Talking about tech is so hard, its a wonder we can do it at all
A while ago I was watching AWS re:Invent, presenting its ever-flowing stream of new and innovative cloud services. And I couldn't help but admire what I saw: Not just the services themselves, but especially the ease with which the presenters navigated the sky-high piles of ever thicker abstraction layers that constitute modern information technologies.
Of course, this talent isn't exclusive to the people we see on stage - we all use it on a daily basis, often without noticing. Just imagining Amazon announcing its new time-series-database-as-a-service offering with all abstractions removed:
"And today, we are proud to announce that if you choose to connect your intricate silicon, copper and aluminum mesh to our optic fiber cable we will send and receive a frequency of blinking lights that will be advantageous to your business goals"
Now, of course, that's a bit of a silly example - a lot of the things we use every day are inherently complex, yet we don't talk about them in terms of the internals that power them. In fact, most end-consumer companies avoid talking about the tech layer altogether - Apple's Facetime commercials show you emotionally charged imagery of a soldier speaking with his daughter back home without ever touching on how it works - but this is a luxury that B2B cloud providers like AWS rarely can afford.
Metaphors from here on out
Instead, we talk about IT almost exclusive in layered metaphors: Data is stored in buckets, streams are pumped through pipelines and glue holds it all together. This isn't a bad thing per se - but it comes with its own set of challenges:
Metaphors inform our thinking
Metaphors aren't a one-way street. Once we start thinking about A in terms of B, we start projecting the properties of B onto A. A classic example of this trap is the way we think about electricity in terms of water flows. You might remember early physics lessons where Volt was the speed of the water, Ampere the diameter of the pipe and Ohm the rocks you've crammed into it. This holds up well until your wire burns through because resistance generates heat - the point where the metaphor stops working.
Metaphors are driven by motivations beyond technology
Finding suitable metaphors for abstract cloud concepts is hard enough in itself. But commercial companies have motivations beyond explaining the pure tech. Business value properties such as reliability, efficiency or low maintenance all need to feed into the choice. It is thus a small wonder that AWS went with something as flimsy as a "bucket" for its storage solution, though other offerings like Kinesis (strong motion), Glacier (massive amounts of slow-moving ice), or Quantum ledger (futuristic, fast and somehow big as we seem to perceive quantum leaps as massive steps forward) seem to rectify this.
The toughest challenge, however, is to find metaphors that make technology appealing on an emotional level. Go's Gopher (a fast, agile, friendly animal), Github's Octocat (Many arms symbolize the network character) or Slonik, the PostgreSQL Elephant (famous for their memory) are all great examples of this.
Translating these concepts into symbols or - better yet - tangible objects can be a great way to communicate their interplay, but comes with its own set of challenges:
Some can be derived from their hardware equivalents: Databases from disk-based hard-drives, save icons from floppy disks and so on, but as technology moves to the cloud it becomes increasingly abstract and the underlying hardware concepts are gradually forgotten.
Visual metaphors, however, come with the benefit of displaying the "what", rather than the "how". ETL jobs can translate to funnels, Docker to shipping containers or computer vision to cameras.
Tailoring communication and messaging to the receiver is crucial. Strong technical audiences might be familiar with a wider range of established metaphors, business people might need more concrete examples.
But ultimately it is the combination of these concepts and the creation of a strong context that makes for the most effective approach to communicating abstract ideas. Context makes iconography and metaphors add up - you might not identify a funnel as a means to process data on its own, but once a cloud pushes data-points into it and they come out to a bucket on the other side, things are a lot clearer.