Categories
Information Architecture Technology

You can’t rely on search alone

People searching through bins of vinyl records. Photo by Anthony Martino on Unsplash

Why is it still so hard to develop great information experiences?

Featured in UX Collective

Think about the last time you ate takeout. We all know the dance of indecision where we have the perfect thing in mind, but have struggled to give it a name. In those moments we settle on one thing only and sometimes realise that when we have it, it was something else entirely we really wanted.

This is the very same process people go through when attempting to locate information in complex information systems. They know the kind of information they want, but cannot for the life of them put that into words, let alone think of a search term.


The responsibility of information design

When developing any kind of system it’s imperative to place people at the very heart of its design. You can build it out in the most structural and logical manner, but if the people who use your system can’t understand how it fits together then then its usability diminishes greatly.

All information systems have a single goal, connecting people to knowledge, if they fail at providing this simple-yet-fundamental truth then it becomes only an exercise in frustration for all involved.

Be it a single document, a website, or an entire knowledge base, they all need to be carefully curated and crafted to ensure that its intended audience is well catered for.


A common path

We know that most people seek out the humble search box when looking for information in digital systems. It’s how we’ve all been conditioned since the advent of global search engines and our reliance on their available features to find information on the internet.

The role of a search engine is to take several related keywords and turn those into a query. Through tokenisation of the query, and quickly parsing them against a database containing known terms, the system attempts to surface information based on the relevance of the keywords in found in available sources.

Because this method of finding information feels more efficient to people who are conditioned to using search, when approaching a new information resource their first reflexive action is to look for the search box and type a query.

So how do we ensure people can find the things they are looking for?


Information is atomic by nature

All knowledge starts with a single word.

Words make sentences, sentences paragraphs and so on until we end up with what we call information. It’s this information when found, consumed, and finally understood by a person that forms a basis of knowledge in that individual.

If this is true, then why do so many information systems tend to surface only broad strokes like titles, tags, and keywords? The structure of sentences, paragraphs, phrases, etc are such an important part of not only identifying information as useful, but also in its comprehension.

With an understanding that knowledge begins at a single word as the smallest element, the atom, we can use this learning to design systems which better connects people with the information they seek.


Role of structure

While searching alone can surface good information it’s crucial you don’t ignore the importance of excellent relational structure. All information has structure (it is atomic after all), so why is it that when developing modern information systems that so many choose to ignore structure in favour for a much flatter architecture?

Information structure ~ knowledge.

Structure is the compartmentalised link which binds collections of information together. It’s the breadcrumb needed to not only access information, but to also form an understanding from its neighbouring relationships.

Let’s look at a simple example.

Say you’ve written a piece that discusses a particular topic. Let’s imagine it’s about a type of cheese.

Immediately you can begin to imagine its information structure.

We already know that the article is about ‘cheese’, but here we’re proffered the information that it’s also about a taxonomically different type, in this instance perhaps ‘blue-veined’.

Here is our initial structure.

The topic Cheese, with a sub-topic of Blue-Veined Cheese

We know just from the structure above and its placement in that structure that it is about a type of cheese called ‘blue-veined’.

The topic of Cheese, subtopic of Blue-Veined Cheese, and the property of Types of Blue-Veined Cheese

Inferring from the base topic, Blue-Veined Cheese, then we can look at the qualities of the item referring to Types of Blue-Veined Cheese.

We can infer from this structure that if there is one type of cheese called ‘Blue-Veined’ there are other Types of Blue-Veined Cheese in the world, so here we can further add properties to the structure.

The further breakdown of the Blue-Veined Cheese, the types of Blue-Veined Cheese, and the properties of the topic.

To extend this further, we can also infer that if Blue-Veined Cheese is defined that other types of cheese must also exist, otherwise the topic would simply be ‘Cheese’.

An extended breakdown of the topic Cheese and its subtopics of Blue-Veined Cheese and Soft Cheese.

Without you needing to read the content written on each part of the structure you can already infer:

  1. Cheese is a common term for a group of items
  2. There are sub-topics of cheese
  3. Blue-Veined Cheese is one of those sub-topics
  4. There is more than one sub-topic under cheese
  5. Each sub-type has its own properties of firmness, moisture, taste, and texture
  6. The sub-type is classified by its properties

All gained from reading the structure.

This is just a simple example of how good information structures allow for a better experience when attempting to understand a topic.

If we were to go further then we could classify cheese as a type of dairy product, which in turn is a collection of food products. Again, without any further content we already know that food, is a complex topic of which dairy, and cheese, comprises a small part.


Build for people with the best of both worlds

It isn’t enough to put all of your effort and energy into developing only search as the singular method for a person to find information in systems you develop.

People don’t know what they are looking for, you have to show them.

While keyword sourcing, tagging, and tight metadata control all provide great ways to surface information from search queries, using a carefully considered and planned information structure enables people to more freely explore the links between related topics.

The truth of the matter is that when initially researching a topic or attempting to gain clarity or understanding, people don’t know what they are looking for, you have to show them.

It doesn’t matter how a person starts to explore the information, be it through structured navigation of topics and sub-topics or simply beginning a search, it’s critical that they have both the information and an understanding of its structure.

Not only is it best practice, people who use your systems will thank you for it.

By Tim King

Tim is a creative thinker, self-proclaimed futurist, and writer based in Central Victoria, Australia. He's been designing and developing digital content online for over a decade and loves digging into the big topics that shape our world, applying his own brand of thought along the way.

You'll usually find thinking far too hard about the future, often with a good glass of red in hand.