Why did you make this fake site?
I like the concept. I don't have any strong mutual connections to Chamath, I made this on the off chance that I can get his attention to learn more about his plan. If you're reading this Chamath, let's chat! I'll happily hand this site over, and I'm happy to work with your team to build the real one.
Thanks I couldn't find any other info on it except what Chamath had said on a podcast, so I gave a link to the "fake" site since it explains the concept behind 80/90.
Do any self hosting sites include an "effort index"? For example, if its a binary which needs Sqlite/MySql and Nginx, it gets a perfect score; if it needs 5 docker containers...
On the contrary. Handwaving "oh bring your own DB" is drastically more complex than having a ready-made docker-compose.yml which includes said DB and whatever else is needed.
Ahh the lost art of writing installable deployable software.
Another thing I don’t like about SaaS is that most of it is a yarn ball held together by snot and twist ties that can run on only one cloud… or only one cloud account with hard coded details about that account. As such it’s encouraged a generation of developers to write this slop.
> Ahh the lost art of writing installable deployable software.
Yeah, because it's so easy to create a deployable package for a variety of distros. Especially if you have any system-level dependencies. Even more so for shitty unreliable ecosystems such as Perl, Python and Node.js.
There is a reason why Docker and friends took over deployment artifacts. Having a single reproducible (if properly done) immutable all-dependencies included artifact that will run everywhere with a container runtime is just magic. Gone are the days of "You need libcgit v2 to use my package... oh Ubuntu calls this libcgit2, but wait, I'm on Ubuntu 16.04 which only has libcgit51".
It just makes deploying much easier. I don't have to worry about system dependencies or some random trash being left on my filesystem, since the container has all of its dependencies and it only has access to whatever parts of the filesystem I give it.
We as Software developers need to be honest with ourselves and really think about pricing. The product should be priced as per the effort and time it took to develop. If application needs maintenance then charge it separately.
But the customer should not have to pay for your SaaS because you don't want to work 9 to 5.
>> The product should be priced as per the effort and time it took to develop
When I started programming I costed projects like this. Now I price things based on their value to the customer.
There were a number of reasons for the switch;
Firstly in the earlier model all I'm selling is my time. All time I don't have work is lost, and I can't make it back. This resulted in years of being underpaid because a month of no work was lost forever.
There are lots of things you need to do running a business test are not "billable". Finding new eork is one of them. So costing just the time you spent on the app is naive.
Pricing based on effort also allows no opportunity to grow. If I hire an employee I now need to find work for her as well. And so on.
Lastly, when making products (to sell over and over) you can't base price on cost because we don't know up-front how many copies we'll sell.
So we look at the value to the customer. Things that offer them higher value cost more. Sometimes there's a surplus, and that allows new, less certain, products to be developed. Sometimes it takes years for a product to return the investment.
We moved some (not all) products to a subscription when it became obvious that it was not possible to support all existing clients from only new-sales revenue. We needed to hire full-time supporters and since they add value to existing customers they need to be paid for by customers.
This means that if sales dried up tomorrow , existing customers could still get support ad infinitum. This is -much- more valuable to them than some small subscription amount.
So here's my suggestion. If a service is costing you more than the value you are getting, just stop using it. By definition you will be ahead. You don't need yo feel aggrieved at how other people price their stuff. That's on rhem, not you.
> When I started programming I costed projects like this. Now I price things based on their value to the customer.
What would the world look like if everyone was using this mindset? Water is VERY valuable to customers for example. So is food. Tech people have a very distorted ways to look at the value of their work and I say this as someone who works in tech.
but everyone is using this mindset. on average in an economy, prices are more or less the maximum what a customer might pay. If it is more, it is not bought obviously, and if it is less, it sure will be more sooner or later, because why not.
What is preventing water (and e.g. health outside of the US) and schools etc. from getting super expensive is usually that they are not produced and sold as a commodity on a free market with competitors, but provided in a regulated, often state-owned way where market mechanisms do not (fully) apply. Where it does, things get ugly. Google "water war bolivia".
edit, to comment the original statement:
Price must be >= cost of production, obviously. But it must also be acceptable to the customer. So you go as high as you can and if that is still not enough to cover your costs, then you stop your business and do something else. If it is very much higher than you need, sooner or later a competitor will notice the extraordinary profits in your sector and do what you do at a lower price.
Saying that "prices are more or less the maximum what a customer might pay" is different than saying "I price things based on their value to the customer."
Because in tech that might be doable but again, the "value to customer" of something like food is obviously infinitely high because without it you ain't gonna go much far.
And there's a reason why generally speaking countries have laws to prevent price gouging around a certain types of goods. Becuase without them, I am 100% sure that some dickheads would try to screw people over because that is what capitalism is producing.
But also, this idea that you should be paid based on the value I get out of your work is bizarre to me. if you are self employed or you run a business, you decide how much you want to charge for a product or a service.
> Because the more you get out of it, the more you are willing to pay for it.
In "normal" contexts, sure. Would you be happy if doctors were using this aproach and just asked you insane amount of money simply because you obviously want and need their service?
The thing I'm trying to say is that this approach doesn't apply equally across the work landscape.
Hypotetical: let's say you're a developer and you code me an ecommerce site. It takes you a month to do it and you charge me 10k. Should you be paid more if I make 1M using that site? What if I make 100M?
Does the value of your work changes based on how much value I get out of it?
The problem with the "how much it cost to make" approach is that means that there is literally no room for error. You're not building in any cushion - no shock absorber. One bad idea, one dud product etc and the company goes under.
On the other hand there are _lots_ of examples of value based pricing outside the tech world. some examples;
a) Any "luxury good". Think Ferrari, Birkenbag, designer clothes etc.
b) Any supply-constrained service - think rounds of golf at St Andrews.
c) Popcorn at the movies
d) Movie tickets (they are not dependent on the cost, or earnings, of a movie)
e) Office rent. It costs the same to build a building, so why do rents vary so much depending on location?
and so on.
value-based-pricing is not a tech thing - it exists everywhere.
I'm not arguing for a "how much it cost to make" approach. I'm arguing agains the "Now I price things based on their value to the customer" approach.
Because the notion of "value to customer" means nothing. What is the value of a piece of software? What is the value of a website? What is the value of branding? The problem I have with all your examples is that those are all tangible things while software work is, for the most part, a service. Coding a website for a client doesn't cost me anything in a traditional, material sense.
I can't quantify how much it cost to make a site. But I do have to assign a monetary value to my work as a developer. And figuring out that value based on how much value my client gets out of my work is a conceptually flawed aproach IMO.
The problem with one-off pricing is that people generally don't want just one drop of software. They want ongoing updates and support. Which means if your pricing model is only one payment, in that world you're going to have a future cost stream but no future revenue stream to support it.
That's why software businesses with a one payment pricing structure often end up going bust - they rely on a structure where they need constant growth in their customer base to sustain support of existing customers. The software coompany going bust isn't in the interests of customers who want support either so this model just doesn't really work for anyone.
If a one-off payment is matched by a customer expectation of a one-off software install with little or no ongoing support then that's probably fine.
That’s not how price works. Price is a function of supply and demand. The cost of goods sold is just part of what goes into price. So if you make something really compelling or scarce that people are willing to pay a lot of money for, you can charge more for it than it costs you to produce it. That’s business.
Lesson one is to never go public. As soon as you have shareholders, they're going to demand growth, and there's only so much you can do to sustain that growth before you have to raise prices.
Agree. However, people are allowed to price apps however they want (ie. SaaS model), but shouldn't be surprised if they are overcharging and someone else makes something similar and charges less.
I find it very interesting that you're been downvoted for this take which I don't personally find all that wrong. But I guess different people have different ways to look at the world.
All coins have 2 sides. Paying once is great, and if uou are happy then that's good. I use a bunch of software purchased on this model.
It's worth understanding though that the seller owes you nothing after this. Not support. Not development. Not keeping up with API or whatever changes. Not bug fixes (although most everyone ships some free updates for a while.) Obviously not hosting.
And for lots of software this model is perfect. For other stuff it's imperfect (as anyone who desperately needs Google Support will sympathize with.)
You get to choose what software I'd on your business-critical path, and what is on your "can replace anytime" path.
The concept is good but the site design is just awful. The page is non responsive. Can't search and the scroll is very slow. Not sure why a such a simple page is designed to take up so much CPU.
First entry I see: Pabbly Connect [0], a subscription web service that normally runs $20 a month and currently has a lifetime subscription offer advertised. Everything is hosted in the cloud, you sign in to access the product, they're SOC2 compliant, etc.
This website has a very different definition of SaaS than I do, and it sounds like it rhymes with "subscription".
A few more examples to make sure it wasn't a one-off:
* Foxly, a URL shortener with a subscription or lifetime deal [1].
* CountVisits, web analytics. This one doesn't offer a subscription, but it's still all hosted on their servers and therefore has usage limits (which you can pay to upgrade) [2]
Yes, I wasn't sure how to categorize these. This is V1 of the site and I will make adjustments based on feedback. There are no Non SaaS alternatives to those products, so I included them even though they are technically SaaS with a lifetime deal.
If I were you I'd drop them outright. Better to have the directory live up to its name than to have it be more comprehensive but contain apps that are clearly SaaS.
8090ish is a fake site.
https://www.8090ish.com/