“They truly understand the problem I'm trying to solve, don't try to upsell me and help me realise the value of their product.”

I'm part of an invite-only community of TA leaders. It's the best community I've ever been part of - very different from a glorified Slack channel where people ask "can you recommend an agency?" It's much higher quality.

When we were evaluating a sourcing tool, I went to that community and said: those of you using AI sourcing tools, what do you like and not like? And I got fantastic feedback - the good, the bad and the ugly from actual end users.

I've been doing this for 17 years. I have a good enough network to leverage. I also speak to people at VCs who've invested in companies I've worked for - if they don't have insights, they can often connect me with somebody who does. And then newsletters, webinars, G2. But my first go-to, if I had to stack rank it, is that TA leadership network.

Completely. The phrase I hear from the C-suite is "do we buy it, build it, or acquire it?" I wouldn't go so far as to say I'm going to build an ATS myself. But our people analytics person has built amazing dashboards that replace what our ATS can't do. And now we're connecting agents to it so you can just talk to the data.

How I'd evaluate an ATS two years ago, I'd evaluate very differently now. The most important factor - does it have an MCP? Can I plug it into the AI platform I already live in? I don't want to work across ten different applications. I just want to do what I need to do in one place.

It's the little things that annoy me, but they add up cumulatively. The CSM was on vacation. My team chased three times - no response. I escalated it. "Oh, sorry, they're on vacation." I'm like, fine, but can you just put an out-of-office on and tell us who the alternative contact is? I've deliberately paid extra for a CSM.

Then there are latency issues. It's not quick enough. I don't know if that's the vendor's problem or something behind the scenes, but it all starts to add up and become a pain.

You get the razzle dazzle in the sales process. It seems to work in the demo. Then five or six weeks in, you're having challenges. And you start to question whether it lives up to what you were promised.

Yes. A lot of TA teams have a deluge of applications and want a tool to review them. I've resisted that. Ethically, I don't know how comfortable I feel - having been a job hunter myself - with AI reviewing applications. The vendor pushes the responsibility onto the customer by saying "you've got a human in the loop." But recruiters are under such pressure, they'll just take the top ten recommendations.

And these vendors are all using the same models behind the scenes. There's a lot of monoculturalism. If you apply to a hundred companies and they all use the same AI to filter, the algorithm could reject you across every single one. That has affected my reason not to move forward with purchasing that kind of technology.

I can't remember the last time a vendor said: "This is what success is going to look like for you after 30, 60, 90 days." Not just onboarding - that's different. I mean agreeing what success looks like for my specific team, with tangible numbers. So I can say: we agreed we'd move the needle here, the needle has moved, and I feel I've got value out of the product.