How do you come up with ideas and validate them?

I have a spreadsheet I could share here but basically I list my business ideas I try to rank with different criterias

• Competition: from 0 (a lot of competitors with good products) to 4 (no competition at all)
• Complexity: from 1 (requires a state of the art solution that requires a lot of r&d) to 3 (easy to have a MVP and first good product in a month)
• Reach: How many people might need this ? I don't do market research but write roughly a number : 100, 1000, 100000, ...
• ARPU: I estimate this to have an order of idea. Having looked at competition helps. If you have no idea it should be tied to what you out for competition and complexity

This then gives me two scores:

• Value = reach x ARPU
• Difficulty = competition x complexity

I then combine again the two scores to grade my ideas with : value x difficulty
A+ if > 10M
A if > 1M
B if > 100k
C if > 10k
Otherwise D

This is far from perfect but helps to eliminate ideas I don't want (for example I don't want to spend months on complex product development if the potential user base and/or revenue is not huge). Then you can invest some more time researching your ideas that have a good grade.

Are there additional pains that your clients have related to looking for talent?
Do they need help with writing the job post? Maybe don't have time to interview all the prospects. Perhaps you could help them with setting up a test to pre-qualify them. Or you could help them calling their references.
I would try to leverage your current client base and industry knowledge. Interview them, buy them a coffee, and ask about their most painful problems regarding what they are trying to accomplish, and how are they trying to solve them now.
The Jobs-to-be-Done methodology could prove useful to get this kind of insights.
Once you define a clear "job" or need, you can brainstorm potential solutions, build an MVP and thest them.

It really depends on the type of idea but in general you have to talk to people about your ideas to start validating them. If you think you can find your potential early adopters on Twitter or LinkedIn go there and try and talk to people there. See what they think. Prepare a form to validate a few of your assumptions and if it's the case make a landing page to validate if there is market. Use it to understand if the way you phrase your idea on a landing page resonates with what people are looking for and iterate on that. Again, only part of this might make sense depending on what kind of idea you have but all I'm trying to say is that you gotta find the shortest path to learn about your ideas if you have plenty of them and don't know which to put the majority of your efforts on.

I like the thoughts of Ash Maury in scaling lean and the value proposition canvas of Alex Osterwalder. Validate the problem by interviewing ~10 people. Use the mom test as guidance. You learn so much about what your customers think and how easily you can reach out to them. That's easily feasible in a week and you will have actionable data to reinforce your decision to move ahead.

I have heard of the book 'Will it fly?" to help with that. I would fill lean canvas for your ideas, create some crazy 8's to develop and maybe pivot the concept. I'd say these two steps will give you a better view on how strong or how much potential the ideas have. Then you need to test with users. The problem, nowadays, is that there is already a lot out there, and with great quality. So cheap prototyping doesn't quite work as it used to. Smoke testing is a way of working around it, for example.

I am somewhat in the same position as you, so I can only wish you the best luck.