Dear Negotiation Explorer,
Welcome to Week 17 of our NegoAI series.
Many of us have seen the same pattern: a genuinely useful assistant is available, the workflow is ready, and yet adoption stalls—not for lack of features, but because people don’t feel fully safe using it when it matters.
In Episode 2 with Prof. Remi Smolinski, we returned to a simple driver that keeps proving itself: training builds confidence, confidence builds trust, and trust drives adoption.
Training → Trust → Adoption
Training That Builds Trust (and Outcomes)
In AI-augmented negotiation, the barrier is rarely technical—it’s confidence under pressure. When teams understand what the assistant does (and doesn’t), where humans stay in control, and how to use it on real work, hesitation fades. Recommendations get followed more often, value stories are defended more clearly, and outcomes improve.
One sentence from last week’s data worth keeping in mind: when both sides prepare with LLMs, joint outcomes improve by 84.4%, with only a 2.2% difference between parties—so value goes up while fairness stays balanced.
A quick stance on the landscape: we prefer assistants over autonomous agents because we want to preserve a clear human-in-the-loop and use AI to augment, not replace, negotiators’ skills and judgment
And a small nudge on adoption: this is change management for negotiators—we’re changing habits, not just installing a tool—so the rollout must be simple, safe, and staged.
Field Notes: when coaching turns a tool into outcomes
A sales team was seeing margin erosion as conversations slipped to price too quickly. We introduced a workflow with two agents built on Cassidy:
One for a crisp value proposition
One for value defense
We paired the workflow with short, focused training and coaching. The tools didn’t change the culture by themselves—practice did. As confidence grew, the value story moved back to the center and margins were protected more consistently.
“Give people confidence they are using a tool—not the tool using them.”
The Training → Adoption → Coaching Playbook
1) Demystify kickoff — 90 minutes
Turn the black box into a glass box: what the assistant does/doesn’t do, where humans decide, how to use it safely on real deals.
2) Guided reps — 90 minutes
Short, realistic drills with good/better/best exemplars tied to current opportunities.
3) Adoption sprint — 2 weeks
Small teams apply the assistant on live opportunities. Aim for two “first wins.” Daily micro-check:
Did we use it? What helped? What slowed us down? One improvement to try tomorrow.
4) Follow-up coaching (small squads) — 60 minutes
Teams return with obstacles from the sprint. We remove blockers, update instructions/playbooks, assign owners, and agree the next small step
Principles that help adoption stick
Make it safe. Clear guardrails and a short human-review step before anything leaves the building.
Make it easy. One-click templates, pinned links, a simple starter checklist.
Make progress visible. Share quick wins and the small rules you’re adding, so everyone sees the system learning with them.
Keep humans in charge. The assistant proposes; negotiators decide.
Data & privacy guardrails. Keep client data out of external training, document AI-assisted work with light audit trails, and write these terms down early to reduce process anxiety.
Why this matters
For Procurement: Reduce supplier risk, accelerate cycle times, standardize option evaluation beyond price, and improve internal stakeholder alignment.
For Sales: Strengthen value defense, reduce reactive discounting, and tighten the fit between customer needs and your value proposition.
This Week’s Exercise
TPick one common scenario. First prepare it with no assistant. Then prepare it again using the assistant.
No role play needed—simply compare after the two preparations:
Preparation time (note actual minutes saved)
Confidence (rate 1–5)
Quality of options
Ability to defend value
Finally, capture one rule to add to your team playbook.
Equipping negotiators with the right skills, building their confidence, and turning AI from a black box into a glass box—with clear guardrails and simple privacy practices—builds the trust to use it under pressure.
Preview of Next week
Next week, we will continue with the third episode of my conversation with Professor Remi Smolinski.