When information architectures collide — using IA to optimise product integration

danramsden
6 min readSep 6, 2024

--

No polar bears were harmed in the generation of this image

Summary: By helping you to consider trade-offs to internal coherence to boost compatibility and external coherence, the idea of ‘concurrent information architecture’ can help you to optimise the points of overlap, interaction and integration between products and services.

What is an information architecture?

Information architecture is the sense-making signals and relationships which exists within an environment — it’s how we categorise, describe and arrange the information in a design. Good IA enables us make sense and make use of products and services.

In 2020 I wrote about two types of information architecture:

  • Structural IA is determined when we architect, design and make — we embed structure and signals into a “place” which will either help or hinder a user to make sense of it.
  • Individual IA are the rules, “grammar” and personal experiences the user brings which can shape how they interpret the built-in environmental signals of Structural IA.

What was concurrent information architecture?

Concurrent information architecture is the idea of different information architectures occupying the same environment, product or service and overlapping. Structural IA and Individual IA are examples of concurrent information architectures. The two coincide and overlap within every experience. But most experiences will also have multiple sets of Structural IA overlapping.

My example of concurrent information architecture in 2017 was Sonos networked speakers — I wrote about how multiple information architectures overlapped — (1) the “speakers” as locations and (2) the music/audio you were playing. You could change either the speaker or the music. Each speaker represented a different place that contained that ‘IA of music discovery and playback.’ Users can find audio and play it in a place. They can use multiple services to find music and audio. And the places also have a fluid structure. You can group speakers to create new places. All this is contained in the Sonos app ecosytem.

In 2017 I was concerned with ways of connecting products into networked services to increase the value BBC users got from our various products. I argued that without shared architecture they remain ‘un-networked’. They could be linked to and between, just like any other parts of a digital ecosystem. But they wouldn’t make as much sense together as they could. I thought that shared or concurrent architecture could create better integration within and between networked or connected system. It supports sense-making when people changed direction or jumped to another structure within the network.

When concurrent information architectures are compatible, the user can make sense of things in richer ways. But if they’re not compatible, because the IA is poorly conceived or multiple architectures interact in unintended or unanticipated ways, then this “unintentional IA” will damage the experience.

What are intentional and unintentional information architecture?

‘Unintentional information architecture’ describes situations where the information architecture in a design has not been consciously considered. The main point of unintentional IA is that information architecture always exists within an enviroment and experience — structural elements and individual sense-making combine as a user constructs a mental model and tries to interact with an environment. If you haven’t intentionally constructed the IA, the unintentional IA in the enviroment might result in a worse experience than you wanted.

I used the example of a wonky set of buttons to control a lift.

Along with other examples of unintentional IA I used the example of these lift controls — where the order of the numbers does not make sense.

I argued that the numbered buttons are a good design if you consider the numbers individually — there’s tactile and visual feedback, accessibility features and it makes use of a well-established convention — the button. But the arrangement, the information architecture of the buttons as a set, breaks the design.

Information architecture has two responsibilities — internal coherence and external coherence — making sense in context. The internal coherence of the number-to-button relationship is good— each button has a single number which represents a floor in a building. It’s a mutually exclusive and (presumably) completely exhaustive set which gives you access to anywhere you’d like to go. But the design fails in the number-to-number-to-building relationships, where we expect the floors in a building to be arranged in ascending numerical order. The two sets of information architecture are incompatible — the structure of the building (ie. floors following ascending numerical order) and the order/arrangement of the buttons don’t coincide in the way we’d expect.

Unintentional information architecture is easy to create when we don’t pay sufficient attention to both the internal and external relationships within and between information architectures. Therefore, concurrent information architectures are a regular source (maybe the main source) of unintentional information architecture — when we haven’t identified an overlap in sense-making mechanisms, unintentional IA can occur.

And while the definition of intentional and unintentional architectures depend on the fact that information architecture always exists — I think I now must extend my definitions to say, concurrent information architecture always exists… your IA always interacts with an interpretative contribution the user makes. But it likely also always interacts with other information architecture — it might overlap, interact or integrate.

Concurrent information architecture and unintentional architecture

Unintentional IA can result in a lack of coherence — a sort of jarring discontinuity when the elements of the IA don’t make sense when experienced together. This might occur when structural IA and Individual IA are incompatible. But it might also happen as you integrate environments or when they interact.

My definition of IA states that it’s about “internal coherence” — providing signals and structure to help make sense of a product/service/environment. Any definition of “internal coherence” relies on a clear sense of the boundary… and I suspect thinking about ‘concurrent information architecture’ is most useful because it prompts consideration (and hopefully debate) about that boundary.

Architecture or architectures, MECE and multiplicity

It’s a good rule of thumb to think about the thing you’re making “in the next biggest context.” By considering the environments which might contain our designed experiences we can insulate them and integrate them as needed. Integration brings efficiency. Interactions between systems can extend their utility. Insulation might help protect and increase resilience. So thinking about the way your design interacts at the level of information architecture (as well as the other points of interaction and integration) can bring two of the key benefits of IA — efficiency and resilience. I suspect there are likely always ways to adapt how your solution delivers internal coherence which can boost compatibility, integration and external coherence. It might involve trade-offs – but that is the essence of good design and intentional information architecture.

In the physical world a book can only be in one place at a time, so the Dewey Decimal system provides a set of mutually exclusive and completely exhaustive categories (MECE) that enable predictable categorisation and location. This act of organisation has one point of integration — the book and the shelf. As content becomes more fragmented, aggregated and disaggregated, describing it becomes more complex… add to this the fact that the same “thing” can be located in multiple places simultaneously and there is more chance of information architectures overlapping. But this unlocks new sources of value and utility — it just complicates the job for information architects.

So, what does this mean for the way you make information architecture?

  • continue to think about mutually exclusive and completely exhaustive categories to provide learnability and predictability to the structure and categorisation in your design — this is a reliable way of providing good internal coherence which boosts predictability and learnability
  • consider the architectures that your architecture could interact, integrate and coincide with — consider these points of connection
  • consider trade-offs which increase efficiency and resilience at points of interaction and integration with other system
  • make choices about your priorities and be prepared to forfeit some internal optimisation to make sure your information architecture survives contact with the real world.

To deliver value, your design must work in the real world. It’s easier to imagine the hermetically sealed perfection of a system with high internal coherence. But this is likely a lie, and when your design begins to interact with real users and the real world, it will likely break because of the unintentional architecture it contains and which contains it. Extending your attention to points of integration and interaction will always result in better designs.

--

--

danramsden

I'm a Creative director at the BBC. I like words, design, data and magic. These are all my own views (apart from retweets. I borrowed those to look clever.)