How do you share your information architecture with developers?

Perhaps more important than "models" is the activity of "modeling", where the people involved discuss a concept using a diagram, spreadsheet, or whatever kind of model is needed to help describe what they mean. Modeling enables groups of people to develop a shared understanding of a problem or solution space in a way that is more efficient and has higher accuracy than textual or spoken communication.

I love using Notion to document content models. It allows you to create a central database that links to (1) individual pages for documenting each template/component/etc. and (2) related Figma files, copydecks, and other supporting artifacts.

I used to be a developer so I was able to appreciate the level of detail in the annotations that a company I worked at used, and I adopted it ever since, especially for complex functionality. I try to limit prototypes with multiple versions for conditional functionality--instead, I point to a feature on the final design (these days in Figma) and add the conditional statements in the annotation, which have gone over extremely well with the Dev teams I've worked with. Here's an example: --- (1) Main Image OnClick display image as follows: If selected product = product x then display image1.png; elseif selected product = product y then display image2.png; else display image0.png.

As an SEO, a large mega navigation bar represents a liability more than an aid for users. Its destinations are hidden links that dilute the PageRank (PR) punch of the page. The worst example I saw had 500 links. When the mega nav is present on every page, each page on the site links to every other page. The topical authority of a page, say, about credit cards, would be higher if only other credit card pages were linked to it. While this isn't practical, a compromise is a two- or more-step navigation menu that keeps links from other products further away. I ask the developers to build some prototypes, then the UX team presents them to real users and gets their feedback. Some large businesses use this.

What does the navigation bar comment have to do with information architecture? It seems irrelevant to this topic,in my opinion.

Low-Fidelity Prototypes: Low-fidelity prototypes are basic, simplified representations of your design. They focus on showcasing the fundamental functionality and interactivity of your AI. These prototypes are often quick to create and are particularly useful during the initial stages of development. By sharing a low-fidelity prototype with developers, you provide them with a visual outline of your ideas, allowing them to understand the core concepts and interactions.

High-Fidelity Prototypes: High-fidelity prototypes take things a step further by providing a more realistic representation of your AI's appearance and performance. These prototypes are designed to closely resemble the final product, incorporating visual elements, graphics, and finer details. Sharing a high-fidelity prototype with developers allows them to experience how your AI will look and feel in its finished state, helping them grasp the aesthetics and user interface nuances.

For communicating info architectural decisions, low-fidelity, or structural prototypes—ones that only include the structural aspects of the website or software— are key to not only validating the effectiveness of the structural design but also communicating to the developers. An effective structural prototype for IA should be able to validate these three things: 1) Do I know where I am within the website or software? 2) Do I know what is available for me to do in this (or any other) place within the website or software? 3) Do I know how to find what I need next based on where I am now?

Warning: Regarding summary #5 (Use standards and best practices), every word after the first sentence is incoherent and offers little value to sharing IA with developers. The following is a more accurate statement: Certain IA specifications can be transformed using standard formats like JSON, XML, and RDF. It's also common practice to provide IA specifications to developers in Microsoft Excel and comma-delimited files. For increasingly complex environments, taxonomy and ontology management software, ER modeling tools, and APIs are commonly used to share IA specifications with developers.

There is an important distinction to be made between feedback about the *clarity of the documentation* provided to developers and the *effectiveness of the IA* as understood by users. It is important to remember that not all feedback is equal or appropriate. Developers (and business stakeholders) are not the users and their feedback about what makes sense from the user's perspective is not the same as research with actual users.