Is getting content “just right” a matter of art or science? Actually, it’s a bit of both. It requires skillful editorial judgment combined with an experimentation mindset. How can content teams get these two capabilities to support each other?
Michael AndrewsPublished on Mar 26, 2020
Structuring content gives writers and editors the ability to refine their messages and lets marketers and business users measure the outcomes of editorial changes and revisions.
Content models enable content items to be fine-tuned. The structuring of content allows different parts of an item to be changed easily.
Writers and editors face choices about how to position their content—the wording to use and the messages to include. And their colleagues in marketing and product roles want to know how these decisions are influencing outcomes. Everyone shares the desire to have content perform as strongly as possible.
Let’s consider a couple of common issues that content teams must decide:
- Which content items should be highlighted to get the best results?
- What parts of a content item should be changed to get the best results?
Content models provide the missing link between editorial decisions and the measurement of performance.
Content models enable experimentation, which has become an essential activity. As Ron Kohavi and Stefan Thomke have noted in the Harvard Business Review, “At a time when the web is vital to almost all businesses, rigorous online experiments should be standard operating procedure.”
Accentuating the Right Mix
Certain content types are aggregations of other types. A “homepage” is the classic example of this, providing links to various items. But many other content items also feature information that is available in more detail elsewhere. Editors must choose which items to highlight. They curate information from other content items and need to decide what items to feature that will generate the most interest from audiences. What mix yields the best results?
Curating which items to promote is like deciding the right mix of ingredients in a recipe. What ingredients to use? The content model indicates what kinds of items can be linked. Editors can choose different items to see which ones generate the most interaction from audiences. The models support analytics on interaction with items.
Editors can optimize which items should appear together. The performance of items can be influenced by their context: where they are placed and which other items they appear with. For example, an item might perform well within an email but not on a homepage. The content model makes changing the context of where the items appear easy to do. The possibilities to experiment are extensive.
Tweaking the Message
Now let’s shift from considering which items to highlight to what those items should say.
Individual content types can be structured by elements that address specific subtopics or messages. When content items are structured, teams can optimize the wording and information in specific elements.
Content types can use a variety of structures. Much content is based on common structural patterns, such as introduction, main body, and conclusions. They may incorporate widely-used elements such as a list—for example, a list of supplies or a list of steps. Other common patterns are:
- Pros and cons
- Problem and solution
- Issue and example
Teams may want to structure their content types to reflect the specific goals they have. Whether the structure they use is a custom or a common one, they can tweak the elements.
Martin Cutts, in his book The Oxford Plain English Guide, introduces a content structure he calls the SCRAP model (for Situation-Complication-Resolution-Action-Politeness). The model is used to communicate information to individuals. The elements are:
- Situation (The triggering event: why the reader is getting the message)
- Complication (The problem to resolve)
- Resolution (The suggested remedy from the sender)
- Action (The proposed CTA for the reader)
- Politeness (The closing message from the sender)
The SCRAP model could be used to structure a content type for an email. The enterprise needs to notify customers when an order item is out of stock and needs to know if the customer wants to wait until the item is back in stock or cancel the order.
Such notifications are recurring and need to be handled carefully (since we’d expect that the customers will be disappointed). With this kind of business-critical content, it can be useful to test different versions of the message. Perhaps the customer’s attitude is influenced by how the Situation or the Complication are worded. Or perhaps changing the wording of the Action will influence whether customers choose to cancel their order. Alternatively, the wording may not be critical; instead, the terms of the resolution might be, such as whether shipping fees get waved if the customer waits.
Because the content type is structured with elements, each of these aspects can be changed and evaluated. Enterprises can make editorial changes and measure different outcomes using whatever analytics or optimization tool they choose.
Planning for Editorial Experiments
The scope for experimentation with content choices is broad. Yet only some enterprises experiment with their content beyond testing headlines or button text. Make sure your content model can offer your team insight into the dimensions of your content you’d like to test.
Kontent.ai integrates with popular analytics and optimization tools. Look into how your tool supports experimentation. These tools will often have detailed tutorials about how to set up experiments and interpret results.