[mikert89]

When software becomes cheap to build, a lot of strange second-order effects kick in. It’s not just that products are easier to create; sales, marketing, and every other business function can iterate faster alongside them. That speed erodes moats. As this reality sinks in, I think we’re headed for a brutal shakeout in SaaS.

People still argue that distribution is the real bottleneck now. But when the product itself is trivial to build and change, the old dynamics break down. Historically, sales was hard because you had to design and refine a sales motion around a product that evolved slowly and carried real technical risk. You couldn’t afford to pour resources into distribution before the product stabilized, because getting it wrong was expensive.

That constraint is gone. The assumptions and equations we relied on to understand SaaS no longer apply—and the industry hasn’t fully internalized what that means yet.

[bflesch]

I think that we should discern between nonfunctional and functional requirements. Right now, implementation of functional requirements can be done very fast, with the effects that you mention.

But nonfunctional requirements such as reliability, performance and security are still extremely hard to get right, because they not only need code but many correct organizational decisions to be achieved.

As customers connect these nonfunctional requirements with a brand, I don't see how big SaaS players will have a problem.

For new brands, it's as hard as ever to establish trust. Maybe coding is a bit faster due to AI, but I'm not yet convinced that vibe coders are the people on top of which you can build a resilient organization that achieves excellence in nonfunction requirements.

[mikert89]

"As customers connect these nonfunctional requirements with a brand, I don't see how big SaaS players will have a problem."

Brand means almost nothing when a competitor can price the software at 90% cheaper. Which is what we are going to see

[bflesch]

I don't think you are correct. Take for example something like workday where companies manage their HR data. It's not only about the web interface but also about the fact that all of the implemented IT processes have been checked to be legally correct for each country, that the website is resilient to hackers and has availability that covers their customer's needs. You can't copy that without building a large org such as workday.

Even on a technical level the interfaces with country-specific legacy software used all over the place are so badly documented the AI won't help you to shortcut these kind of integrations. There are not 10k stackoverflow posts about each piece of niche software to train from.

[dgreensp]

By that logic, Microsoft’s brand means nothing when OpenOffice is free.

[mikert89]

Microsoft is a robust business, with corporate contracts going back 40 years. There are going to be exceptions and winners, and microsoft is probably a winner

[immibis]

Have you noticed how rapidly Microsoft is setting its own brands on fire? Most recently by abandoning the brand "Microsoft Office" and renaming that product to "Microsoft 365 Copilot App" instead.

[bflesch]

I also cannot understand how they mix up their brands so much, even people working in the MS ecosystem need to learn new brands every year.

[cosmic_cheese]

Quality may become the defining factor. Anybody can vibe code software, but only someone who is actually capable of programming themselves and has some understanding of what good UX looks like, how to fix bugs (often subtle and not caught by LLMs), and how to achieve high performance and low resource consumption can build high quality software.

I think developers who have an inclination towards UI/UX and a good grip on the technical side are particularly well positioned right now.

[mikert89]

Right now, you can take the raw html and css of a competitors website/saas product, give it to claude code, and it will rewrite your react/nextjs app to match the quality and style of the assets you gave it. You are underestimating how quickly these supposed moats have been decimated

[cosmic_cheese]

I’ve been using Claude and it’s decent… sometimes. It’s still capable of making rookie mistakes and it will happily let those accrue if you don’t point them out (and even then, occasionally it won’t see the problem). Personal technical familiarity with the domain still brings a lot to the table.

[anonu]

> Historically, sales was hard because you had to design and refine a sales motion around a product

while your statement is true - this is actually a very minor reason why sales is hard.

[mikert89]

I’m well aware of what makes sales difficult—I’ve lived it, both in early-stage, venture-backed environments and in long, enterprise sales cycles. In SaaS, almost everything about how sales works ultimately ties back to the cost of building software. Relationships and customer trust absolutely matter, but when building SaaS becomes trivial, the underlying equations change—and many of the old assumptions stop holding.

[sdf2]

Only people who rely on things like frameworks live on assumptions. I guess you fit into that group of people.

[baq]

Only a third of the most is gone - development. Validation and infrastructure/operations are still alive and well, though LLMs make for decent system investigators.

[jongjong]

No moat? In the software industry?

My decades of experience suggests that the opposite will happen. People will realize that the software industry is 100% moat and 0% castle.

People will build great software that nobody will use while a few companies will continue to dominate with vaporware.

[andsoitis]

> few companies will continue to dominate with vaporware.

that makes no sense. "Dominate" implies people use or buy you software. If you produce nothing ("vaporware") how can you dominate?

[jongjong]

Most modern software is horrible. Linux is objectively better than Windows and cheaper but it has much smaller market share. Same story with OracleDB vs PostgresDB. Developers recently identified a critical vulnerability in MongoDB that had been there for 10 years, yet RethinkDB, a superior product was shut down (at least the company) because they couldn't find enough customers. There are so many such cases.

They're vaporware in relation to what else is available.

[mikert89]

Yes, the moat is disappearing. Go look at any stock index of publicly traded saas companies, they have all been selling off 50% in the last six months. This is going to be a bloodbath

[usefulcat]

Some are down over the previous 6 months, but not by 50%. At least among the largest.

    Company:     6 month change:
    ============================
    Palantir:    +20%
    Salesforce:  -9%
    Shopify:     +40%
    Intuit:      -26%
    ServiceNow:  -30%
    Adobe:       -17%
    CrowdStrike: -1.5%
    Snowflake:   -3%
    Cloudflare:  0%
    Autodesk:    -9%

Above companies are the 10 largest from this list:

https://www.mikesonders.com/largest-saas-companies/

[kubb]

OTOH there’s no reason not to hire a small in house dev team to make bespoke software for any company.

Except for the token cost maybe.

[anonu]

Thats just not true. $CRM is at worst flat looking back 1 year. $IGV, software ETF is the same. Definitely not down 50%.

[jongjong]

The moat is marketing and social media algorithms. These tech companies are subsidized by investors and drive up the CPC to levels that don't make financial sense this is the moat.

If a superior product launches without marketing, it will not get a single customer; it's incredibly hard to spread through word of mouth these days... With paid advertising, you end up paying Facebook inflated CPC to get bot traffic!

[sdf2]

Lol what a nonsensical post. Stock prices? Your argument would hold more weight in regards to talking about revenues which are a reflection of sales performance. But stock prices? hahaaha.

[corimaith]

The moat for artists hasn't actually gone down though. Even for those in Patreon with their styles directly trained are raking the same subscribers they have.

[raincole]

What does 'even' here mean? Patreon artists are doing fine because they're more like influencers. Artists looking for jobs in the industry (especially graphic design, but also film and game) are the ones who are in trouble.

[mikert89]

Really? Because UX Design, Graphic Design, and other "corporate creative" roles are essentially obsolete and these people cannot get hired.

[reb]

Canva destroyed graphic design well before LLMs caught up, but UX is still (somewhat surprisingly) on the ropes.

My bet: front end devs who need mocks to build something that looks nice get crowded out by UX designers with taste as code generation moves further into "good enough" territory.

Then those designers get crowded out as taste generation moves into "good enough" territory.

[puzzlingcaptcha]

You can easily tell UX design roles are obsolete by the recent threads about Tahoe and W11. There is going to be whiplash.