The answer has nothing to do with rationing. When all software is designed to grab the most attention, serve the most adds, collect the most data, deliver the maximum amount of the cheapest content and generally be as exploitive of its users as possible, no amount of rationing or not rationing will effect it.
McMaster Carr, a company worth a little over a billion, sells metal, bolts, casters, and many other industrial bits and bobs. Their website is infinitely better than Amazons. Why? It has nothing to do with rationing and everything to do with trying to actual deliver a good customer experience.
I'm even more cynical; most software from large organisations is bad because it was never designed for the user, it was designed to make the product manager's career advance.
Two factors:
- middle management do not have contact with either customers or shareholders. They report to managers, and only have managers reporting to them. They're insulated from the actual consequences of their decisions.
- the only way to get recognised for promotion is to create change. Any change. Doesn't matter if it's bad (blame can be shifted), it just needs to be noticed.
If a feature of the product hasn't changed for a long time (because it works for the user), it becomes more of a target for changing. If it changed recently, it creates less impact if it changes again, so isn't a good target. So the product becomes worse over time, not better.
> it was designed to make the product manager's career advance.
Not just product managers, lots of programmers are guilty of implementing unnecessary technology (e.g. microservices) or the framework du jour in the name of resume driven development.
Ironically, so many modern tech companies are arguably less incentivized to provide a good technology experience than non-tech companies, since the tech company's value/revenue primarily comes from directly exploiting their users for their data and/or attention. Improving the user experience frequently directly reduces their revenue. No amount of free LLM usage can allow a video serving company to both serve ad-free videos and make money from ads in their videos.
Though, maybe the irony only exists because we use the broad term tech company instead of the more generally accurate terms data harvesting company and attention consuming company.
Mcmaster also isn't (somewhat famously) competing on price. They won't rip you off, but you can't sort by price. There are no sales. With limited exception, they won't list brands on the site. (But if you contact them, they'll gladly tell you who so you can buy direct.)
Their target market is people who need to fix something quickly or get some industrial supply item tomorrow so work isn't stopped.
Massive corporations save money via the end user by taking a chainsaw to customer service. Large tech companies (and many other industries) will just quietly fuck the regular customer over because they can't profitably do customer service at scale. Even Amazon's effectively no questions asked return policy is a band-aid over the forgeries and trash that fill their hyperscale market.
If you have a problem with an item that Mcmaster sends you, they're going to go have a stern talk with the manufacturer. They sell reliability, not race to the bottom capitalism.
(I've also heard that they really take care of their employees. People in jobs that are automated away are always retrained; they'll pay for college with no qualifications, because education is good even if they don't directly benefit, etc.)
> McMaster Carr, a company worth a little over a billion, sells metal, bolts, casters, and many other industrial bits and bobs. Their website is infinitely better than Amazons
I’ve bought a few dozen things from McMaster. For very competent people who know precisely what they want it is good, but I do find I spend an hour looking at the specs of “doodad” which will almost meet my needs when I stumble onto “doodaddy” which is a better fit. After I buy it, a friend points out they sell “doodada” which is even better.
Note I’m an engineer, but didn’t study in North America, so the names of things are different.
McMaster is terrible for product discovery, you have to know the precise name of the thing you want.
For the average person on the street Amazon is better.
I find McMaster awesome for product discovery. I can go and just start scrolling, see a interesting category and start digging into what is there. How on earth is Amazon better for this? Instead of just scrolling down and starting to click I have to open the side menu, expand categories, click on industrial & scientific, click on show all, get smacked in the face by a full page add, and then I get a nonsensical mix of lists of categories, new products, then back to a list of categories to hopefully get to where I want to go.
You need help finding something? On McMaster, scroll down, click contact and send a message. On Amazon? Google to find out where the shortest path to getting assistance that they have buried under a dozen submenus.
I will admit, there is one superior thing for product discovery, which is of course their paper catalog.
Agree that incentives explain a lot of why software is bad. But I don’t think they explain why it all looks the same. For instance, every code forge copied the pull request, including the ones with no ads to serve.
Crucially, McMaster-Carr is a private company, so their customers are the people giving them money for goods and services. As a public company, Amazon's customers are the gamblers on the stock market.
Software engineers need to grow up, at the turn of the century when software was starting to remake the world, they developed an enormous conceit and started believing that the world revolved around their work and they were going to shape it.
It doesn't. The world revolves around economics and politics, just like it always did. It took a decade or two for established institutions to catch up and re-assert control, but catch up they did. In the end that 10-20 years was so brief that in a few generations it'll be entirely forgotten.
Now during that time who knows? Maybe software engineers could have held their ground and used their leverage to influence the political and economic system and turn it into something better. Did they do this? 99% did not. Instead, they coveted cushy jobs at Big Tech where they gave up their freedom and their ability to disrupt anything in exchange for security, and resigned themselves to trying to sound smart in blog posts where they produce grandiose but ultimately sophomoric opinion articles about their industry. No one with actual power could care less about these articles as they follow the rules of power, not the rules of engineering. They own the IP. They decide what "owning IP" even means. Their economy of IP is essentially one of monopolies. That is why everything is basically shit. Read Zero to One and perhaps a few chapters out of Wealth of Nations. The full explanation of "why our software isn't 10x better" will be laid out for you in a weekend.
The software engineers gave up their agency for a salary. Hey, that's fine. That's what 99% of all engineers in human history did. They've gone to work for a government or a big corporation, and let the guys with power keep making the rules. You just did what other engineers did because you ultimately needed to pay bills, take care of your families, etc. You're no different. But software engineers, pretending you are is getting old.
The rebel path where you actually make your own shit is still there. You can start a company. Or a nonprofit. You can donate your time toward building, marketing, and lobbying the government for something better. Because god knows all three of these are needed in spades. Without evidence of that though, honestly I'm tired of software engineers proclaiming something new from Big Tech is going to change the world. It's not. Innovation alone will solve nothing. The guys in charge have put their feet down and you work for them now. A revolution will be required in a pretty literal sense of the term to change this. Using the latest Anthropic models harder isn't going to do shit.
I'm an engineer too. Let's grow up already. Let's either get serious about our conceit that tech can improve the world, or let's resign ourselves to being just like every other engineering profession in the world, getting paid well to obediently build what the investor class tells us to build. You can't have it both ways. Choose your destiny and accept it.
(Fyi - serial bootstrapped open source adjacent founder here, and I made my own decision 15 years ago, one way or another, I'm dying free.)
The answer has nothing to do with rationing. When all software is designed to grab the most attention, serve the most adds, collect the most data, deliver the maximum amount of the cheapest content and generally be as exploitive of its users as possible, no amount of rationing or not rationing will effect it.
McMaster Carr, a company worth a little over a billion, sells metal, bolts, casters, and many other industrial bits and bobs. Their website is infinitely better than Amazons. Why? It has nothing to do with rationing and everything to do with trying to actual deliver a good customer experience.