Content Modeling with Structured Content

The content modeling process in Crystallize starts with determining what kind of content you need as per your use case. The next step is defining the atomic components of the information you are managing and, finally, relations between different types of information.

A component could be a label, an image, GEO location, or a list of technical specifications. Components are given semantic meaning and chunked together, making up the shape of your information. Information modeling is the core of product information management (PIM).

Structuring information enables efficient reuse of content chunks across multiple channels. For product information, you want to future-proof your product information models to serve your website, app, point of sales, digital signage, syndication, and any potential future channel.

As an example of a product content model, let's consider cooking recipes. A recipe could be explained in a single field of rich text. However, structuring, adding semantic meaning, and organizing the content in reusable chunks can open up possibilities for reuse.

By separating out ingredients as a product shape and recipes as separate structured documents that refer to ingredients, you create a flexible content model for many use cases like:

• Being able to show all recipes with a specific ingredient.
• Adding all ingredients to a shopping basket for a given recipe.
• Finding recipes that are made in 30 minutes or less.
• Finding all recipes with only organic ingredients.
• Generating structured data for recipes.

Shapes are used in Crystallize for content modeling. You can define the structure of products and structured documents. Use the right multilingual eCommerce strategies to model your content and product data for efficient reuse across channels and languages.

From an eCommerce perspective, you have an additional layer of complexity to take into account. The product universe, if you are selling cars, is very different from a product universe for food. The basics are the properties that describe your products, e.g. horsepower and weight compared to nutritional contents.

Products often have relations that define related items, compatible items that fits your selected product, bundles of products or even complex product configurators.

Product configurators take into account compatible combinations of a product. It could be as simple as the width of a mattress needing to match the bed frame. More complex product configurators could define usage, e.g., how many meters of fabric are used to calculate the cost of a sofa in a specific combination of modules.

In summary, some of the topics to consider when designing your product universe are:

• Properties, specifications, and descriptions.
• Related products.
• Compatible products.
• Configurable products to create composable products.

For a more in-depth discussion of the ins and outs of content modeling, please check out our which comes with a free downloadable PDF. This introduction to content modeling covers semantic content modeling and channel-specific content modeling. Additionally, you get perspectives from different stakeholders, such as information architects, developers, designers, and editorial users.

For even more on content modeling, check out the Content Modeling Livestream playlist, where in five recorded streams, we go over basic concepts, types of content modeling approaches, composable content modeling, the role of product storytelling with bundles of products and product variants in mind, and modeling out the Tesla online car configurator using the Crystallize example.

To understand content modeling better, check out our Creating online Tesla car configurator Livestream. It goes in-depth about architecting a product page (model) for a Tesla Model 3 to be sold via an online shop. It shows how to implement it in a car configurator that allows your users to personalize their car.