Negotiation for HR: BATNA, ZOPA, and the Harvard PON model applied to people work

Most HR negotiation training is communication training in disguise — active listening, empathy, win-win language. Useful, but downstream of the actual mechanics. The Harvard Program on Negotiation provides the structural vocabulary that makes the communication work.

The PON vocabulary

BATNA — the only concept that always matters

Fisher and Ury’s central insight (Getting to Yes, 1981) is that leverage in any negotiation comes from your best alternative, not your demands. The party that can credibly walk away wins disproportionately.

• Know your BATNA in numbers before you enter the room.
• Estimate their BATNA — what happens to them if no deal? Most over-estimate their counterpart’s BATNA (assume they have better options than they do).
• Improve your BATNA whenever possible. A better #2 candidate, a backup vendor, a documented investigation — each shifts the entire negotiation.
• Disclose your BATNA only when it’s strong and credible. A weak BATNA disclosed is a position weakened.

ZOPA and reservation prices

If the seller’s reservation price (lowest acceptable) is $80 and the buyer’s reservation price (highest acceptable) is $100, ZOPA is $80–100. Any number in that range is a deal both prefer to no deal. Your job is to find ZOPA and split it favorably.

If ZOPA does not exist — seller’s minimum exceeds buyer’s maximum — no skill in the room bridges it. Recognize this fast. The most expensive mistake in negotiation is spending weeks trying to close a negative-ZOPA deal.

Principled negotiation, applied

Application: offer negotiation

• Your BATNA: the #2 candidate × probability they accept × salary delta. Calculate this concretely before extending the offer.
• Their BATNA: usually their current job (with all its frictions) or a competing offer. Verify what you can; assume more than you’re told but less than they claim.
• ZOPA: from your max budget down to their walk-away. The candidate’s walk-away is rarely just compensation — it includes title, scope, manager fit, equity story.
• Anchor with the philosophy first, then a specific number. ‘We pay at the 60th percentile of market for this level, which puts you at $X.’ The philosophy makes the number defensible.
• Use multiple-issue trades. Base, sign-on, equity, start date, title, scope, location — never negotiate one issue at a time.

Application: exit negotiation

• Your BATNA: a unilateral termination with statutory severance only, plus litigation risk.
• Their BATNA: litigation × probability of winning × time (1–3 years) × cost ($50–200k of legal fees) × emotional cost × reputational cost.
• ZOPA: from the floor of what you’ll offer for a clean release down to the ceiling of what protects the company from realistic claims.
• Interests over positions: ‘What does a soft landing look like for you?’ often unlocks creative options — transition title, reference, extended health, outplacement — that are cheap to give and high in value received.
• Anchor with the release-and-severance package. Most counter-anchors come in with multipliers; the framing of the original package matters disproportionately.

Application: internal pay pushback

• Your BATNA: the employee stays at current comp, possibly disengaged. Or they leave — cost of replacement is typically 50–200% of salary.
• Their BATNA: a competing offer, an internal move, or staying-and-complaining. The credibility of each varies.
• ZOPA: between current comp and the band ceiling (or the next level’s band, if mid-cycle re-leveling is in play).
• Objective criteria: market data, internal band, calibrated peer comp. ‘We just moved you from the 45th to the 55th percentile of the band; the 75th would put you above 8 of your peers at the same level.’
• When the right answer is no: have a written reason, deliver it the same day, and follow up at the next review cycle. A pushback declined with no follow-up is the highest-correlation predictor of regretted attrition.

Anchoring, framing, and tactics

• Anchor first when you have strong information about ZOPA. Ask first when you don’t — let them anchor and respond from data.
• Make the anchor specific. $87,250 outperforms $87,000 in laboratory studies (Mason et al., 2013) — specificity signals informed analysis.
• Frame as gains, not losses, where possible (Tversky & Kahneman, 1981). ‘You gain $X over 2 years’ outperforms ‘we’re only $X apart per year.’
• Use silence after the anchor. The instinct to fill silence with concessions is the most expensive instinct in negotiation.
• Never split the difference reflexively. ‘Splitting the difference’ is itself a form of anchoring — the average becomes the new ceiling.
• When stuck: ‘What would have to be true for us to get to a yes?’ — opens interest-based exploration without conceding.