The Technology Adoption Curve
--
When I sent it to a friend, he responded: “Wow, who hurt you?”
Everyone.
Uber hurt me when I tried to rent JUMP bikes in LA, and every single one broke. Airbnb hurts me when the price triples at checkout. Even Apple hurt me when the Catalina upgrade deleted two years of notes after I clicked “sync with iCloud.”
This illustration doesn’t apply to all technology companies. Google, as an example, doesn’t give me nervous tremors. I attribute the “Get Fucked” phenomenon partially to the fact that Google’s core product is exclusively in the virtual world. Uber and Airbnb, like most sharing-economy technology apps, operate partly in the meatspace and therefore have to deal with the reality of our physical world.
Back in the early-minority phase, opening novel apps and interacting with technology was fun. Hours of talking to Lyft drivers and getting to know Airbnb hosts. Now, a host refuses to let me drop bags off early unless I pay a $40 fee.
The only solution I’ve found is getting out of these apps before the early-majority phase. Because once too many people join, you’re going to get fucked. Don’t worry; you can come back in a later stage; hopefully, things will be better.
As an employee of a hyper-growth growth company, I understand how this happens. As a customer, I don’t like it. As a product manager, I detest it. Although we have limited recourse as consumers, the product manager in me had to pull apart the reason why using inventive apps has started to feel like stepping into a nightmare.
The two main offenders I’ve found are code complexity and process capacity.
Code Complexity
As time passes and a company grows larger, the software running the product is likely to become infinitely complex. Think about the house your grandma had for three decades. It’s full of things that were collected and never thrown out. The same goes for codebases. Similar to your grandma’s house, cleaning up a codebase requires effort no one wants to put in. Unlike your grandma’s house, the ‘clean-up’ is not one you do with your body but one you do with your mind, which is a different type of exhausting.
As a small representation, this is the flow used in Slack to decide to send a notification. This logic is for the simple task of sending a notification, one small part of the total app.
This complex flow for a tiny part of the app has to get integrated with all the other, even more, complex flows to make up the entire slack app.
Once the complex slack app is complete, it integrates with other complex systems like Mac, Windows, Chrome, Safari, Firefox, App-store, Play store, etc.
Essentially, technology is a never-ending, ever-growing set of complex interactions. I’m astonished this shit works at all.
Process Capacity
Software, for the most part, can be scaled exponentially (add AWS instances!), but the meatspace can not: physical supply chains are slower, errors have consequences, and issues are less uniform. However, the root of the problem is that software companies ignore these realities.
Physical supply chains are slower
At Coinbase, we 10x’d our website capacity in a few weeks. Try to 10x the supply of JUMP bikes or Uber drivers in a few weeks. You can’t because physical supply chains don’t scale as quickly as virtual ones.
Errors have higher consequences
The speed of iteration in the meatspace is slower because repercussions are costly, further slowing down process capacity.
Bugs are less uniform
When something in the meatspace goes wrong, a bug, if you will, it will be less uniform and more challenging than bugs in the virtual world. Not only will solving these challenges take more capacity, but they also increase the quantity and complexity of support tickets.
Root Cause: ignoring reality
If you look outside the software industry, process capacity in the meatspace is a solved problem. Car manufacturers and airline companies (typically) don’t endlessly frustrate customers with their processes. At tech companies, it’s an attitude problem. When organizations dub themselves a “software company,” they ignore issues that don’t fit into the category of “software.” Eventually, the problem can no longer be ignored, at which point they attempt to solve the non-software problem as though it were a software problem and completely overlook the meatspace conditions of the underlying issue.
For the information age to reap its full benefits, and stop pissing off consumers, the next set of innovations need to focus on increasing meatspace process capacity. Innovations will use technology but can only be successful if they start by acknowledging the constraints of our physical world.
