The CEO who paused their EVP over AI search

I recently heard in a conversation about an employer brand rollout that got held up.

Not because of budget. Not because of messaging. Not because stakeholders couldn't align.

Because C-suite leadership started asking, "how do we control our perception and narrative through AI models?"

Most of us are still optimizing for a world where candidates start on Indeed or LinkedIn. Where they search "marketing jobs near me" and click through five job boards before landing on a career site.

That's not how discovery works anymore.

Around 40% of candidates are already using generative AI tools (ChatGPT, Perplexitiy, Gemini, etc.) to make their job search more efficient.

Not for mass-applying (that's a different AI problem). But for research. For comparison. For those early-stage questions that shape whether someone even considers your company.

• "What's it like to work at ___?"

• "Best healthcare companies for work-life balance"

• "Companies hiring remote product managers"

And here's what matters: LLMs are forming opinions about your company whether you participate or not. They're scraping Glassdoor. Reddit. LinkedIn. News articles. Job descriptions from aggregators. Your career site (if it's crawlable). And they're synthesizing all of that into a narrative.

If that narrative is messy, contradictory, or missing entirely—you're losing candidates before they ever see your apply button.

According to Semrush, the average Google search query is 3-4 words long.

AI prompts? Around 23 words on average—nearly 7x longer.

That's a completely different intent model.

Google: "software engineer jobs"
ChatGPT: "What are the best companies for mid-career software engineers who value work-life balance, want to work on AI products, and prefer remote-first culture?"

If you're still optimizing for keyword strings, you're solving for the wrong search behavior.

We need to adapt as an industry to how candidates are actually discovering and evaluating opportunities now.

At CollabWORK, we've been distributing jobs through trusted professional communities and newsletters for years: Morning Brew, 6AM City's 400+ local newsletters, security-focused Discord servers, niche Slack communities.

We always knew it worked because candidates told us: "I found this role in my newsletter." Trust-based discovery.

But what we're seeing now is that those same placements are becoming the sources LLMs reference when candidates ask about companies and roles.

When Morning Brew features a company. When Raytheon roles show up in security Discord servers. When a healthcare newsletter highlights opportunities—LLMs take note.

That's not SEO. That's not your career site. That's your job content showing up where attention has already been earned.

Which is why we've started thinking differently about XML feeds.

Most companies treat job feeds like utility infrastructure—something that exists to power aggregators and programmatic buying. Clean it up enough to not break Indeed's import. Make sure titles aren't garbled.

But if LLMs are ingesting job data at scale, and if candidates are asking nuanced questions like "remote nursing jobs in Arizona with weekend flexibility paying above $75/hour," your feed needs to answer those questions clearly.

Structured data. Clean categorization. Actual FAQs embedded in each listing.

Not just for job boards. For AI search.

Most teams don't know where to start because they can't see the problem yet.

So we're approaching this in three layers:

Layer 1: Monitoring
Where do you actually show up when candidates ask AI about your company? Which prompts surface you vs. competitors? What sources are being cited? You can't improve what you can't measure.

Layer 2: Content Optimization
Are you answering real questions? Is your job content structured for 23-word prompts, not 3-word keywords? Can LLMs extract meaning from what you publish?

Layer 3: Distribution
Are you showing up in the trusted sources LLMs cite—not just on your own channels? According to Muck Rack, over 80% of LLM citations still pull from earned media sources. If you're not in those spaces, you're not in the conversation.

This is the flywheel: Monitor where you are. Optimize what you control. Distribute where trust already exists.

Technical crawlability matters too—schema markup, site structure, all of that. But if you don't know where you stand first, you're optimizing blind.

The CEO asking those questions wasn't wrong to be nervous.

Employer brand used to be about controlling the narrative on channels you owned or rented. Career site. LinkedIn page. Glassdoor strategy. Paid campaigns.

Now it's about showing up where candidates are actually forming opinions—often before they even know they're job searching.

And if you're not monitoring that, you can't manage it.

We've been quietly pressure-testing this with a few partners—tracking how employers show up across ChatGPT, Gemini, Perplexity, and Google's AI overviews. Then connecting that to what we already do: optimizing job feeds and distributing them through communities and media that LLMs trust.

Not rankings. Not gaming the system. Just clarity—and action.

If you're curious how your company shows up in AI search—or want to see what we're building—reply to this email. Happy to run a quick diagnostic and share what we're learning.

More next week as we all learn together.

Best,
Summer Delaney
CollabWORK Founder and CEO

• YouTube Is Now the Top Citation Source in AI Search — New Adweek reporting shows YouTube has overtaken Reddit as the most frequently cited social platform in LLM responses—signaling a major shift in how AI systems source authority. A clear reminder that AI visibility now extends far beyond written content.

• Why Traffic Is Down but Pipeline Is Up (and Why That’s Not a Bug) — Kevin Indig breaks down the growing disconnect between traditional SEO traffic and real business outcomes—arguing that AI search didn’t kill demand, it just removed the click. A sharp read on why brand authority and influence now matter more than pageviews.

• The LLM Visibility Playbook — I caught Happydance’s webinar on AI Visibility and thought this was a great resource to reference for quarterly plannings.