Creating a Framework for Digital Media Success

Chaotic. Unpredictable. Disruptive. No matter how you view the media and entertainment business, there’s one thing it’s definitely not: boring. The digital megatrend has created a seismic shift across the industry with continuing aftershocks, affecting how media companies approach the creation, production, marketing, distribution, and monetization of content.    

More people than ever are consuming content using digital devices:

• More than three-quarters (76%) of people in the US and UK watch video online every week
• More than 68% of people in the US and UK subscribe to a video-on-demand service
• Digital content revenues are predicted to reach $180 billion in 2017, up nearly 30% compared to 2015

Meanwhile, television is losing ground to digital as the latter is expected to surpass the former as the top advertising medium in 2017. And the print newspaper industry continues to see declines in circulation and revenues. Combined print ad spending in newspapers and magazines is expected to fall by 14% in 2017.

As new ways to consume media and entertainment continually enter the market, born-digital companies—the digital disruptors—are going head to head with digital re-inventors (those stalwarts in traditional media businesses that are shifting to digital) in the battle for audiences and ad revenues.

For all media companies, digital success is imperative. But it’s not impossible. Especially if you take advantage of the one unifying force for managing digital media businesses: data. You need data to improve customer experiences and drive ad revenues, make sure you’re ready for major events, and empower your team and your stakeholders by demonstrating the cultural shift you’re leading is delivering tangible results.  

This ebook gives IT leaders and their business counterparts in media companies a data framework for creating a performance culture to drive digital success.  

If you can’t beat them, join them. Digital disruption has forced media and entertainment companies to embrace digital transformation as a strategy to compete against the industry upstarts, adapt to continued device proliferation, and deliver on changing consumer expectations.

Digital reinvention is all about attracting new customers, generating more revenue, and creating new revenue streams. But achieving these goals requires lightning fast time to market and top quality content and delivery, all while reducing costs.

To meet these difficult challenges requires various stakeholders—business, product, marketing, content creators, developers and operations teams—to collaborate and innovate. They have to work together to identify new approaches that lets their company quickly and securely produce and deliver the right content, at the right time, to any device.

Here are some of the top digital strategies that media and entertainment companies are putting into action:

Cloud-based digital workflow 

• Move data collection, storage, and management from physical media to the cloud to reduce costs and accelerate time to market with new content 

Multi-channel delivery 

• Create individualized customer experiences that provide the right format to the right platform for the right audience 

• Use newly created, library-stored, or licensed content to attract new customers 

• Create new over-the-top (OTT) offerings to take advantage of new customer habits as they are being formed 

Direct-to-consumer relationship 

• Combine multiple technologies (cloud, mobile, Internet of Things) to develop new types of content and ways to engage with audiences 

• Forge and maintain direct relationships with consumers 

• Attract new audiences

While everyone agrees that content is still king, the digital shift has vaulted all aspects of the creation, presentation and delivery of content into the spotlight. This intense focus on the technical components that deliver the consumer experience means that companies need to understand at any given moment how their content assets and delivery infrastructure are performing.  

Measuring and tracking digital performance centers on data. Performance and usage data—on content, applications, databases, infrastructure, and more—becomes the common language to collaborate across your business workflows—content creation, ingestion, processing (encoding, transcoding, security, and more), distribution, and playout/delivery.   

To get this data-driven approach right, you need to connect key performance indicators (KPIs) from across the business with the front-end customer experience and backend applications and infrastructure. And of course, you need these metrics to be available in real time and accessible to all the different stakeholders in the business.

Let’s take a deeper look at the metrics you need for each of the three major digital strategies we described earlier.

Leading the transformation to digital

Fairfax Media Limited owns more than 300 websites which see approximately two million visitors each day. The company has been in the forefront of the transformation of media to digital channels. “Everything is going digital, which means that we are constantly building more digital products,” says Cheesun Choong, head of product platforms at Fairfax Media. “We need to do software and technology really well to survive in the media space.”

Fairfax understands that its websites and services must be available and performing well at all times. “Our infrastructure must be running 100% of the time,” says Michael Lorant, senior systems engineer at Fairfax Media. “There is no room for downtime in this business.”

Application and Infrastructure Metrics

As media companies consolidate brands and rationalize their architectures, they are increasingly adopting the cloud to enable nearly unlimited scalability and elasticity. To take advantage of the benefits of the cloud, organizations are re-architecting their applications and choosing modern programming languages such as Node.js, PHP and GO. The cloud also makes it easier to deploy newer, more dynamic architectures such as microservices.

As cloud-based applications become more complex and the lines blur between application performance and cloud infrastructure, it is critical for engineering and DevOps teams to have integrated visibility to manage the availability and performance of critical customer-facing systems. Measuring and tracking the following KPIs will help you do that.  

Apdex 

• Description: The measure of response time based against a set threshold. It measures the ratio of satisfactory response times to unsatisfactory response times. 

• Why it’s important: Apdex provides a single value between 0 and 1 which gives a concise indication of how many of your customers are experiencing slow response times while engaging with your web application. It is calculated by setting a threshold that is considered ‘acceptable’ and by weighing and counting users who experience wait times that are far in excess of that range. 

Availability 

• Description: Measures your site’s uptime. It’s the ratio of throughput (calls per minute) and errors per minute for the selected time period for the service. 

• Why it’s important: If your application or your content isn’t available, you’re not meeting audience expectations and they’ll look elsewhere, costing your company ad revenue and de-valuing the brand. 

Throughput 

• Description: The amount of traffic a particular service is experiencing. It is requests per minute. 

• Why it’s important: An unexpected dip in throughput is often a signal that there’s an issue, such as consumers not being able to reach your site or they’re navigating away in frustration. 

App/API/DB response times 

• Description: Response time/latency duration of a transaction from the perspective of the requester, and not necessarily the total sum of all time spent within a transaction. 

• Why it’s important: While Apdex represents concise aggregates about your application’s response times, you can also drill into very granular views of individual actions as well. 

Error percentage 

• Description: Percentage of application errors over a period of time. Along with details on the error types, related frequency and diagnostic traces. 

• Why it’s important: An error is an unhandled exception which terminates a request. This often means that your customers were directly impacted and experienced some sort of degraded experience. Being able to quickly understand how frequently and where errors are manifesting is one of the most integral parts of maintaining a high quality user experience. 

Memory footprint 

• Description: Amount of memory being used by the service.

• Why it’s important: High memory usage can contribute to slow page-load times, which frustrates consumers and could cause them to navigate away from your site. 

CPU workload 

• Description: Amount of CPU cycles being used by the service.

• Why it’s important: High CPU usage can also contribute to slow page load times or even make your site unresponsive. 

Disk workload 

• Description: Amount of time the disk was not able to keep up with the read/ write speed. 

• Why it’s important: For disk-intensive applications (e.g. apps which write files, upload pictures, or read data from local storage), it’s important to know where your bottlenecks are. Further, if your memory is at high usage, your application will page to disk, which is extremely non-performant. Disk workload will help you immediately troubleshoot these common, low-level issues. 

Network throughput 

• Description: Amount of data moved successfully from one place to another in a given time period, and typically measured in bits per second (bps), as in megabits per second (Mbps) or gigabits per second (Gbps). 

• Why it’s important: In today’s microservices-centric world, almost every application is dependent on some other application (which could be internal to your organization, a third-party API, or a mix of both). You should know how much time your app spends waiting on data to move across the wire (and where the data is going and coming). Network throughput will help you understand your application’s healthy communications and quickly troubleshoot when something’s gone wrong. 

Note: You’ll want to be able to filter or group the above KPI metrics by: geography, business unit, product, user type, user device, application name, programming language, or cloud provider.

The big event

Live events influence how customers view your brand. For significant events such as the Super Bowl or a presidential election, you need to make sure your applications are available, scalable, correct and fast.

For instance, you might plan to elastically scale your cloud infrastructure, choosing to implement Kubernetes before the big event to facilitate change before and during the event as well as for its container-centric auto-scaling.

For the event, you’ll want visibility into the performance of Kubernetes and to understand how many masters and workers are running, along with their CPU and memory consumption. Moreover, you’ll want to view the Kubernetes performance in context with the throughput and response time for key transactions overall. Tracking KPIs through a performance monitoring solution can help you do that.

Customer Experience Metrics

While the core premise of a multi-channel delivery strategy is providing the right content format to the right platform for the right audience at the right time, it’s the customer experience that will determine success. To that end, the most important KPIs to measure customer experience will be those that tell you how well your front-end, customer-facing applications and services are performing.   

For instance, page load times are inversely proportional to ad viewability and revenue. That’s why many media companies are adopting single-page applications to help improve ad viewability. This makes the front-end application a critical code-processing tier, in some cases contributing to more than 90% of the total response time. As such, the same type of rigorous optimization that occurs for backend code must be applied to the front end.

Simply put, you want your content and your ads to be viewable as quickly as possible. Therefore, while it’s important to measure overall page load time, it’s equally important to monitor key components of the page load such as DOM Ready and Page Render. Here’s an explanation of the most important customer experience KPIs to focus on for your front-end applications.

Session count 

• Description: A group of user interactions with your website that take place within a given time frame. For example a single session can contain multiple page views, events, social interactions, and ecommerce transactions. 

• Why it’s Important: Provides an indicator of the amount of usage and engagement on your site. The trends you see in a session count can indicate if a particular event or story is driving interest. Measuring page availability and load time compared to session count is a good way to understand if your site performance scales. 

Session duration 

• Description: Time spent by a user in a given session. 

• Why it’s Important: Shows you how engaged users are on your site. Combined with page views per session, this can provide an indicator of how compelling content is (or is not). 

Page views 

• Description: The number of pages viewed in a particular time period. 

• Why it’s Important: Helps you understand the level of interest in a particular page. This can help you measure the effectiveness of marketing campaigns, the audience for a special event, and more. You can correlate page view to page speed or site performance. Additionally, you can calculate the number of ad impressions per page view and how many impressions are lost when performance degrades. 

DOM (Document Object Model) readiness 

• Description: A popular measure of page timing because it indicates the minimum amount of time before the user sees the page. Single-page apps avoid a DOM reload for each page view (or route change). 

• Why it’s Important: Useful to measure over time to understand if you are reducing page load time by improving page templates, going to single-page applications (SPA), etc. 

Page render 

• Description: The time between the DOM being complete and the window-load event. This phase measures browser-side processing of the page content, and often includes time for scripts and static assets to load. 

• Why it’s Important: Indicates how long it takes to render the elements of the page that are visible to the user. You can further segment this to onload time, which is an indication of how long it takes the visible portion of a page to load. Use the trends of these numbers to measure the success of optimization efforts. 

Error percentage 

• Description: Percentage of application errors over a period of time. Along with details on the error types, related frequency and diagnostic traces. 

• Why it’s Important: An error is an unhandled exception which terminates a request. This often means that your customers were directly impacted and experienced some sort of degraded experience. Being able to quickly understand how frequently and where errors are manifesting is one of the most integral parts of maintaining a high quality user experience. 

Apdex 

• Description: The measure of response time based against a set threshold. It measures the ratio of satisfactory response times to unsatisfactory response times. 

• Why it’s Important: Apdex is an industry standard for measuring and reporting application performance in the context of actual end-user experience so it provides one uniform measure of performance across multiple applications. 

Note: You’ll want to be able to filter or group the above KPI metrics by: geography, business unit, website name, page name, device type, browser type, browser version, or error type.

Loading a single-page application quickly

A major media company was moving to an infinite scroll, single-page application. The early versions of the site had an average page load time of more than 9 seconds and roughly 28% of page loads had JavaScript errors.

After employing a targeted optimization effort with New Relic, the media company was able to get page load times down to 2 seconds and the percentage of JavaScript errors to just 3%.

Video Quality of Service and Engagement Metrics

Almost everyone loves video, just ask YouTube, the second most popular site in the world (second only to Google).

For many media and entertainment companies, video is the top revenue producer:

• User engagement rates and ad revenue are both higher for video content

• Short-form video is the content of choice on social media and mobile devices

• Long-form, OTT video is an increasing percentage of the content watched on TVs (the “best screen available”)   

That’s why it’s critical for the success of your direct-to-consumer strategies to make sure your video experiences are meeting customer expectations. If you’ve followed the advice of the previous sections, you have measured and optimized your backend infrastructure and front-end page performance. Now you need to measure and optimize the customer experience at the video player and video levels.  

It’s essential to measure how long it takes the video player to load on the page, exits before video start, video starts, ad plays, errors, bitrate, buffering and more. Having this visibility every day (and during big events) allows development teams, operations, and product teams to collaborate to deliver stellar customer experience and, as a result, the desired business outcomes.

Video play count 

• Description: Number of times a video is played. You can set a threshold for how many seconds a video needs to be played before it’s counted. 

• Why it’s Important: Understand your most popular content to optimize delivery (e.g. rendition sets), improve customer engagement (put popular content in the carousel), and increase content sharing across brands. 

Video play duration 

• Description: Amount of time a video is played. 

• Why it’s Important: Are consumers watching the entire video? Your content teams need to know if consumers are losing interest half-way through the video. 

Video bitrate 

• Description: Bitrate of video being played at the time the sample is taken. 

• Why it’s Important: If the bitrate decreases, your video quality degrades, which impacts the quality of the experience for the consumer. 

Video buffering time 

• Description: Amount of time video spends buffering during feedback. 

• Why it’s Important: Because buffering time directly correlates with audience satisfaction, you’ll want to understand which geographies, networks, devices, and more are experiencing re-buffering. 

Video success rate 

• Description: The rate at which video plays successfully. 

• Why it’s Important: Problems with the video player can affect the customer experience and ad impressions negatively. 

Errors (video, ad) 

• Description: Number of video errors and number of ad play errors. 

• Why it’s Important: Did the player/stream fail to load? Did the ad stream generate an error? These are also critical aspects of the consumer experience. 

Buffering annoys consumers the most

A recent survey showed that more than half of respondents cited video buffering during playback as the most annoying aspect of online video viewing. The survey also reports that 48% of people will stop watching a video if it re-buffers twice, and 78% will stop watching if it re-buffers three times.

Once your backend media workflows are optimized and your frontend and video performance are great, then your teams can collaborate with the marketing organization to improve user engagement. For instance:

• Page design decisions can be made in real time to take advantage of trending content.

• Carousel content can be targeted by region and demographic.

• Customer journeys for trial and subscription signup can be scrutinized to improve conversion rates.

Active users 

• Description: Daily active users (DAU) is the total number of users that engage in some way with a web or mobile product on a given day. To be considered active, users simply have to view or open the product. Monthly active users (MAU) is the aggregate sum of daily active users over a period of one month. 

• Why it’s Important: Indicates the level of adoption of your service. Measure active users against your sales projections. Also, use the trends of this number to measure the results of optimization trial/buy processes. 

Content views 

• Description: The number of times a piece of content is viewed. 

• Why it’s Important: The greater the viewership, the more popular the content is. Your content team can then create more content in the same genre, style, and so on. 

Session duration 

• Description: Amount of time a user spends in a session. 

• Why it’s Important: The more time consumers spend on your site, the more content they consume and more ad impressions you get. 

Note: You’ll want to be able to filter or group the above KPI metrics by: geography, business unit, product, user type, content genre, or media device. 

Measuring and tracking the right metrics requires end-to-end intelligence across all the areas discussed earlier: application and infrastructure performance, customer experience, video quality, and user engagement. Getting the right data isn’t difficult if you have digital intelligence capabilities that include: real-time analytics, full-stack visibility, and scalability to support even your busiest day.  

New Relic is the only full-stack performance monitoring platform that gives you complete visibility into your digital media business. From your CMS to the CDN, to the page and the player, New Relic provides end-to-end monitoring and analytics that connect your authors to your audience with lightning fast page loads and mobile app performance, along with bufferless, end-to-end video and audio performance.

With New Relic, you can quickly identify, isolate, and fix problems whether they are in backend code, external services or scripts, third-party mobile app, or an ad provider.

Why New Relic for digital media companies?

• SaaS-powered platform

• Future-proof language and framework support

• Application, audience, and outcome-based analytics

The media and entertainment industry is undergoing fundamental changes in the way content is created, distributed, consumed, and monetized. Competing and winning on the rapidly shifting digital playing field requires a commitment to delivering the best possible customer experience.

Data is the lifeblood of a performance culture—and nowhere is that more true than for media and entertainment companies. New Relic can help you get the data you need to demonstrate the success of your digital strategies and lay the groundwork for continued innovation.

“New Relic is easily the most straightforward and all-encompassing monitoring tool I have ever used. It not only provides insights into how the application is performing, but shows exactly where in the code, database, or caching layer a potential bottleneck could be.”  – Jim Mortko, Vice President of Engineering, Hearst Corporation