The hard part is knowing what "minor" changes needs to be made.

Unless you have evidence of what the problem is, you're just stabbing the dark. $100k is probably best spent on a session capture tool that allows you to see how/when/why users are dropping off, hire a UX designer who can think up solutions for the highlighted problems, then use the capture tool to gauge effectiveness and sneak up on a solution.

I won't deny that there are great UX people out there with a natural intuition for this stuff. But there's still the problem of figuring out which ones are great.

Old school UX used to involve video taping users interacting with software.

This is a qualified statement that really doesn't add up in my head. No minor change is going to triple your revenue, no matter what ui UX designer says.

I've seen redesign fail spectacularly and deliver incredible results - and they were done by the same people.

Sometimes following the best practices makes things 10 times worse.

Even when you throw AB testing and pretend you are following the data, the sample size changes so much you can't be sure off anything. I've seen A/B tested changes winning in the benchmarks perform terribly in the long run.

So would you have the same chance to hit the needed changes to triple your revenue by hiring a developer for 20$ per hour on freelancer? Absolutely, yes.

I'm not spending a penny on agencies, let the big companies do that

Sure, but that's not 2/3 minor changes.

> offering deals for people who buy right away without abandoning their carts or removing roadblocks from the purchasing process

There is a job in spam advertisement for you yet.