I think pure flat rate pricing could be difficult for some AI products, especially when usage and inference costs vary so much, a simple base price with clear usage limits or predictable tiers might be a better balance, customers get transparency without putting the entire cost risk on the SaaS company.

I think the moat point is one of the most important parts here, if AI makes basic features easier and faster to build, having good code alone won’t be enough to stand out, deep integrations, useful proprietary data, and real understanding of a specific industry are much harder for competitors to copy.

I think that last point is really important, people can be genuinely interested in a product and still never pay for it, early feedback is useful, but actual willingness to pay is what really shows whether you’ve solved a problem people value enough to spend money on.

I think the hybrid model is probably the most practical middle ground, predictable pricing makes budgeting easier, while usage or value based add ons still let SaaS companies capture the upside when customers get more from the product, the hard part is keeping the pricing simple enough that customers actually understand what they’re paying for.

What stands out to me is how much this changes the idea of what customers are actually buying, if the software is doing the work instead of just giving people tools to use, then paying for access starts to feel outdated, the companies that adapt around actual results will probably have the strongest position.

I think the biggest shift is really from paying for access to paying for what the software actually accomplishes, SaaS isn’t necessarily going away, but companies that don’t adapt their pricing and value proposition to that change could have a much harder time competing.

The interesting part is that pricing will probably have to become much more tied to actual usage and results, if one AI agent can handle work that used to require several people, charging simply by the number of human users starts feeling pretty disconnected from the value the software provides.

I think the point about looking at what actually changes after using a tool is spot on, a feature can sound impressive, but if the team still has to spend time fixing, checking, or managing everything, it hasn’t really saved much work.

I agree, especially with AI changing how SaaS is priced, i’d rather pay a little more for a tool that genuinely saves time and fits my workflow than choose the cheapest option and end up dealing with extra work or limitations.

The move from seat based pricing to outcome based pricing could be one of the biggest changes here, if an AI agent is doing the work instead of just helping a person do it, customers will naturally start asking what results they’re actually getting for their money, that’s going to force a lot of SaaS companies to rethink how they package and price their products.

One thing that stood out to me is how AI is changing the developer’s role rather than simply replacing parts of it, if agents can handle more of the implementation, knowing how to review the output and spot problems becomes a much bigger skill, faster coding is great, but knowing whether the code is actually right is where the real value is.

I like that it focuses on practical value instead of just listing new tools, with so many SaaS products launching, seeing how they actually perform, scale, and fit into real workflows makes choosing the right one much easier.

I’d look at real world performance first, a product can have great features and still be frustrating if it’s slow, unreliable, or doesn’t actually save time for me, the real test is whether it makes the workflow easier after you’ve used it for a while.

I’d say the biggest sign of value is whether you actually miss the product when it’s gone, if it saves time, fits into the workflow, and quietly handles work in the background, that’s much more meaningful than having a huge list of features.

The knowledge base point is probably one of the most overlooked parts of AI support, you can have a great AI agent, but if the information behind it is outdated or messy, the customer experience will suffer, good documentation really is the foundation.