The most expensive thing in your ATS is invisible

Hi everyone,

Last week I drove to a hospice and home health conference in upstate New York.

I knew almost no one there: I have never worked in hospice or home health. I went because the math of how those industries hire sounded different from anything I had heard in our world, and I wanted to learn.

I walked out with the clearest framing I have heard in a year for a problem every TA leader is sitting on.

Home health, hospice and the agencies that staff them live in a world where every unfilled shift has a price tag. A private-duty home care agency that cannot send an aide to a client loses that hour of billable revenue. A staffing agency that cannot place a nurse loses the shift to a competitor. Even on the Medicare side, when an agency cannot meet referral demand from a hospital discharge planner, that referral goes to the next agency on the list and rarely comes back.

That changes how operators think about their applicant database. The people who applied last year, the silver medalists who almost got hired, the aides who worked a few shifts and went quiet, the nurses who moved out of state and came back — none of them are abstractions. They are unfilled shifts, lost referrals and dollars walking out the door with names attached.

Every operator I met at that conference understood this immediately. They did not need to be sold on the idea that their CRM was an asset. They needed help acting on it.

Every employer with a database is sitting on the same dynamic. The cost is just hidden in a different line item.

For most enterprises, an inactive applicant pool shows up as recruiting spend you did not need to make. Roles that stayed open three weeks longer than they should have. Agency fees on a placement you could have filled internally. Silver medalists who took an offer from a competitor because nobody called them back. Productivity loss that nobody charges to talent acquisition because it does not have a line item.

Home health and home care operators are the canary, not the exception. They feel the cost faster because their unit economics make it visceral. The cost is identical for the hospital across town, the retailer, the staffing firm or the tech company with thirty open AI engineering roles. Most TA leaders just cannot see it in real time the way a home care operator can.

The conversation about your ATS shifts. It stops being a system of record and starts being an audience you paid to acquire and are not talking to.

That is a different category of asset: it changes what you are willing to invest to keep it warm, what you expect from your vendors and what you measure. Open rates, reactivation rates and source of hire from the existing database, these are the kind of metrics a marketing team would build a quarterly review around.

For our Nurse Ascent newsletter, the open rate is 58 percent. The reason is not the AI generation or the matching tech. It’s because it leads with editorial value and treats the audience like people, not pipeline. The same logic applies to any database.

Job alerts get ignored; useful content does not.

What is the dollar value of the database you already paid for and stopped talking to?

Most TA leaders I ask cannot answer this. Not because they are not smart, but because no one has framed it as a question worth answering. It just sits in the ATS, getting bigger every week, costing more every year and invisible on every line of the budget.

The operators I met last week could tell me their version of this number to the dollar. Their P&L forces them to. Yours does not…yet.

This newsletter is me practicing what I am preaching. I built this list the way most TA teams should be thinking about their ATS. I met you at a conference, or on a sales call or through a mutual connection or somewhere in the space. I added you because I thought you would find this useful.

If I sent you a static job alert every week and called it a newsletter, you would stop opening it inside a month. So I treat this group like an audience, not a list. That is the whole bet behind Nurse Ascent and the product we are building for enterprise TA teams now.

If you have run the math on your own database, or you want to argue with the framing, hit reply. The best material here comes from you.

More next week.
Summer Delaney
CollabWORK Founder and CEO

• Tracking Conversations: Measuring Content and Identity Exposure on AI Chatbots — A new study found 17 of 20 popular AI chatbots share user information with third parties, including prompts, chat URLs and in some cases plaintext conversation snippets. As I frequently discuss, candidates are asking ChatGPT, Gemini and Perplexity what it is like to work at your company, whether your benefits are competitive and what the interview process is actually like. Those prompts are now data points. The question is whether your employer brand team can see what is being asked and act on it. (Full study here.)

• Lessons from 25+ conferences — I posted a video this week on what I have actually learned showing up at 25+ events across HR tech, healthcare, media and AI. Two things that keep proving true: the biggest conferences are not always the most valuable, and the best insights for your category often come from conferences outside it. The hospice trip that anchored this issue is the most recent example.

• LinkedIn posts that robots can't resist — Scrunch analyzed 12,000 LinkedIn posts to figure out what ChatGPT actually cites. The headline finding: engagement does not predict citation. Reaction count has near-zero effect. What does work is technical detail, named entities and topic specificity. What hurts is unicode formatting (the fake bold tactic) and link-in-comments. Useful if your team is rethinking how you write on LinkedIn for AI visibility.

• How long does it take to be cited in ChatGPT and Claude? — A Profound early employee dug through billions of logs and roughly 900 newly published marketing pages to track how long it takes for content to get cited by AI engines. Median time from publishing to first citation is under 7 days. 90 percent of pages that get cited get there within 37 days. If your content is older than that and still has not been cited, the issue is on your end, not the engine. Happy to share our learnings and help you out.

CollabWORK is hiring visibility infrastructure. We help employers get found, distributed, and re-engaged across the channels that now define how candidates actually discover jobs: AI search, professional communities and their own talent networks.

Three products, one thesis.

• Employer AI Discoverability monitors how you show up across ChatGPT, Gemini, Perplexity and other LLMs, and optimizes your job descriptions and XML feeds so you can actually be surfaced by the models candidates are searching.

• Community Distribution places roles inside 500+ vetted newsletters and communities including Morning Brew and 6AM City, reaching over 14 million professionals.

• Talent Community Activation turns the dormant ATS and CRM data you've already paid for into a living pipeline.

The companies winning at hiring right now aren't the ones with the biggest job boards budget. They're the ones showing up where great candidates actually are.

Find time to learn more. You can also find us on LinkedIn and X.