If you ask almost any sign company owner what they want, faster estimates will usually be near the top of the list.
The interesting part is that speed stops being the priority the moment confidence starts to disappear.
People don’t really want fast estimates.
They want estimates that are fast enough and trustworthy enough.
Finding the balance between those two things turns out to be much harder than I originally thought.
⸻
One thing I’ve noticed while building Ryvet is that everybody loves the idea of faster estimating.
That shouldn’t be surprising. Estimates create bottlenecks in a lot of shops. Customers are waiting, salespeople are waiting, projects can’t move forward until somebody puts a number together, and there’s never quite enough time to get through all the work sitting in the queue.
On paper, faster sounds great. The problem is that speed isn’t actually the goal. Confidence is.
I’ve seen this play out repeatedly during beta testing. A customer uploads a project, Ryvet generates an estimate, and one of the first things they do isn’t look at how fast it finished. They start reviewing it.
They check the labor.
They review the materials.
They compare it to how they would approach the project.
They look for things they agree with and things they don’t.
In other words, they do exactly what an experienced estimator should do.
That got me thinking about something. Most of us have had experiences where we got a number quickly but didn’t really trust it. Maybe it came from software. Maybe it came from a spreadsheet somebody else built years ago. Maybe it came from a vendor budget price. The source almost doesn’t matter.
When confidence is low, speed loses a lot of its value. You end up double-checking everything anyway.
In some cases, you spend so much time validating the estimate that you give back most of the time you saved.
That’s why I don’t think the future of estimating is about removing people from the process.
The more customer conversations I have, the more convinced I become that people want help, not replacement.
They want a starting point.
They want a second set of eyes.
They want something that helps them move faster without increasing risk.
Those are very different goals than fully automated estimating.
Owning a sign company has probably made me more skeptical about automation than I otherwise would be. I’ve seen too many situations where small details make a big difference. A drawing can look straightforward and still contain assumptions that completely change the job. A project can appear simple until somebody visits the site. A customer can approve everything and then decide they want something different.
That’s why judgment matters.
It’s also why trust matters.
One of the encouraging things I’ve seen during the Ryvet beta is that customers don’t seem interested in blindly trusting software. They challenge it. They modify it. They use it as part of their process instead of replacing their process.
Honestly, that’s how I use it too.
If I generate an estimate, I still want to understand it. I still want to know why it arrived at certain conclusions. I still want to review assumptions before sending anything to a customer.
The goal isn’t eliminating human involvement.
The goal is eliminating unnecessary work.
I think that’s where a lot of software companies get things wrong. They focus on how much faster something can be without spending enough time thinking about confidence. Customers care about both.
An estimate that’s 20% faster but introduces doubt probably isn’t an improvement.
An estimate that’s significantly faster while maintaining confidence is a completely different story.
The longer I’ve worked on Ryvet, the more I’ve come to believe that trust is the real product.
Speed gets people’s attention.
Trust is what determines whether they actually use it.