A Guide to Big Data Tools for Publishers


BEST PRACTICES SERIES

Article ImageAfter a period of relentless hype, clarity is emerging on use cases for Big Data, and the toolsets are also showing signs of maturity. Digital media publishers may have been slow to join the Big Data party, but it is not too late. This article provides a quick overview of the Big Data tools for publishers, and the focus here is not on internal business intelligence use cases, but on the content-creation side of the business.

Publishers care about two main stakeholders-readers and advertisers. In catering to them, three factors are paramount:

  1. What content to create
  2. How to personalize the content for the reader 
  3. How to effectively monetize the generated traffic

Let's look at the Big Data tools and techniques available to address each one of these issues.

Content Creation

Your content strategy, which is based on the needs of your target audience, mainly drives decisions around what content to create. To a large extent, cracking this part of the puzzle is more of an art than a science. You rely on editorial expertise and sound journalistic judgment-tools only have a limited role to play here. However, there are several tools to inform and aid your decisions.

Of course, there is web analytics software that lets you discern important information such as trends in traffic and content consumption patterns. This technology is reasonably mature and examples of tools are Google Analytics, Webtrends, Adobe (Omniture), and IBM (Coremetrics). Typically, these tools incorporate data from mobile devices and apps as well.

Readers are also likely to appreciate it when relevant advertisements and offers are targeted at them rather than irrelevant ads.

In addition to arming yourself with knowledge of the readers of your own digital properties, you also want to get a feel for the zeitgeist, or what's going in the wider world outside. Social media monitoring and intelligence tools can help you with that. The tools are relatively new, and examples include Chartbeat, Spike, Sprout Social, and Hootsuite. It is beyond the scope here to dwell on the functionality of these tools; think of them as radars to track key current events on the horizon, including social media trends, content popularity, how readers interact with your content, and some high-level competitive intelligence. All of them can be valuable inputs into your editorial and content decisions.

Content Delivery (aka Personalization)

Content is king when it is both high-quality and highly relevant-which is achieved by targeting the right content to the right audiences (i.e., personalization).

Before Big Data, personalization approaches ranged from the basic (e.g., "Most Read" or "Most Emailed") to content recommendations based on predefined rules around content and visitor attributes. Many web content management (WCM) systems (such as Adode Experience Manager, Sitecore, and SDL Tridion) come with such recommendation modules (typically, separately licensed from the WCM).

Another approach to personalization relies on collaborative filtering (i.e., recommendations based on what content users who are similar to you have consumed). Here, the publisher's data on its users is combined with third-party user data for content targeting. Tools such as Cxense let you do this.

In addition, you also have tools that let you do multi-variate testing (a subset of this, A/B testing, is a more popular moniker) of different content elements to empirically validate and generate content that is optimized for different target audiences. All of this can be done dynamically while the visitors are interacting with the content. Examples of those tools are Adobe Target, Maxymiser, Optimizely, and SiteSpect.

Revenue Optimization

Publishers typically have user data based on activity on their sites, but there is greater value to be had in combining such first-party data with third-party user data to create richer user profiles and granular audience segments. Data management platforms (DMPs) let you just do that. You can use DMPs for content-targeting as well, but most DMPs are currently used for ad-targeting. Brands and marketers prefer publishers that are able to help them reach very narrowly defined audience segments and may even pay a premium for such placements. Using a DMP, publishers are able to create niche and micro-segments of audiences that marketers increasingly prefer. 

Readers are also likely to appreciate it when relevant advertisements and offers are targeted at them rather than irrelevant ads. However, taken to the extreme, this can feel creepy; user privacy concerns should be respected. Examples of DMPs include Adobe Audience Manager, Lotame, Krux, and Oracle Bluekai.

In closing, publishers have very specific use cases around content. A single tool does not address all of them. But smart publishers can leverage a variety of Big Data tools and techniques-from content, marketing, and advertising technology domains-to create compelling content, grow their audiences, and increase their revenues.    

(Image courtesy of Shutterstock.)