HRBP Metrics: Engagement, Attrition, Internal Mobility, Manager NPS

Picture the QBR. The CRO is presenting Q3 sales numbers. Pipeline is fine, but quota attainment is sliding because the West Coast AE team just lost four reps in 60 days, three of them top quartile. The CRO turns to you. "Did People see this coming?"

You did, in fact, run an engagement survey eight weeks ago. eNPS came back at 34. You put a green smiley face on the People slide. And now you're sitting in a room with the CRO, the CFO, and a CEO who is doing math in his head about whether ramp time means he calls the board to lower Q4 guidance.

This is the moment most HRBP careers stall. Not because we don't care about the people. Because we walked into a room full of finance-trained operators with a metric that measures average sentiment and called it a leading indicator. It isn't. It never was.

The fix isn't more dashboards. It's six metrics that hold up under cross-examination, plus an honest conversation about which ones lead and which ones lag. Let me walk you through what I actually put on the QBR slide, why each one earns its spot, and which numbers you should be reading instead of eNPS when you want to know what's going to break next quarter.

eNPS: Keep It, but Stop Pretending It Predicts Anything

eNPS measures the average emotional temperature of people who chose to fill out a survey. That's it. It's a sentiment snapshot, not a leading indicator. The score moves when something feels different (a new perk, a layoff rumor, a comp cycle that went well or badly), and it moves with a lag of weeks behind whatever caused it.

Why keep it then? Because the trendline matters even when the absolute number doesn't, and because executives understand it. A 12-point drop in eNPS quarter-over-quarter is a real signal. An eNPS of 34 versus 36 is noise. Use it for direction, not altitude.

The B2B SaaS benchmark I work to: 30+ for high-trust orgs, 20-30 for healthy, under 20 means you have work to do. But here's the trap. eNPS goes UP after a layoff, every time, because the people who survived feel grateful and the people who would have rated low aren't there to take the survey. Survivor bias bakes a 5-10 point lift into your number for two quarters. Annotate your eNPS chart with headcount events or you will mislead your own executive team.

The green-eNPS trap: a 34 eNPS on a survey with 62% participation, where the 38% who didn't respond skew toward your highest-tenure ICs and the team that just had three exits. That number is meaningless. Always report eNPS alongside participation rate AND segment cuts by tenure band, manager, and team. If sales eNPS is 12 and engineering eNPS is 48, your overall 34 is hiding the fire.

Regrettable Attrition Rate: The Only Attrition Number That Matters

Total attrition is a vanity number. It includes people you wanted to lose and people you fired for cause. Regrettable attrition is the population that actually hurts: voluntary departures of people you wanted to keep.

How I define regrettable, and the checklist I use so it doesn't become a vibe check:

1. Voluntary: they resigned, you didn't terminate
2. Performance was meets expectations or above at last review (not last week's mood, the last formal calibration)
3. The role is still budgeted: you're backfilling, not absorbing
4. The manager would re-hire them tomorrow: answered yes/no in a 5-minute exit-classification call, not a free-text exit interview

That fourth one matters. The manager-rehire-yes question filters out the "we would have managed them out anyway" rationalization that creeps in three weeks after someone quits. Lock the classification within five business days of the resignation, before the team rewrites the story.

B2B SaaS healthy range: 8-12% annualized regrettable attrition. Above 15% is red. Above 20% in any single function is a fire. Go look at managers and comp before you look at anything else. Under 5% sounds great but often means you've stopped hiring or your bar moved down.

Cut it three ways every time you report it: by tenure band (0-12 months, 13-24, 25+), by manager, and by department. Never by role alone. Role-level attrition averages hide the fact that one manager is responsible for 70% of the SDR exits.

Manager NPS: The Upward Signal That Catches Problems Two Quarters Early

Manager NPS is the single most predictive HRBP metric I track, and almost nobody reports it correctly. It's the NPS-style question asked of direct reports about their manager: "On a scale of 0-10, how likely are you to recommend working for this manager to a peer?"

The math: % promoters (9-10) minus % detractors (0-6). Same as eNPS. Run it quarterly. Report it per manager, never as an org-wide average — averages hide outliers, and outliers are the entire point.

The first time I caught a manager problem in manager NPS before regrettable attrition was the autumn we acquired a 40-person services team and rolled them under a director who'd been a strong IC and a mediocre people manager. Q1 manager NPS for that director: -8. eNPS for the team: 28, totally normal. Q2 we ran a coaching plan. Q3 he resigned, two of his reports resigned the same month, and one of them was the highest performer in the function.

Had we only watched eNPS or attrition, we'd have caught it nine months later, after the damage. Manager NPS in Q1 led regrettable attrition in Q3 by two quarters. That lag is the predictive engine of the whole dashboard.

The benchmarks I use:

• Manager NPS 20+: healthy, leave them alone
• 0 to 19: monitor, ask the team specific questions, watch for repeat dips
• Below 0: 90-day fire-or-coach trigger. You and the manager's manager build a coaching plan or you start the conversation about whether they should be a manager at all

The trick to making manager NPS land with executives: pair every score with the trailing 12-month regrettable-attrition rate from that manager's team. Two numbers, one slide. A manager at -5 with 18% attrition is a different conversation than one at -5 with 4% attrition (the second is probably a brand-new manager whose team is still adjusting).

Internal Mobility Rate: The Leading Indicator for A-Player Retention

Internal mobility = (promotions + lateral moves) / average headcount, annualized. It tells you whether your A-players see a future inside the company or are quietly opening Carta and updating LinkedIn.

B2B SaaS healthy benchmark: 15%+ for orgs with healthy career paths. Below 10% means people leave to get promoted elsewhere. Above 25% can mean either a wonderful career-development engine or a chaotic re-org culture, so check the texture of the moves before you celebrate.

Lateral moves count, but only when they're real. A lateral move with a new comp band, new manager, and a new scope counts. A title change with no new responsibility doesn't. Don't let the number get gamed by HRIS hygiene fixes.

The pattern I watch for: a team where eNPS is up but mobility is zero. This is the mobility-less morale bump, and it's a future flight risk masquerading as a healthy team. People feel good RIGHT NOW because last quarter was calm, but they haven't seen anyone get promoted in 18 months. The good ones are 4-6 months from quitting and they don't know it yet. You do, if you're reading both metrics together.

Pair internal mobility with regrettable attrition by tenure band. If 25+ month tenured employees have 14% regrettable attrition and 6% internal mobility, your career path is broken. Fix that before you fix anything else.

Time-to-Productivity: The Onboarding ROI Number That Ties HR to Revenue

Time-to-productivity is how long from start date until a new hire performs at 80% of a tenured peer's output. For an SDR it's meeting their full quota. For an engineer it's shipping at the team's average velocity. For a CSM it's owning their full book at standard NRR. Your operating number, not HR's.

B2B SaaS targets:

• Individual contributor: 90 days for sales/CS, 90-120 days for engineering, 60-90 days for go-to-market ops
• Manager: 120-150 days
• Director and above: 180 days

Where this metric earns its slot on the QBR slide: it's the only People metric that translates directly to revenue. If sales time-to-productivity slips from 90 to 120 days, the CFO can model the cost: 30 extra days of unproductive ramp times the loaded cost of every new AE you hired this quarter. That's a number she can present to the CEO. eNPS is not.

The diagnostic question I ask when time-to-productivity is slipping: is it the candidate quality, the onboarding program, or the manager? Pull the cohort of new hires whose time-to-productivity is over benchmark, group them by hiring manager, and look. Ninety percent of the time the answer is the manager, not the program. That's when you walk over to People Ops with a coaching trigger and a comp-adjustment recommendation.

Comp Ratio Compression: The Silent Regrettable-Attrition Driver

Comp ratio = an employee's salary divided by the midpoint of their pay band. Comp ratio compression is what happens when new hires come in at 1.05 of band and tenured employees in the same role are sitting at 0.92 because they got 3% merit increases for three years while the market moved 8%.

This is the comp-compression silent killer. It shows up as regrettable attrition with exit-interview answers like "great experience, just a better offer." Of course the offer was better. They were 13 percentage points underwater versus internal new hires doing the same job, and you didn't catch it because comp data lives in Workday and exit interviews live in your inbox.

The metric: pull every role with three or more incumbents. Calculate the average comp ratio for tenured employees (24+ months) versus new hires (under 12 months). If the gap is more than 5 percentage points, you have compression. If the gap is more than 10, you are losing tenured A-players right now and you don't know it yet.

Run this calculation every quarter, before merit cycle. Bring the list of compressed roles to the CFO with a name-by-name re-leveling proposal. The cost of fixing comp for 12 tenured ICs is roughly the cost of replacing one of them. The math is not hard. Most CFOs say yes when you put it that way.

Reading Attrition Signals: Cohort + Tenure + Manager, Never Just Role

The biggest mistake HRBPs make in attrition analysis is reporting it by role. "Sales attrition is 18%" is a number that triggers no useful action. Cohort plus tenure plus manager is what surfaces what's actually happening.

The 18-month cliff: in B2B SaaS, voluntary attrition spikes hard between months 15-21 of tenure. New hires figure out within 18 months whether their next promotion is real or a polite stalling exercise. If you report attrition by tenure band, you can see the cliff forming and intervene with promotion-path conversations before the resignation letter.

The "one bad manager" pattern: three or more regrettable exits in a 12-month rolling window from one manager's team is a coaching trigger, full stop. Not a "let's see if it continues," but a coaching plan within 30 days, calibrated against manager NPS and skip-level feedback.

Why exit-interview themes alone are useless: recall bias and exit politeness. People departing rationalize their decision into a clean narrative ("I needed a new challenge"), and they soften criticism because they want a reference. Run exit interviews, but treat them as one input out of five, not the source of truth. The five inputs: exit interview themes, manager NPS trajectory for the departing employee's manager, comp ratio at exit, internal-mobility events for that employee in the prior 24 months, and the manager's classification of the exit (regrettable yes/no).

The QBR Slide That Holds Up to the CEO

One slide. Six metrics. Trailing four quarters. Each metric has a "so what" annotation in plain English. Each "so what" answers an unasked CEO question.

Here's the structure I use:

The annotations are the slide. The numbers are evidence. If you can't write a "so what" in 12 words, the metric doesn't belong on the slide.

The predictive lag I want every HRBP to internalize: manager NPS in Q1 leads regrettable attrition by two quarters. The number you watch in Q1 is the number that resolves itself in Q3, either through coaching working or through people quitting. By the time regrettable attrition spikes, the action you needed to take was 180 days ago. Manager NPS is your early-warning system. Treat it that way.

Vanity Metrics to Cut from Your Dashboard

• Survey participation rate as a success metric. Participation is a hygiene number, not an outcome. Optimize for honest signal, not response rate.
• Anonymous "manager effectiveness" averages. Org-wide rollups hide the outliers that matter. Always report per-manager with a minimum sample threshold.
• Total attrition without the regrettable cut. Mixing voluntary, involuntary, and excellent-departures into one number tells you nothing actionable.
• Diversity ratios without flow data. A static representation number doesn't tell you whether you're hiring, promoting, or retaining underrepresented talent. Hiring funnel + promotion rate + attrition rate, segmented, is what matters.
• Engagement-survey question scores in isolation. "82% of employees feel valued" sounds great until you cut it by manager and see four teams under 50%.

What an HRBP Owns vs. What People Ops Owns

The most common career trap for HRBPs is getting absorbed into running the survey tool. That's not your job. People Ops collects and reports. The HRBP interprets and acts.

Your job on these metrics:

• Read the trends and segments
• Identify the manager, team, or cohort issue
• Build the intervention (coaching, comp adjustment, promotion conversation, re-org recommendation)
• Present the QBR slide with the so-what annotations and the proposed actions
• Track whether your intervention moved the metric

If you're spending more than 10% of your week formatting survey reports, you've drifted out of HRBP work and into HR Ops. Hand the tool back. Take the analysis.

The CEO doesn't care about your eNPS. The CEO cares which teams will miss plan because of a manager problem, and whether you saw it coming. The six metrics in this playbook, read together with the predictive lag baked in, are how you walk into the next QBR with that answer ready before the question gets asked.

Learn More

• HRBP Operating Cadence: Weekly, Monthly, Quarterly
• Calibration Sessions That Don't Devolve
• Manager Coaching Plans for Underperforming Leaders
• HR Business Partner JD Template