Navigating the Complexities of Information Architecture vs Data Architecture
Thanks to an ever-growing digital world, businesses are spoiled for information and data. The more complex the business, the more information there is and the more complicated the business requirements are. But where there are challenges, there is opportunity. That’s where information architecture and data architecture come in.
Information and data architecture both seek to make sense of the plethora of information a business handles. However, the two have different roles to play in the way businesses use, move, maintain, and present data - both to internal and external stakeholders. So what are they and why should businesses take note?
Defining Information Architecture 🗺️
Information architecture is the structure used to organize and label content on websites, mobile applications and other digital environments. Its primary purpose is to enhance user experience by ensuring information is structured in an accessible, usable and relevant way.
Information architecture seeks to understand user needs and goals by analyzing both existing and required information, then building an information framework in a logical and user-friendly way. It deals with three main components:
Labels: How information is represented
Navigation: How users make their way through the information
Search: How users look for information
Whilst this information sits in the background, it’s the layer upon which you build the design of your digital products.
Information architects bring data from file systems and databases to life by building meaningful narratives and stories. Outputs can include site-mapping, information architecture diagrams and content inventories. These outputs are supported by user research techniques such as card sorting, tree testing, user surveys and first-click testing.
Defining Data Architecture 💻
Data architecture bridges the gap between business needs, goals, and system requirements related to data handling. It sets out a framework for managing data assets, the flow of data and the maintenance of data systems. As such, it has a slightly more macro view than information architecture and concerns itself with emerging technologies such as artificial intelligence, machine learning, and blockchain.
Where information architecture centers around the end-user interaction, data architecture centers around practical handling and operation of data processes i.e. collection through to transformation, distribution, and consumption. Because of this, data architecture must take into account the businesses ability to scale operations, integrate with third party systems, support real-time data processes and the reduction of operating costs. Modern data architecture may point to artificial intelligence to tackle some of these challenges.
The Importance of Enterprise Architects in Information and Data Architecture 🏗
Enterprise architects are big-picture people. Data architecture and information architecture both fall within their remit, and they often oversee other data management job specialities within an IT department.
As a leader (and often, visionary) within a business, enterprise architects shoulder the responsibility of ‘mission critical’ projects. As a result, they tend to have several years experience with IT systems, backed by a bachelor’s degree in computer science, IT management, data science or similar. Many will hold a master’s degree and specialty certifications.
The role involves collaborating with senior business leaders, solution-delivery teams and external stakeholders, and requires creative problem solving and excellent communication skills. Therefore, enterprise architects very much steer the ship when it comes to information and data architecture. Combining high-level business strategy with knowledge of ‘the nuts and bolts’ of IT data systems and processes, they command an annual salary in New Zealand between $150,000 and $200,000 per annum..
Continuous improvement within any business that has substantial IT infrastructure calls for serious investment in enterprise architecture.
Designing and Implementing an Effective Information and Data Architecture 𝞹📈🧠📚
Once overarching business goals are aligned with the scope of data and system requirements, information and data architecture design (or redesign) can begin.
Crucial to the design and implementation process is developing an architecture framework. This is a set of guidelines that lays out principles, practices, tools and approaches required to complete the design. It supports system design decisions, assigns key tasks and provides project guidance throughout the design process. The framework essentially aims to unite disparate teams and maintain business and IT alignment.
The choice of architecture design is also critical. It should consider scalability, performance, maintainability and adaptability to emerging technology. Which is why cloud platforms feature so heavily in modern data architecture. Cloud architects will navigate the architecture design and technical requirements of cloud-based delivery models, which offer the solution to those scalability and adaptability challenges. They are responsible for bridging the gaps between complex business problems and solutions in the cloud. Modern data architecture tends to involve some form of cloud delivery component.
Throughout implementation, data and information architects will work closely with designers and engineers until testable architecture is ready. User research and testing will be carried out, and a feedback loop will commence until requirements are met. Users, as always, should be at the center of your digital product.
Summing Up the Complexities of Information and Data Architecture 🧮
Whilst the difference between information and data architecture can appear nuanced on the surface, they hold unique roles when delivering a cohesive, user-friendly digital product.
Think of a sliding scale where business operations sit at one end, and users sit at the other. Data architecture addresses challenges closer to the business: aligning business requirements and goals with how data flows through the system. On the other hand, information architecture addresses the challenges related to how this data is organized and interpreted for the end user.
At the end of the day, both information and data architecture need to work in harmony to satisfy the user and the business.
Sitting inside any beautifully crafted and designed digital product, there must be a fully functional and considered information architecture.
As much as information architecture shouldn’t be developed in a vacuum. Neither should the design and look of digital products. In fact, a large proportion of the function of digital designers is devoted to supporting users locating content they need and driving them towards content that the product owners want them to find.
Incorporating visual markers to make sure that certain content is distinct from the rest or creating layers that demonstrate the diverse content on a product.
If you do not have quality content, it is impossible to design a quality digital product. It all comes back to creating a user experience that makes sense and is designed to make task completion simple. And this relates back to designing the product with the content planned for it in mind.
8 Principles of information architecture, according to Dan Brown 🏗️
As a designer, the more you know about information architecture, the better the products you design will meet your user requirements and deliver what they need. If you work with an information architect, even better. If you’re still learning about information architecture the 8 Principles according to Dan Brown is a great place to begin.
If you haven’t come across Dan Brown yet, you have more than likely come across his 8 principles. Dan Brown is one of the UX world's most prolific experts with a career that spans most areas of UX designs. He’s written 3 books on the subject and experience across a multitude of high profile projects. Aiding large organizations to make the most of their user experience.
The principle of objects: Content should be treated as a living, breathing thing. It has lifecycles, behaviors, and attributes.
The principle of choices: Less is more. Keep the number of choices to a minimum.
The principle of disclosure: Show a preview of information that will help users understand what kind of information is hidden if they dig deeper.
The principle of examples: Show examples of content when describing the content of the categories.
The principle of front doors: Assume that at least 50% of users will use a different entry point than the home page.
The principle of multiple classifications: Offer users several different classification schemes to browse the site’s content.
The principle of focused navigation: Keep navigation simple and never mix different things.
The principle of growth: Assume that the content on the website will grow. Make sure the website is scalable.
It’s highly likely that you’ve already used some, or all, of these IA principles in your designs. Don’t be shy about mastering them, or at the very least be familiar. They can only help you become a better user experience designer.
Wrap up 🌯
Mastering the 8 principles, according to IA expert Dan Brown will see you mastering the complex tasks of information architecture. Understanding IA is key to creating digital designs with a content structure that is functional, logical and just what your users need to navigate your product. Design without good IA doesn’t work as well, just as a content structure without a well designed interface will not engage users.
As information professionals, we work with the “stuff” of information in our everyday work. We search for information, we spend time analyzing and synthesizing it, and we carefully create and structure it. Whether you elicit information from users and stakeholders, explore large data sets, design ‘journeys’ or interfaces, or create information architectures, understanding the information you are using and creating as information can help you do your work better.
Kat King, Business Intelligence Analyst at the University of Michigan Library, recently spoke at UX New Zealand, the leading UX and IA conference in New Zealand hosted by Optimal Workshop, about understanding exactly what information is, and where it is, in our work.
In her talk, Kat uses simple examples to teach you to “see” the information around you and understand what makes something “information” in the context of working as a human to accomplish something.
Kat King bio 🎤
Kat King is an Information Architect interested in language, meaning, and the things we make. She currently works as a Business Intelligence Analyst for the University of Michigan Library.
Information theory can be dense and jargon-filled, and discussions in academic texts can feel divorced from the practice of actually working with information. We’re all told that information architecture is much more than website navigation. So, what is it? IA has a reputation for being difficult to understand, and in her talk, Kat attempts to help us understand what it is, where the information is, and what is it that we’re doing when we use IA methods.
Kat defines IA as “the practice of ensuring ontological alignment”. ‘Ontological’ relates to concepts, categories, properties, and relationships. ‘Alignment’ means arrangements into appropriate relative positions. Therefore, information architecture is “the practice of ensuring concepts, categories and their properties and relationships are arranged into appropriate relative positions.”
To align information then, you need to begin by sorting it into concepts and categories, which is difficult because information can sometimes be “slippery and abstract”. Kat argues this is the real reason that IA is sometimes hard to wrap our heads around. So, getting to the heart of the question, what is information?
Kat defines information as “a patterned relationship between differences that reduces uncertainty”. The key word here is ‘differences’. The trick to understanding and taming information is to identify what is different about sets of information. The next trick is to identify consistencies between these differences.
This can be a little confusing, so Kat uses the example of picking fruit. We tend to use color (the difference) to identify when fruit is ripe and sweet. We know for a fact that, at some point, the fruit will be at its sweetest and, while there is a scientific way of identifying this point, we have to use the information we have at our disposal instead i.e. the colour of the fruit. The skin of the fruit in this example is like an interface - allowing a flow of information from the fruit’s ripening process to our eyes.
Information categories 🧺🧺🧺
The relationship between the information described in the fruit example can be split into two categories. “Information 1” is a factual, objective description of when the fruit is ripe (i.e. the science of why the fruit is the color that it is right now), whereas our subjective observation, based on color, is “Information 2”.
Information 1: Matter and energy, and their properties and interactions i.e. the laws of physics and universal truth or rules
Information 1 poses challenges for us because we have a narrow range of perception, attention, and aggregation, which means we, as humans, can’t possibly understand the laws of nature just by observing. We have evolved to be simple, efficient observers of what is important to us. In other words, we don’t need to understand everything in order to get things right. We see patterns and generalize. Going back to the fruit example – we only need to know the color of ripe fruit, not the exact chemistry of why it is ripe.
Information 2: This is Information 1 that is given meaning by humans. This is done via processing semantic information, or “differences and structures that create meaning for people”.
We use semantic information by processing concepts, patterns, categories, mental models, and even language as inputs to form our understanding. As social animals, we tend to reinforce general ‘truths’ about things because we’re constantly cooperating using shared information. General ‘truths’ are good enough.
Kat uses the following interaction to demonstrate the interplay of different information.
Person 1: If the raspberries look good, can you get some for me?
Person 2: How can tell is they’re good?
Person 1: Get the ones that are the most red.
In this interaction, the different pieces of information can be broken down by category:
Semantic information = Words and concepts
Information 1 = Meaningful signs
Information 2 = Perceptible differences
Real life information = Raspberries
Using our ability to communicate and understand concepts (words “red”, “good”, and “raspberries”) helps us to understand Information 2 (processing the words and concepts to understand that a red berry is good”), which aligns with Information 1 (the evolutionary science and ongoing consistency of red/ripe berries being sweet) that helps us decide when processing all of this information.
So, now that we understand a little more about information, how does this influence our roles as designers?
Why it matters 👀
Thanks to our individual lived experiences, people have many different inputs/concepts about things. However, Kat points out that we’re pretty good at navigating these different concepts/inputs.
Take conversations, for example. Conversations are our way of getting a “live” alignment of information. If we’re not on the same page we can ask each other questions to ensure we’re communicating semantic information accurately.
When we start to think about technology and digital products, the interfaces that we design and code become the information that is being transmitted, rather than words in a conversation. The design and presentation become semantic information structures, helping someone to understand the information we’re putting forward. This highlights the importance of aligning the interface (structure and semantic information) and the users' ontology (concepts and categories). For the interface to work, IA practitioners and designers need to know what most people understand to be true when they interact with information, concepts, and categories.
We need to find some sort of stability that means that most users can understand what they need to do to achieve a goal or make a decision. To do this, we need to find common ground between the semantic information (that might vary between users) so that users can have successful Information 2 style interactions (i.e. absorbing and understanding the concepts presented by the interface).
To wrap up, let’s remind ourselves that information architecture is “the practice of ensuring concepts, categories and their properties and relations are arranged into appropriate and relevant positions”. As IA practitioners and designers, it’s our job to ensure that concepts and categories are arranged in structures that can be understood by the nuance of shared human understanding and semantic information – not just in some physical diagram.
We need to present stable, local structures that help to reduce uncertainty at the moment of interaction. If we don’t, the information flow breaks and we aren’t reducing uncertainty; instead, we create confusion and disappointing user interactions with our digital products. Making sure we present information correctly is important (and difficult!) for the success of our products – and for better or worse, it’s the work of information architecture!
The Ministry for Primary Industry’s (MPI) customers have some of the most varied information needs — possibly the most varied in New Zealand. MPI provides information on how to follow fishing rules, what the requirements are to sell dairy products at the market, and how to go about exporting honey to Asia. Their website mpi.govt.nz has all the information.
However the previous website was dense and complicated, and MPI’s customers were struggling to find the information they needed, often calling the contact center instead — one of several indicators that people were lost and confused on the website.
Ruth Hendry, Head of Strategic Growth at Springload, recently spoke at UX New Zealand, the leading UX and IA conference in New Zealand hosted by Optimal Workshop, about how new IA helped MPI’s broad range of customers find the information they needed.
In her talk, Ruth takes us through the tips and techniques used to create an IA that met a wide variety of user needs. She covers the challenges they faced, what went well, what didn’t go so well, and what her team would do differently next time.
Background on Ruth Hendry 💃🏻
Ruth was Springload’s Content Director; now she’s Head of Strategic Growth. She has broad experience in content, UX, and customer-led design. A data nerd at heart, she uses analytics, research and testing to drive decision-making, resulting in digital experiences that put the customer at the forefront.
At Springload Ruth has worked on large-scale content and information architecture projects for organisations including Massey University, Vodafone and Air New Zealand. She got into the world of websites in her native UK, working on Wildscreen's ARKive project. After she arrived in Aotearoa, she spent four years looking after Te Papa's digital content, including the live broadcast of the colossal squid dissection. She's Springload's resident cephalopod expert.
She finds joy in a beautiful information architecture, but her desk is as messy as her websites are tidy.
Food recalls, fishing rules, and forestry: creating an IA strategy for diverse audience needs 🎣
Ruth begins her talk by defining IA. She says, “If IA is the way information is organized, structured, and labeled, then an IA strategy is the plan for how you achieve an effective, sustainable, people-focused IA.”
Considering this, applying an IA strategy to the Ministry of Primary Industries (MPI) website was a challenge due to its diverse user groups. MPI is responsible for a range of things, such as publishing food recalls, looking after New Zealand’s biosecurity, outlining how much fish can be caught, how to export products, and even how to move pets between countries. Needless to say, the scope of this IA project was huge.
The current state of the website was challenging to navigate. In fact, one customer said, “It’s hard to find what you need and hard to understand”. MPI Contact Center staff often found themselves simply guiding customers to the right information online over the phone.
So, in solving such a massive problem, does having an IA strategy work? Ruth says yes! And it can have a huge impact. She backs up her strategy with the results of this project before broadly outlining how she and her team achieved the following improvements.
The project achieved:
37% decrease in time spent on the home page and landing pages
Customers found where they needed to go, faster, using the new IA and navigation elements
21% decrease in on-page searches
People could find the content they need more easily
53% reduction in callers to MPI saying that they couldn’t find what they needed on the website
Users could more easily get information online
Developing an IA strategy 🗺️
Ruth attempts to summarize 14 weeks' worth of work that she and her team delivered in this project.
Step one: Understanding the business
During this step, Ruth and her team looked at finding out exactly what MPI wanted to achieve, what its current state is, what its digital maturity is, what its current IA was like (and the governance of it), how the site got to be in the way that it was, and what their hopes and aspirations were for their digital channels. They conducted:
Stakeholder interviews and focus groups
Reviewing many, many documents
Domain and analogous search
Website review
Step two: Understand the customers
In this step, the team looked at what people want to achieve on the site, their mental models (how they group and label information), their main challenges, and whether or not they understood what MPI does. They conducted:
A review on website analytics and user needs
In-person interviews and prototype testing
Card sorts
Intercepts
Users surveys
Treejack testing
Step three: Create the strategy
This talk doesn’t cover strategy development in depth, but Ruth shares some of the most interesting things she learned (outlined below) throughout this project that she’ll take into other IA strategy projects.
Why it matters 🔥
Throughout the project, Ruth felt that there were eight fundamental things that she would advise other teams to do when creating an IA strategy for large organizations with massively diverse customer needs.
Understand the business first: Their current IA is a window into their soul. It tells us what they value, what’s important to them, and also the stories that they want to tell their customers. By understanding the business, Ruth and her team were able to pinpoint what it was about the current IA that wasn’t working.
Create a customer matrix: Find the sweet spot of efficient and in-depth research. When an organization has a vast array of users and audience needs, it can often seem overwhelming. A customer matrix really helps to nail down who needs what information.
Card sort, then card sort again: They are the best way to understand how people’s mental model works. They are critical to understanding how information should be organized and labeled. They are particularly useful when dealing with large and diverse audiences! In the case of the MPI project, card sorts revealed a clear difference between business needs and personal needs, helping to inform the IA.
Involve designers: The earlier the better! User Interface (UI) decisions hugely influence the successful implementation of new IA and the overall user journey. Cross-discipline collaboration is the key to success!
Understand the tech: Your IA choice impacts design and tech decisions (and vice versa). IA and tech choices are becoming increasingly interrelated. Ruth stresses the importance of understanding the tech platforms involved before making IA recommendations and working with developers to ensure your recommendations are feasible.
Stakeholders can be your biggest and best advocates: Build trust with stakeholders early. They really see IA as a reflection of their organization and they care a lot about how it is presented.
IA change drives business change: You can change the story a business tells about itself.Projects like this, which are user-centric and champion audience thinking, can have a positive effect throughout the business, not just the customer. Sometimes internal business stakeholders' thinking needs to change before the final product can change.
IA is more than a menu: And your IA strategy should reflect that.IA captures design choices, content strategy, how technical systems can display content, etc.
Your IA strategy needs to consider
Content strategy: How is content produced, governed, and maintained sustainably going forward?
Content design: How is content designed and does it support a customer-focused IA?
UI and visual design: Does UI and visual design support a customer-focused IA?
Technical and functional requirements: Are they technically feasible in the CMS? And what do we need to support the changes, now and into the future?
Business process change: How will business processes adapt to maintain IA changes sustainably in the long term?
Change management and comms plan: How can we support the dissemination of key changes throughout the business, to key stakeholders, and to customers?
Finally, Ruth reemphasizes that AI is more than just designing a new menu! There’s a lot more to consider when delivering a successful IA strategy that meets the needs of the customer - approach the project in a way that reflects this.
