Pharma Trade Bootcamp · SOP · Scripts · Cases

A controlled field experiment covering 336 salespeople over several months (one group with an AI assistant, one without) found roughly 12% more orders in the AI group — statistically robust. Field evidence, not a demo.

The gain does not come from working people harder, but from copying the top performer's ineffable judgement — negotiation feel, intent reading, need spotting — to everyone else. AI learned from top behaviour guides ordinary decisions: the master in your pocket. Effects: faster communication, fewer expert hours, sharper intent judgement.

Experts gain little beyond time saved; mid-level staff gain most — they can absorb AI-transferred experience and still have room to grow; the weakest benefit but struggle to internalise it. Design training and tooling for the median: mid-level staff get AI plus training; weaker staff get AI plus mentoring.

Anyone can rent the same models; private data is the moat — product libraries, CAS and quality dossiers, deal records, correspondence: high-value behavioural data. Garbage in, garbage out. For pharma traders, every quote, every email and every audit conclusion is an asset to be banked.

Within one task AI is strong at some steps and weak at others — a jagged frontier. Decompose work into define → generate → verify → execute → accept: delegate generation boldly, keep verify and accept with humans, and route compliance red lines (sanctions, dual-use, end-use) to humans, always.

Three judgements: the CEO's understanding of AI sets the possible; the firm's data maturity is the precondition; private data quality is the ceiling. Three enablers: data quality, scenario-matched models, and re-cut roles and workflows around AI. Let 'ordinary rep plus AI' take on judgements once reserved for experts — released gradually, not overnight.

Across six scenarios — development, inquiry, quote, negotiation, delivery and red lines — recite the standard scripts and explain the reasoning.

From inquiry to post-mortem, know every step: what to do, who does it, what to log. Process over inspiration.

Sanctions, dual-use, end-use and confidentiality: know how to check, how to speak, and when to escalate.

By the end of mentoring, run an inquiry-to-delivery chain solo and submit your first reviewed real quote.

Self-study plus one 60-minute Q&A. Output: a confirmed personal learning path.

Read aloud, transcribe, then rewrite each scenario in your own words. Output: a personal script card — three sentences per scenario.

In pairs, one plays client or supplier while the other responds; the supervisor scores with the rubric and replays the tape. Output: scorecard plus an improvement list.

Run one real chain — inquiry, quote, negotiation, delivery — under a mentor, logging everything. Output: two work decision cards (next step / owner / metric / review date).

Pass all four gates to work solo; failures follow the remedial path — and the red-line exam is non-negotiable.

Who you are and what you supply to whom — no corporate chronicle. In three seconds the reader knows you are a pharma API/excipient supplier.

One verifiable proof: pharmacopoeia compliance (USP/EP/CP), GMP audit history, registration support (DMF/CEP), delivery track record by type. Let the proof speak.

Ask for a 15-minute call or their specification and pharmacopoeia standard — not an order. The cheaper the reply, the higher the response rate.

Spell out spec, quantity, COA, shipping terms and an end-use declaration. Free samples go to clients with registration intent; pure price-shoppers pay shipping or sample fee.

Tier 1 (public): spec sheet, sample COA. Tier 2 (under CDA): process overview, stability data, registration file list. Tier 3 (after project award): DMF/CEP dossiers and audit support. Release tier by tier — never show all cards at once.

Day 3 confirm receipt, day 10 ask testing progress, day 20 propose a registration timeline. Each follow-up carries new information — never 'any update?'

Tell the client the day you know. A delay they discover themselves doubles the trust damage. Proactive notice + a firm new timeline + compensation options.

“I need to inform you early: production of lot X is delayed 10 days due to [reason]. Impact on your shipment: new ETD [date]. Options: partial shipment of 60% now, or full shipment with a 3% credit on the next order. Which works better for your planning?”

“I understand the impact this caused, and I take it seriously. Let us agree the facts first — I will send the batch records and shipping documents today, and propose a written resolution within 48 hours.” No blame-shifting, no blanket admission, no off-book settlement — follow the written process.

Bring every quote to one frame: same spec, monograph, Incoterms, payment terms, freight and tax treatment. Comparison on mixed frames is fake comparison.

Politely ask for the cost build-up (materials / labour / energy / loss / packaging / overhead) to locate their room to move. Negotiation needs an anchor, and their cost is it.

Round one on price, round two on terms (volume / credit / lead time / annual commitment), round three lock the deal. Each round trades a concession for a condition — one-way squeezing pushes suppliers toward cutting corners.

Sign a quality agreement before the first order: spec and test methods, deviation and change control (any process/material change requires written notice and assessment), retention and traceability, complaint response times. Change control is the lifeline of pharma procurement.

“Before we place the first order, we would like to schedule a GMP audit — it protects both sides. Alternatively, send your latest audit report and we can start with a documentation review.”

Critical findings (quality or data integrity): stop orders until remediation is verified. Major: deadline plus report. Minor: tracked observation. State the grading scheme upfront and put remediation in writing.

On receipt: retain samples, test data, photos and batch numbers. Without fixed evidence, everything after becomes a shouting match.

Script: “Batch [X] failed assay at [Y]% against spec [Z]. Findings attached. Per our quality agreement, we request root-cause analysis and CAPA within 5 working days.” State batch, metric, standard, ask and deadline.

Choose per agreement: return and replace / concession (discount plus restricted use) / on-site re-test. Any concession needs written sign-off and an impact assessment on downstream registration.

First confirm what broke: capacity, logistics or compliance qualification? Then size the gap (days covered by open orders), activate the backup source or alternate spec, and only then communicate outward — facts first.

“We are qualifying a parallel source to protect supply continuity — this is standard risk management, not a vote of no confidence. Existing allocation and terms remain unchanged.” To the incumbent: no bad-mouthing, no abrupt replacement — keep the way back open.

Mark every key material: single / dual / alternate-spec. Re-assess single-source risk quarterly (supplier finances, geographic concentration, compliance qualification validity) and qualify a second source in parallel. Before a break, dual sourcing costs management time; after one, it saves the business.

Before touching price, reply with the four questions: spec, monograph, annual volume, end-use. Answers come back: EP standard, ~60 t/year, for finished-dose export — two levels of information richer than the first email.

Run the cost waterfall: beyond direct cost, 90-day credit funding, FX and freight leave a thin margin. With the floor computed, negotiation stops running on feeling.

Probe the cause (budget line or fishing?) → change dimension (price firm, terms tradable) → leave a ladder (two versions to choose from).

Confirm by phone and in writing: cause, restart date, goods in transit, whether we get priority. Facts come by official letter, not chat screenshots.

Build the coverage table: stock plus in-transit = 20 days, gap 25. Allocate by client priority: registration validation first, volume orders negotiated into batches.

Activate the second source qualified in parallel half a year earlier (thank heavens): rush testing, a spot contract, first air-freight lot in five days — covering half the gap; the rest is negotiated into phased deliveries with clients.

To downstream clients: coverage first (your volume is safe to date X) → actions (second source plus phased plan) → fixed Monday updates. To the broken supplier: reserve rights in writing, without burning bridges.

Check parties: screening raises a related-party alert. Check items: the product sits at the edge of dual-use control — its regulatory status depends on use. Check use: implausible end-use, unknown end-user. All three in doubt.

The AI assistant only pre-screened and logged (screening record, full correspondence archived); the judgement went to the compliance lead: no pricing on sensitive items, no commitment of any kind, until an end-user and end-use statement clears internal review.

After two written rounds the party still refused an end-use statement, so we declined under compliance policy — every word in writing and archived. The deal died; nothing we said could be held against us.

Finish units 01-02; transcribe your role's scripts and read them aloud three times; sit in on two real mentor calls.

Rewrite every scenario script into your own voice (personal script card); use them at least three times in real work — three structured replies or one RFQ.

Complete the role-play (sparring plus rubric); pass the red-line exam; score 80+ on every unit quiz.

Run one real chain solo (inquiry → quote → negotiation, or sourcing → RFQ → negotiation); produce two decision cards; hold one review with your supervisor.