Legacy Software Is Getting Stickier
I am observing a pattern where small businesses’ relationships with their existing software providers are becoming MORE sticky due to AI. And this is not from any AI tools being released by those providers (there are some great ones… and there are some poor ones).
I’ve been conducting one-hour 1-on-1s with small business owners for the past few months. When it comes to AI, I have met with the full spectrum: from novices to businesses that have implemented impressive workflows. In every follow-up conversation, each business owner has told me they are doubling down on an existing software solution that was previously cited as a pain point. This has included multiple acknowledgements that they were dropping plans to migrate to another software provider.
Four Businesses That Stayed
The clearest example was a fitness studio owner. Her studio’s website had been in the middle of a move to a new platform for over a year. The move was paid for, it still wasn’t live, and every month it stayed unfinished was another month she put off looking for new members.
In one hour she handed Claude her existing site and the artwork she already loved, and by the end of the session she had a homepage she was proud to market. By the next morning she was editing copy and swapping photos herself, and she had canceled her call with the website designer. The migration she had been living with for a year was gone. She texted me that morning:
“I was awake at like 2 AM thinking about all the ways the robot can help me… my mind has just gone bananas about ways that this cannot only help me but help so many people in my field.”
(Her full story is on the results page.)
Three other businesses faced the same choice. One company was weighing a complex marketing platform against staying on its existing website platform. Another ran a small business ERP and was considering the leap to a much larger one. A third was deciding whether to buy a data collection product or keep using the one it had hand-rolled.
The complaints about the existing software were legitimate. But in every case, extending the relationship with the existing software turned out to be less painful than migrating. Once the owner saw how far a few small tools could stretch what they already had, the question changed from “which platform do we move to” to “how do we make this one last longer.”
I think the reason is plain economics. Migration carries a cost that is easy to underestimate. Data has to move, staff have to relearn, a vendor relationship has to be rebuilt, and there are months where nothing quite works. A helper that sits next to the existing software carries almost none of it.
What They Build Instead
What I’m not seeing is just as interesting. There’s a lot of talk right now about “roll your own SaaS,” the idea that a business with these tools will build its own replacement for the software it pays for. None of these businesses are doing it. Building your own platform carries every cost of a migration plus the burden of maintaining it forever.
What they build is smaller: a widget, a helper, a skill that takes the class schedule from a spreadsheet and turns it into the screen on the studio wall, or a report the vendor never offered. Each one fills a gap in the existing software rather than replacing it.
Most of the software these businesses run has no official support for Claude or ChatGPT, and it doesn’t need it. The studio owner’s website had no AI integration at all. Claude worked on it through the browser, the same way she would. If a person can log in and click through it, these tools can work with it.
What I’m Watching
This is anecdotal, not a study, and it may be a short-term effect. AI-native applications are coming, and in five years the software these businesses run may be competing with programs that don’t exist today. A product built from the ground up around these tools could change the math.
For now, the pattern is consistent enough that I’m spending more time thinking about it. The pain points that used to push a small business toward a new platform are the first ones it is now able to work around. That makes the software that business is currently invested in stickier than it has ever been.
This is a field note from the Urnicus practice. For more writing, visit the field notes.