Your AI
Copilot
Handbook

Why it worksAsking the AI to "be" Rob grounds it in your company context. Requesting JSON output forces structure instead of prose, making it easy to copy individual fields directly into your ATS or email. Combining extraction + boolean + templates in one call saves ~20 minutes per req.

Why it worksThe explicit bias-free instruction reduces AI tendency to pattern-match on names or schools. Asking for a structured table makes output scannable in seconds. Injecting both JD and Resume via the global panels means you don't have to re-paste for every screen.

Why it worksNegative rules ("no generic opener", "no emojis") are just as powerful as positive ones — they prevent the AI from defaulting to clichéd recruiter language that candidates ignore. The pain-point angle doubles reply rates compared to feature-listing outreach.

Why it worksLinkedIn DMs are a completely different format from InMail — shorter, more casual, read on phones. The 300-character constraint forces the AI to lead with value immediately. Three variants let you A/B test and find what lands best with your target candidate pool.

Why it worksThe title expander is the hidden gem here. A React Native engineer might list themselves as "Mobile Dev", "RN Specialist", or even just "JavaScript Developer" — all invisible to a narrow Boolean. AI knows these synonyms across thousands of job postings in its training data.

Why it worksSharing a rubric with the client before interviews positions [YOUR COMPANY] as a strategic partner, not just a resume sender. The 1–5 scoring grid also makes debriefs faster — interviewers fill in numbers, not essays.

Why it worksAcknowledging one gap proactively builds enormous credibility with hiring managers — it signals you've actually read both the JD and the resume, rather than bulk-submitting. This is the single biggest differentiator between a trusted partner and a resume vendor.

Why it worksCounteroffer fallout is one of the most preventable losses in staffing. The key is referencing the candidate's original motivations for leaving — which you captured in intake — and showing how a salary bump doesn't fix culture, tech debt, or career growth.

Why it worksMost agencies send no feedback at all, or a two-line form rejection. A thoughtful rejection email turns a rejected candidate into a referral source and future placement. The "keep the door open" closing is critical — the market is small and memory is long.

Why it worksMost agencies disappear the moment a placement is made. A 30-60-90 cadence catches early issues before they become falloffs and generates referrals at the exact moment the candidate is most enthusiastic about their new role.

Why it worksThe bottleneck analysis is the part that saves the most time. It forces you to look at the pipeline structurally — not just "who's good" but "where is time being lost." Clients also appreciate receiving a structured status update rather than a scatter of individual emails.

Why it worksAsking for a CSV output means you can paste the result directly into your CRM. The HOT flag cuts through the noise — most sales reps spend equal time on all leads, but 20% of leads drive 80% of revenue.

Why it worksEmail 1 should never pitch — it should earn the right to pitch. The "pain before product" structure is what separates a 30% reply rate from a 3% one. Email 2's insight-first approach positions you as a knowledgeable advisor, not a vendor.

Why it worksLinkedIn DMs get read on phones in 3 seconds. The binary question has dramatically higher response rates because it requires almost no mental effort to answer. The 300-character hard limit forces AI to write like a human, not a brochure.

Why it works"Just checking in" is the most deleted phrase in B2B sales. Re-engagement works when you can answer: "what's different now?" An external trigger (funding round, new job postings) is gold because it's about them, not you.

Why it worksThe acknowledge-pivot-offer structure prevents defensive responses that kill deals. "We're not hiring right now" almost never means that — it means "I don't see why talking to you is worth my time." The pivot reframes their objection as the exact reason to talk now.

Why it worksMost clients think the markup is pure profit. When you transparently walk through employer taxes, compliance costs, and payroll risk, the conversation shifts from "you're expensive" to "oh, you absorb all that." The "risk transfer" framing is particularly powerful with finance-oriented contacts.

Why it worksThe "5 minutes before the call" framing forces the AI to be scannable — not a 10-page essay. The red flags section is the most underused part: identifying deal risks early saves you from chasing unwinnable deals for weeks.

Why it worksThe weighted percentage model forces realism — a Proposal Sent isn't worth full value. The stalled deal flag is often the most actionable output: a deal sitting for 3 weeks without contact needs a Hail Mary move, not another gentle nudge.

Why it worksThe 1-hour follow-up email is the highest-leverage sales action that most people skip. Research consistently shows reply rates drop 40%+ after 24 hours. The retrospective trial-close questions also help you spot the moment where deals typically stall in your pipeline.

Why it worksTelling the AI to "keep all output anonymized" is an extra safeguard. The structured 5-point output maps directly to what your approval system needs.

Why it worksTimesheet vs. invoice mismatches in IT staffing are extremely common. This prompt structures the analysis and tells you exactly what document to request, saving a 3-email back-and-forth with the vendor.

Why it works1099 misclassification is one of the highest-risk legal exposures in IT staffing. This prompt applies all three major US tests simultaneously. It's designed as a first-pass flag, not a replacement for legal counsel.

Why it worksIT staffing firms run 60–90 day payment cycles, making AR collections a monthly reality. The 3-tier escalation preserves the relationship at each step. Email 3's "payment plan" offer often unlocks stuck invoices without going to legal.

Why it worksThe "just below threshold" pattern is one of the most reliable fraud signals in AP — it suggests someone knows the approval limit and is deliberately staying under it. AI is excellent at spotting this across large datasets.

Why it worksThe Confidence column tells you which categorizations to trust immediately vs. which ones to double-check. Low-confidence items are usually edge cases that benefit from human judgment.

Why it worksThe Day -5 to Day +2 sequencing is what makes this genuinely useful vs. a generic checklist. The "staffing-specific mistakes" section catches things like unbilled revenue from contractors who submitted timesheets late — a cash flow leak unique to this industry.

Why it worksThe "contract clauses to add at renewal" output is the sleeper value here. Most vendor risk reviews end at "medium risk — monitor." This prompt turns that into specific legal language to add to your next contract, converting analysis into action.