What makes something feel digitally lightweight?
If you do a quick search, you'll find the heaviest object in the world is considered to be the Great Wall of China, the next heaviest being the Three Gorges Dam. While the heaviest object ever weighed is the Revolving Service Structure of launch pad 39B. It turns out many of the "heaviest" objects are actually calculated, not weighed.
Just like their enormous, physical counterparts, you can calculate the weight of a digital object. This is useful for making sense of digital things: products, services, protocols. It can also help you design digital objects.
Digital weight
What makes something feel digitally lightweight could be figured out by just qualitatively guessing.
A text message might feel lighter than a blog post. It fits in a smaller space, there's less "weight" from formatting like headings, emphasis, quotes. A text message is easier to move. I can move the same text across devices without losing information. Moving data without losing information was essentially the birth of information theory.
The weight could also be a function of how easily it was replicated across devices and platforms. A text, for example, can easily be replicated across multiple carriers, hardware providers, and operating systems. Any combination of these three will result in the same message being communicated.
whether or not a message was sent
hardware | tmobile | at&t | verizon |
Apple | yes yes yes
Samsung | yes yes yes
Google | yes yes yes
As it turns out, a text message is lightweight enough to survive multiple permutations.
What defines digital weight?
Qualitatively, a text message feels lighter than a blog post. But how could we quantify digital weight?
When I first started thinking about this, I used the format count to define how heavy a digital object was. As an example, for a tweet we might have: text, hashtags, images, and mentions. We could set a tweet's weight at 4. For a text we could set it at 1—just text or just image—disregarding all the crazy unfurls iOS might do. Let's look at a few more:
object | weight
text message. | 1
tweet. | 3
markdown. | 9
instagram. | 5
blog post. | 12+
This felt like the definition for weight was lacking. If we look at text messages compared to blog posts, we can see a couple more differences:
The number of objects (text, images, maybe a poll, etc.)
Number of formats (headings, emphasis, etc.)
How hard it is to move or replicate
The number of objects is important, because you could have few or many; but extra points for being optional. The formats—bolds, italics, lists—the ways you can manipulate weight. Finally, can you move it? How easy it is to move or replicate is a factor.
The (working) equation could be something like:
(objects + formats) / replicability = weight
objects : the digital thing (images, text, whatever) + multiply by .5 if it’s optional
formats : how you can format an object (height, width, headings, bold, etc.)
replicability : score of 1-5, how easily replicated or duplicated is this?
Let's look at a couple of examples:
Text Message
Objects: 1 (1 image or 1 text, not optional)
Formats: 1 (can't bold or anything)
Replicability: 5 (high score because it can go anywhere, no formats)
Slack Message (currently)
Objects: 5 (text, mention, channel mention, image or file, emojis?, all but one optional)
Formats: 9
Replicability: 3 (easily replicated on Slack, replicated with some data-loss other places)
Texts have a weight of 0.4
here. Slack messages have a weight of 3.83
. Our algorithm seems close.
The specifics of digital weight
I think it makes sense to go back to the qualitative side and explore some of these ideas, specifically: optionality, replicability, and format count.
We can start with the easy one: format count, which was addressed above. The more formats a digital object has, the heavier it is. This doesn't mean the format count is bad, it's just relative. Docker reduced the weight of code deployment and jobs by creating a system where any code could run anywhere. Independent from platform (AWS, Google, Azure). It's technical, and format count is high, but the replicability of these objects is also high.
Which brings us to our next point: replicability. Thank goodness this is a wordy post and not an out-loud talk, or I'd choose a different word. Put simply, this is how easily replicated something can be. A text message can be replicated pretty easily as it's just an image or text, and the software/hardware/carrier handles the rest. Have you ever tried to migrate an entire blog, or database before? Less easily replicated = more weight.
Optionality is one of my favorites, but one we haven't talked about yet, it contributes to the object score. The basic idea is: if something is optional, it feels much lighter (and faster) than if something is required. Notion—my productivity tool of choice—gives you optional custom properties on objects, allowing you to do all kinds of crazy stuff. A layman's example of this is "tags". You can "tag" a lot of digital objects, and you can use many, but it's not required. Like hashtags on Twitter.
What are the benefits of digital weight?
If you’re designing or providing a digital object spanning across platforms—devices, websites, differences in bandwidth—it could benefit to be lightweight. Lightweight could mean: less maintenance and development time, less confusion, better communication.
Though you don't always want something to feel light. Gold per pound sells for a pretty penny. The heavyweight champion could knockout the lightweight champ; sometimes weight is good.
You might be able to use digital weight in other weird ways we could explore later:
predicting the ubiquity of a digital object or platform
predicting profitability of a digital product
designing using weight (not just hindsight) depending on your needs
There's more to this. But in the interest of finishing this post, and going to sleep, I'm just going to let it into the world. Here's a closing question to ponder: What could be made digitally lightweight that hasn’t been yet?
Cheers, and Godspeed.