Login is restricted to DCN Publisher Members. If you are a DCN Member and don't have an account, register here.

Digital Content Next


InContext / An inside look at the business of digital content

Shakespeare, meet Einstein—Making the leap from data insights to artificial intelligence for content marketing

March 27, 2019 | By Jon Shalowitz, CEO—LiftIgniter @jonshalowitz

There isn’t a company leader out there today who doesn’t realize that their ability to harness and interpret data will make or break their business. When it comes to data analytics, the bar is constantly being raised. What was a theoretical concept just one year ago is now mainstream practice in major tech companies. Leveraging data to become more AI-driven is (or should be) on every CEO’s mind.

In fact, research suggests that by 2020, 30% of digital commerce revenue growth will be attributable to Artificial Intelligence (AI) and analytics.

As a content marketer, imagine if you were able to microsegment your readers in the moment and understand which content to best show them to drive subscriptions (or continued subscription). If that sounds appealing to you, read on.

First Party Data

Just look at the renewed importance of first-party data (data you own). In the past, third-party data, or data collected from aggregators, was considered the best way to market  to customers. However, privacy concerns (it includes things like purchase and browsing history) have made it less viable.

First-party data, on the other hand, includes website traffic statistics, owned email marketing data, e-commerce data, and the content you publish and promote on your site. The best part is that it’s completely within your control. Also, it builds on itself over the years, which gives you deeper insights than you could have obtained through third-party data alone.

Today, content companies that leverage first-party data are better-positioned to make content recommendations and provide better customer experiences. But to give your customers, users, and readers what they want, you need to be able to process and react to data in less than one second.

And assembling, analyzing, and interpreting the data is no small feat.

From Data to Intelligence

According to 2018 data from Nielsen, Americans spend more than 11 hours consuming various forms of media. And adults between the ages of 18-34 spend as much as 43% of their time reading and viewing media on various digital platforms—a number that’s almost certainly higher today.

All of which is to say: There’s more first-party data up for grabs than ever before.

However, today’s users also expect that content to be relevant. That’s why the most effective content providers not only need to provide content that readers care about, but also content that’s highly tailored to their experiences in that moment.

And in the era of digital transformation, where AI-driven businesses are poised to take in close to $3 trillion, you need a data strategy that performs efficiently.

AI, combined with traditional analytics, can help your company comb through huge amounts of data in real time. And every day you put it off, the data set grows even larger and harder to manage.   

Here’s how to harness the power of the data you already have and deliver the content your users want:

Data scientists are often too bogged down with grunt work to focus on what matters.

Today’s tech clutter (the average business today uses roughly 129 apps) creates data silos that make effective communication nearly impossible.

According to a Deloitte survey, 52% of respondents said they couldn’t make information-backed decisions because the data they needed is trapped in siloed departments. When data is isolated, it’s hard to wrangle and figure out how to analyze it. So data scientists find themselves spending 80% of their time on mundane work like figuring out which data to use, managing data movement on and off the Cloud, and understanding what attributes make the most sense to use in your models.

And after all that grunt work, there just isn’t time to do the high-value analysis.

So when you ask your data science team to tackle a portion of data, you may not understand there are many different moving parts that have to come together besides the analytics. And just because your data scientists have PhDs in mathematics or statistics doesn’t mean they have the expertise in data engineering and management to get it all done.

All CEOs need to make sure there’s a programmatic and structured way to enable data science to be ingrained in the data culture—not an afterthought. This means prioritizing building and updating the infrastructure so that there’s time for the most important analysis.

A strong data science platform makes room for truly valuable analysis.

A lot of the businesses we work with at my customer engagement company, LiftIgnigter, say it’s the operationalizing or the engineering part of data analysis that’s the most time-consuming.

It is easy to become overwhelmed by the challenge of turning data into an integral component of business operations—companies often don’t know where or how to begin.

To access the value of your business’s data, your data science team needs to develop not only a clear data strategy, but also find a programmatic way to make this second nature. Typically, this requires either building a platform or leveraging a third-party platform. A third-party platform helps data science teams aggregate the data, ultimately making data analysis a much more democratic process.

But it’s equally critical that the platform is built and fine-tuned for your company’s specific needs. A generic platform may give you lots of different options to run model testing, but at the end of the day, you need to focus your effort on modeling that is the most likely to yield meaningful insights.

However, with the right platform, which typically combines data science and machine learning, a team of five people can be effective as a team of 50. Imagine how much more efficient your team would be if they could focus their time on higher-level analysis.

AI augments the work your data science team can do, in real time.

With AI becoming increasingly ubiquitous, people are worried it will take their jobs. But AI isn’t replacing the data scientist. It’s actually making them that much more effective.

Here’s why: Training user targeting models is one thing. But then your team has to figure out a way to operationalize the continuous updating of the data.

For example, a large publisher company with 40 data scientists can still take several days to do all of the analytics that their editorial or product teams are demanding in real-time. They can hire 1,000 people—but that’s not cost-effective. The data science team is often called on to find compelling nuggets in your data in real-time. They’re constantly bombarded by your revenue team and marketing teams to come up with the perfect segment that will help drive greater revenue.

Large media companies like Facebook have a keen advantage because they have armies of data scientists to dig through their mounting data and figure out what’s relevant. And this is where AI comes in.

In order to compete, your business needs to make that process as standardized, real-time, and with zero wasted effort.

Third-party platforms enable businesses to create, tune, and refresh high performing models efficiently. Whether it’s generating demographic data for audience creation, or live streaming user behavioral signals to train a targeting system, a third-party platform can do a lot of the heavy lifting for you.

If you can create a unified strategy across your organization, optimize the data you have, and work only with trustworthy third-party partners—you’ll be prepared to navigate this new landscape.

Liked this article?

Subscribe to the InContext newsletter to get insights like this delivered to your inbox every week.