It seems like every CRUD application under the sun has been built, and it is difficult to carve out any value that can be productized. Anyone know of anyone that started with Saas(Non AI) recently and was successful?
I felt the same way as a naive undergrad - "every piece of code worth writing is already in a library you can import". Very clearly not the case.
When it comes to SaaS, no one cares about AI vs non AI. What your customer cares about is "do you solve this problem that makes my job hell?"/"does this tool make me awesome at my job?"
How do I know this? I built the 200th uptime monitoring alternative to Pingdom four years ago, and I'm still getting customers in this age of AI.
And to add to that, as programmers I think we sometimes miss how inefficient most people’s workflows are. Ever watched a family member, friend, etc use a computer and you just want to pull out your hair? They are just used to it, “that’s how things are”. Even a simple tool can do wonders for someone’s workflow and can worth quite a bit to them even if you think it’s easy/obvious/just-use-X-tool.
Heck, you can even solve “simple” problems for developers and they will pay to avoid having to implement it themselves. Take a thorny process, wrap it up in a clean api with a nice UI and you’re set. I pay for multiple dev tools that I could write if I wanted to put in the effort [0] but it’s not worth the effort to keep up with it and paying $XX makes the problem go away, I’m sold.
[0] and I think we often underestimate what it would actually take to re-create a tool/SaaS. The core feature might be “easy” but all the other parts are what’s hard. Coding is rarely the hardest part of building a product IMHO.
Not even close. I work in German immigration. There are SaaS opportunities left and right. The same is true in every niche. Developers just tend to look at what's immediately around them, and not at all the industries that are chronically underserved by tech.
German everything is chrnonically underserved because the Germans refuse to automate anything.
I had to do residence and immigration documents in triplicate with old school carbon copy forms and show up to multiple appts in person. Meanwhile, I can renew my WA drivers license online from a different country.
This is a bit of an understatement. Germans are both actively hostile to changing processes, and eager to add layers of process to every imaginable thing.
The result is a growing bureaucracy with no matching effort in efficiency or digitalisation.
Exactly. It's not that SaaS is 'dead', the SaaS hype-cycle has passed it's peak and all the low hanging fruit has been snatched up already. You need solutions to actual problems to succeed now not just clever ideas.
There are apps that are solutions in search of a problem. They may be great solutions, but they're just not solving pain points that anyone wants to pay for.
SaaS products are just businesses. This like asking "Is retail dead? Every retail store seems to already exist."
There are always business opportunities to be had. You could probably open a business selling gravel and be wildly successful despite there already being tons of gravel suppliers in your area.
The term "CRUD application" triggers me, because it is mostly an oversimplification. Technically almost any application can be boiled down to these fundamental operations, but this level of abstraction is not very useful to talk about in terms of business value. Only because AI is hyped today means that it could solve any problem, let alone design applications anyone wants to use.
I think of a "CRUD application" as, hopefully, a thin wrapper around simple database transactions with a web front end. They're mind numbing to work on with most tasks boiling down to "add column to database and update the ui". They also tend to be in charge of more MRR than most well funded startups. There is a huge demand for them and if you're good at banging them out, your kids will never go hungry.
If anyone is looking for a good business model, write an application that makes it so that Person Type X can manage and access a database easily. There are plenty of examples. Salesforce is CRUD for Sales people. A lot of Oracle's software is CRUD for Industry Y.
Kind of, but if you've seen people trying to implement a chess game in Oracle Forms or similar you'll know that most people never stop at "a thin wrapper around simple database transactions"...
Like systems built in excel or locode alternatives, every CRUD application gets more and more convoluted over time and isn't "just CRUD" very soon after it's created.
> every CRUD application gets more and more convoluted over time and isn't "just CRUD" very soon after it's created.
I'm guessing tht's in large part because stuff just keeps getting added, with no refactoring / restructuring of the database. If you just keep adding columns (and tables) without refactoring your data model, stuff can easily get convoluted AF without actually needing to; the same funcionality could probably still be "just CRUD" if it were built on top of an appropriate data model. (I may be eaxggerating somewhat, but I think probably pretty much anything could.)
the majority of software is just a bunch of rats nest business logic and CRUD DB ops. hence it's a CRUD app. Queues, microservices etc are all over engineering for no reason
World would become a better place, and it didn't, so probably not.
And thinking that business is being made by fulfilling needs is very naive. Next big and successful SaaS do not need to do anything new, or anything better for that matter. It is enough that its marketing team would convince people otherwise.
An interesting detail is that if you look at that Wiki entry, there is an informal study showing that this “rule” is not accurate.
> In 2015, a study of 26,000 articles from 13 news sites on the World Wide Web, conducted by a data scientist and published on his blog, found that the majority (54 percent) were yes/no questions, which divided into 20 percent "yes" answers, 17 percent "no" answers and 16 percent whose answers he could not determine.
There is a difference between green field development done by already established companies and new companies entering the field. What was the last major acquisition or IPO from a new startup SaaS company?
This is what I meant by the lens one looks at this through. In this case, you're defining success for a SaaS as IPO/acquisition; neither are required for a thriving business. I use multiple SaaS products that I suspect will never be big enough for, in many cases due to their niche.
Mmm tough one, but I do not think so. I haven't personally seen anyone strike it rich with a straight-up, generic CRUD app lately. But I've seen some smaller successes where people have built SaaS tools to solve very niche problems within specific industries or communities. Think about it – there are still tons of industries that are surprisingly underserved by technology or even within established markets, there might be a very specific workflow or pain point that existing software doesn't address well - I came across Petex the other day, a Scottish company geared towards oil and gas and are making $$
No.. it’s not dead. But it is at a point where the bar to get traction is raising significantly. You can’t make something mediocre and get traction just because the alternative is pen and paper. Need to design something that sells well and provides value.
that means the lean startup method is dead because if you need to come out of the gate with a polished product, you need to build for a while before launching
Are you talking about the business model or the product?
Both a clear No. But the times of endless money for SaaS is likely gone and people are either looking for a complete SaaS&Hardware or for more intelligent SaaS that solve their problem differently.
One of the most interesting metrics I came across this week was the rise in custom built software over SaaS as AI enabled the “Build” in the “Build vs Buy” equation that has typically pushed companies to SaaS.
It's scary to think of how much software is being created right now by ChatGPT, Claude, etc by people who can't even read code. The house of cards being built today is going to lead to an epic collapse
A SaaS for something creative, sure. But a SaaS that deals with any kind of data, of course not, because you can't productize "we operate on your data but you can't trust the outputs lmao".
You would be surprised. If you throw a good implementation of AI into a call center, which still hasn’t been done that often even though I have done a couple, if it even has a 30% success rate at helping users without you having to talk to an agent that’s a huge win.
Every time you interact with a live person at a call center it costs $3-$5 and many people would much rather not deal with outsourced call agent.
When it comes to SaaS, no one cares about AI vs non AI. What your customer cares about is "do you solve this problem that makes my job hell?"/"does this tool make me awesome at my job?"
How do I know this? I built the 200th uptime monitoring alternative to Pingdom four years ago, and I'm still getting customers in this age of AI.