Focus on performance marketers who have verifiable performance. You want someone who speaks in numbers (e.g., ask vlookup vs index/match type questions) and is driven to test-- so look for someone who can explain different strategies (e.g. alpha/beta, how they segregate kws/match types, thoughts on long tail, negative management, manually managing bids vs automated systems, campaign structures, even basic things like constrained budget allocation) in a manner you feel comfortable with. Given your niches and scale, maybe place more emphasis on testing novel channels to help establish both CACs and LTVs, because much of automotive search will be likely weirdly high funnel because it is a highly considered purchase category.

Marketers are going to be good self-promoters (it is their job, after all), and the good ones are worth a lot because they can save multiples of their comp in wasted spend, but most are unfortunately flunkies who parrot Google's talking points, always use automated bidding and constantly change KPIs.

That's the thing though; I have only passing familiarity with many of the concepts you listed. In order to evaluate someone on them, I would have to learn enough to differentiate a real expert from a talented bullshitter, and I'm frankly not confident in my ability to do so in the time I have available.

We've tried working with several marketing firms specializing in ppc, and after a couple of failures have found one that's getting decent results. They understand our business and can write reasonable ad copy. And the traffic driven by the ads is at a CPA that works for us, and has meets the KPIs I'm interested in on-site, in terms of conversions, time on site, retention, etc. However, I have no idea whether it's possible to do considerably better than they are. And as I mentioned initially, I feel the amount spent might be better put toward an in-house employee, who could devote greater effort than I assume the agency does, as well as performing non-ppc marketing tasks. But beyond just trying people who seem good until I find one who gets results, I'm not sure how to find that person. Maybe that's exactly what I should do.

I'd pick 3 people of your team who think they feel like they could produce some PPC ads. Then let each spend 100 bucks on ads. Then call up an agency, say you had bad experience. You give them 300 bucks but if they perform worse than the ones produced by your team with no expertise, you won't pay. If they outperform & you're happy, say what you're looking to spend.

Either way: you're just a few clicks & bucks away to test your ability. Especially in this space.

I locked myself out of a girls flat. She was hungry & freaking out, though i knew i was almost getting in, if i just had strong metal to slide-open the bolt. But stores were closed and i couldn't steel a spokes of a bike. So we called a locksmith. I told him i know we're being ripped off, but i am happily paying if he explains to me what he is doing and what alternatives there are. He opened the door in a few seconds and i paid 100 bucks. Inside i immediately ordered a lockpicking set & started to practice. I am not good at picking locks. But now i at least know when it's feasible to try myself and in which cases i should get ready to be ripped off again.

The ad space is full of people waiting to BS your money out of your pocket. Find out when it's ok to take BS without it leading to BS results – and when the BS predicts a BS outcome.

Or: Trust your gut. It's always right. And if it's not, try again & pretend it's never been wrong!

You need someone who understands PPC on the interview, even if you don't need to hire someone of that level or expertise.

Otherwise they can just unfortunately obscure performance and the way they manage their accounts won't be clear.

There's some good advice in this thread but some of the specifics aren't consistent with current best practice.

Context: 8 years PPC experience