computing contexts (esp. memory caching).”  Today, the vernacular “hits,” as in, “my blog got 987 hits today!,” means something a little different. Instead of a simple record of interaction, a hit refers to a page view. With the advent of more sophisticated browsing instruments, like proxies and dynamic IP addresses, log file analysis became obsolete. In the place of log files, Javascript tags allowed analytics providers to more accurately track user behavior.  Although Google bought Urchin’s log file system, it opted to develop its free service with a Javascript snippet.

The slippage of “hit” from a log file context into a more casual reference to page views reflects the rise of Google Analytics. We began to fetishize the page view because of log file logic, and even though we have moved beyond that technology, we still hold “hits” supreme. 

GA Was A Called Shot

Google Analytics disrupted the web analytics industry because it made previously premium services available to all. Paul Muret, then a Google engineering director and one of Urchin’s founders, told Business Wire on November 14^th, 2005:

we want to give all online marketers and publishers access to powerful web analytics to help them better understand what their customers want. With this knowledge, businesses can create more accurate advertising and build better websites. By making this powerful service free, we aim to give all websites—large and small—the tools they need to better serve their customers, make more money, and improve the web experience for everyone.

In combination with Google AdWords and search engine optimization strategies, Google Analytics fulfilled Muret’s promises. All those tempting but tired lists, like top ten ways to peel an orange or change your motor oil, seven best Tony Danza cameos, or 27 rap songs from Portland? The segmentation of content into advertising rich and search engine sticky sections (as in this article)? The proliferation of internal links (again, here, here, and here)? These content innovations, not so subtle shifts in the way media gets made, are the children of the Google Analytics era. Analytics is not just retroactive knowledge; that is, it does not only tell you about past behavior on a website. Rather, it allows designers and content generators to make inferences about which content is likely to drive the most traffic, produce the most hits, and most importantly, make the most profit. Yet, Google Analytics set its own expiration date. Once everyone had Google Analytics, everyone had essentially the same competitive advantage. Armed with equal data, developers could differentiate themselves only on the basis of insight and web smarts. A once radical idea, a new way of handling statistics, has become so mainstream that it is a prerequisite to competition in the marketplace—but it no longer constitutes any meaningful edge.

Enter Sabermetrics 

Bill James has become something of a baseball myth. While working the night shift at a pork and beans factory, he started publishing The Bill James Baseball Abstracts, which approached complex baseball problems—like “how can you tell who is a good fielder” (edition the first, circa 1977)—from an elegant statistical perspective. The Bill James story has been so readily mythologized because it is a variation on a classic American trope: average Joe, secretly brilliant, labors in obscurity and after terrible tribulations, achieves recognition and success: which in turn adheres to a more general structure of heroism: mysterious man brings knowledge and power to the world, is rejected, and finally triumphs. See Moneyball for a transformation of the Bill James myth into the Billy Beane myth. (Britney Spears or “the pop star” is another example.) “Sabermetrics,” the field of statistics that James cooked up wandering around the cannery after-hours, is now accepted by the powers that baseball be; in effect, the sabermetrics approach, which privileges the value of players, measured in runs or wins, over conventional stats, has become conventional wisdom. Having incorporated marginal elements, the mainstream swallows the disruptive and innovative capacity of the outlier. The man, no longer radical, slipped into legend, the stuff of aspiration for sports geeks. As with “outsider art” that gets wall space in major galleries, there is a point at which a dominant system absorbs its counter-narratives. In the moments before incorporation, the outsider achieves greatest disruptive potential. In the aftermath, the revolution is first a memory, and then a tall-tale. 

Google Analytics Is Not Sabermetric

Given a convoluted analogy between baseball and the Internet, Google Analytics are not the equivalent of sabermetrics. No matter how “disruptive” contemporary analytics platforms might color their products, they have never reoriented our fundamental conceptions of Internet traffic. We still slavishly worship the hit. In our analogy, conventional analytics providers, like Google Analytics, correspond to the most traditional scouts and managers, who are the fans of baseball card stats like batting average and earned runs average. What are the baseball card stats of the Internet? Page views and visitors.  They’ve been around forever, we collected and compared them in our teens—“how’s your blogspot doing these days?”—and now, they are a comfortable yardstick for web site performance. Not surprisingly, advertisement revenue is typically paid out in impressions and clicks, rough correlates for “hits.” It seems as though the idea of a “hit” has been so thoroughly entrenched in our analytics frameworks that we cannot reimagine Internet behavior. 

Unfortunately, hits, page views, and visitors are not particularly useful ways of thinking about user interactions with web content. If the goal of a website is quality traffic—sustained user engagement—then volume of traffic is a poor measure of performance. Although joining the 3,000 hit club is an impressive feat for a baseball player and indicates an increased likelihood for other offensive records, it is not a reflection of direct contribution to a team’s wins. If all those hits came with no one on base and were followed by strikeouts, then they’re all vanity, all personal glory, no practical value. Hits alone, sabermetrics tells us, do not contribute to wins. Likewise, Internet hits alone do not contribute to sustained user engagement, wins for advertisers, or wins for e-commerce. Advertisers need new ways of understanding how to extract value from their hosts—and content generators, the host bodies, need new ways of understanding how to provide value to their advertising partners. In such a mutualistic relationship, it is the advertiser who usually loses. Imagine a world in which bees pollinate flowers but obtain no nutrition in return. The asymmetric exchange of resources between content generators and advertisers will be rectified by the next iteration of analytics tools.

Publishing Sabermetrics

What will the sabermetrics of web publishing look like? Baseball sabermetrics both defined new types of statistics like the “ultimate zone rating”—a measure of fielding performance—and reconfigured old statistics, as with “runs created,” a complex combination of various baseball card stats into a more direct estimation of value. At Parse.ly, we’ve developed some unique statistics, like “momentum,” which helps publishers understand the flow of traffic into their site. We’re also extracting information on topics—what an article is about—with a little help from natural language processing. And the simple addition of real-time data—the lack of which cripples Google Analytics—is changing how publishers think about content distribution. Eventually, user behavior analytics, like YouTube’s audience retention statistic and new e-book analytics platforms that evaluate reading preferences, will supplant Google Analytic’s baseball card stats. In combination with those conventional metrics—the “hits” paradigm—user behavior analytics will grant advertisers a better understanding of how web sites contribute value to campaigns. Although the “hits” paradigm has defined advertising value thus far, web analytics has the potential to demonstrate deeper and more profit driven value to advertisers. Subsequently, digital publishers will need to produce content that persistently immerses and engages readers, exposing them to advertising systems more efficiently. 

• Tags
• analytics
• baseball
• history
• sabermetrics
• Google
• Google Analytics
• Dash
• Parse.ly
• innovation