There's a reason we hire plumbers...to fix DIY projects gone wrong. But, we don't hire plumbers to work on our teeth. They are great at what they do, they know that, they stay in their lane. And, the DIYers learn some very valuable lessons along the way.
This is why building your own CRM is a really bad idea. That being said, not all plumbers are the same and not all their solutions are the same. Some are clearly better at what they do. It's up to the customer to do their research and hire the best for the job to be done.
The custom CRM arc gets underweighted here I think. The build cost went to near zero but the ownership cost didn't move, whoever ships v1 ends up the accidental product manager of their own creation. In regulated industries the single-user software wave lands as a governance problem well before anyone gets around to pricing it. Two dozen personal agents inside a bank is two dozen tools producing outputs nobody reviewed, and once machine-generated code outruns anyone's understanding of the codebase, explaining to an examiner why a tool did what it did gets genuinely hard.
Connecting the dots, it's possible that single-user (and small-team) software built on top of SaaS can be value-accretive to both parties.
Even after accepting the advantages of SaaS—security, maintenance, data complexity, and the need for a shared source of truth—limitations still exist. High cost is often one, but more significant is that these platforms are built to be most things to most people. As such, they tend to be either overly abstracted, built to serve the lowest common denominator, or both. This sets a ceiling on the value available to customers.
In comes AI coding tools, which make it tractable for single users and small teams to build bespoke solutions that leverage the benefits a SaaS platform can provide. What was a ceiling becomes the floor. This increase in value capture by users, combined with the lower baseline cost for SaaS noted in the article, expands the provider's addressable market. This could be a case where lower margins are offset by higher volume.
There's a reason we hire plumbers...to fix DIY projects gone wrong. But, we don't hire plumbers to work on our teeth. They are great at what they do, they know that, they stay in their lane. And, the DIYers learn some very valuable lessons along the way.
This is why building your own CRM is a really bad idea. That being said, not all plumbers are the same and not all their solutions are the same. Some are clearly better at what they do. It's up to the customer to do their research and hire the best for the job to be done.
The analogy between writing bespoke software and creating Excel/PowerPoints is apt.
The custom CRM arc gets underweighted here I think. The build cost went to near zero but the ownership cost didn't move, whoever ships v1 ends up the accidental product manager of their own creation. In regulated industries the single-user software wave lands as a governance problem well before anyone gets around to pricing it. Two dozen personal agents inside a bank is two dozen tools producing outputs nobody reviewed, and once machine-generated code outruns anyone's understanding of the codebase, explaining to an examiner why a tool did what it did gets genuinely hard.
Connecting the dots, it's possible that single-user (and small-team) software built on top of SaaS can be value-accretive to both parties.
Even after accepting the advantages of SaaS—security, maintenance, data complexity, and the need for a shared source of truth—limitations still exist. High cost is often one, but more significant is that these platforms are built to be most things to most people. As such, they tend to be either overly abstracted, built to serve the lowest common denominator, or both. This sets a ceiling on the value available to customers.
In comes AI coding tools, which make it tractable for single users and small teams to build bespoke solutions that leverage the benefits a SaaS platform can provide. What was a ceiling becomes the floor. This increase in value capture by users, combined with the lower baseline cost for SaaS noted in the article, expands the provider's addressable market. This could be a case where lower margins are offset by higher volume.