How do I know if my recruiters are actually using the AI we bought?

Look at usage data by user, not by team. Most tools show you logins, queries run, and features touched per person. If 3 recruiters drive 80 percent of the activity and everyone else touched it twice in month one, you have a non-adoption problem hiding under a decent-looking average. Pull that report this week. It tells you the truth faster than any survey.

Should I make AI use mandatory to force adoption?

No. Mandates without manager buy-in create malicious compliance. Recruiters will open the tool, run one query to hit the metric, then close it and work the way they always have. You get the appearance of adoption and none of the productivity. Train the managers to coach the tool into daily workflow first. Adoption follows the manager, not the mandate.

What if my best biller refuses to use the new AI at all?

Your top biller is your biggest risk and your biggest opportunity. Do not fight them in front of the team. Sit down one-on-one and ask what part of their day the tool could give back. Tie it to their numbers, not the company rollout. When a top biller says the tool saved them 5 hours a week, the rest of the floor moves. When they stay quiet, so does everyone else.

Why Recruiters Ignore AI (And the Manager-First Fix)

Your recruiters are logging into the AI tool once, then going back to their old workflow. The problem is not the software. It is that you trained the recruiters and skipped the managers. Here is how to fix the sequence before non-adoption turns into turnover.

Your recruiters are not using the AI you paid for. They logged in once during the kickoff, ran a search or two, then quietly went back to the workflow they trusted. The tool still shows up on your invoice. It does not show up in their day.

This is the most common story I see in staffing right now. A firm spends six figures on AI tooling, runs a launch, and six months later the usage report looks like a ghost town. Leadership assumes the software was oversold. It usually was not. The rollout skipped a step, and that step was the managers.

The data says this is a training problem, not a tech problem

Bullhorn's Engage Boston 2026 numbers point the same direction as SHRM's research on workplace change. Adoption does not stall because the AI is bad. It stalls because the people who supervise the daily work were never brought in first.

SHRM has been clear for years that employees take behavior cues from their direct manager, not from a company-wide announcement. If a recruiter's manager never uses the tool, never asks about it in a one-on-one, and never ties it to a number, the recruiter reads that silence correctly. The tool is optional. So they opt out.

Here is what actually happens in most staffing rollouts. The vendor trains the end users. The recruiters sit through a 90-minute demo. Everyone nods. Then they walk back to a desk where their manager is still asking about submittals and calls the same way they did last quarter. Nothing about the manager's questions changed. So nothing about the recruiter's behavior changes either.

You trained the wrong people first. That is the whole problem in one sentence.

Why non-adoption turns into turnover if you wait

Non-adoption is not a neutral state. It gets worse. When a recruiter decides the tool is not for them, that decision hardens fast. Three months in, it is a habit. Six months in, it is an identity. They are now the person who does not use AI, and they will defend that position because changing it means admitting they wasted six months.

Meanwhile your top billers are watching. If your best recruiter treats the AI as a threat or a toy, the newer recruiters follow. Top billers set the floor's real culture, not your leadership deck. When a top biller opts out loudly, adoption dies quietly across the whole team.

Then it gets expensive. The recruiters who lean into AI at other firms start pulling ahead on placement speed. Your people feel the gap. Some blame the firm. Some leave. You end up losing the exact people you needed to keep, and the AI you bought to help them gets blamed for a problem it never caused.

The manager-first change sequence that unsticks adoption

Build. Change. Adopt. is the framework I use, and the Change step is where most firms trip. Change is not the launch event. Change is the sequence of behaviors that has to shift before adoption is even possible. It starts with managers.

Here is the order that works.

Step 1: Train managers two weeks before the recruiters

Give your managers a head start. Two weeks minimum. They need to use the tool on real reqs, hit the same friction the recruiters will hit, and figure out where it actually saves time. A manager who has not used the tool cannot coach it. Full stop.

Step 2: Rewrite the questions managers ask in one-on-ones

This is the lever almost everyone misses. If your one-on-one script does not change, behavior does not change. Add specific questions like these:

Where did the AI save you time this week?
What did it get wrong, and how did you catch it?
Which candidate did you find faster because of it?

When a recruiter knows their manager will ask these questions every week, they use the tool so they have an answer. That is not pressure. That is a manager showing the work matters.

Step 3: Convert one top biller before the full launch

Pick your most respected biller. Not the loudest. The one people copy. Sit with them one-on-one before the rollout and find one task the AI takes off their plate. Tie it to their numbers, never to the company initiative. When that biller stands up in a team meeting and says the tool gave them back four hours a week, you have done more for adoption than any vendor training ever will.

Step 4: Measure adoption by person, not by team

Pull usage by individual user every week for the first 60 days. Team averages lie. They hide the fact that three people carry all the activity while everyone else drifts. When you see an individual go dark, that is a coaching conversation for their manager, not a company-wide email.

What this looks like when it works

A firm that runs this sequence sees adoption curves that hold. Managers coach the tool in every one-on-one. Top billers talk about time saved in team meetings. Newer recruiters copy the top billers because that is what they always do. The usage report by person shows the whole team climbing, not three heroes and a graveyard.

The tool did not change between the firm that failed and the firm that worked. The sequence did. Managers first. Then a converted top biller. Then the floor.

If you rolled out AI in the last year and the usage looks thin, you did not buy the wrong tool. You ran the wrong sequence. The good news is you can rerun it. Start with the managers you skipped the first time.

Do this one thing this week

Pull your AI usage report broken down by individual recruiter, not by team. Find your top biller on that list. If they are not in the top three for usage, book a 20-minute one-on-one with them this week and ask one question: what part of your day would you hand off if you trusted the tool to do it? Their answer is your entire adoption plan. Build the rollout around that.

Why Your Recruiters Are Quietly Not Using the AI You Bought