The Need for Topic Modeling

At ComBlu, part of our approach to content strategy includes an insights process involving a deep dive on topics. One interesting trend I have noticed in the last six months is an increase in the number of clients certain they know the topics in high demand. However, after deeper conversation, we often find that their methods and perceptions can be limited. Basing a content strategy without a clear and multi-dimensional understanding of high-value topics is like building a house without a foundation. It just doesn’t work.

Delivering assets that meet the information needs of your reader, viewer or customer is essential to the content value proposition. To determine what topics they are most interested in, you can’t just look at the content already offered and see which topics get the most downloads. It’s too internal of a view—and you are limiting the total scope of topics. A complete approach to topic modeling requires inputs from multiple sources, including:

• SEO reports
• Listening reports
• Marketing plans
• Market research
• Interviews with internal experts
• Hash tag analysis

Each one of these sources brings in a different perspective. SEO reports can provide information on keywords and volume and thereby one aspect of popularity. Listening provides a great way to look at the terms and language people are using and dive beyond keywords to learn what aspects of an issue or topics are of most interest. Listening also helps to see what concerns or questions are hanging out there unresolved—highlighting unmet content needs.

A marketing plan is critical to consults, as they let you know key campaigns and topics on which they focus. Marketing plans also let you know how to time your content waves to support planned initiatives. Market research can provide additional information on the expressed information needs and purchase-drivers of key audience segments.

Another great source of information is internal experts. Frequently we find talking to product development and sales teams can yield a very specific list of topics on which customers are already asking for content. Hash tag analysis tracks the frequency a hash tag is used on Twitter and provides details about the hash tags that are most popular with your key audiences—two great sources for high-volume topics.

Through analysis, you can layer each of the information elements to get a clear (and sometimes surprising) perspective. As you consider the next update on your content strategy, take a critical look at your topic modeling process. A process that looks well beyond your branded and owned assets will reveal a broader view of what content your customers want most.