system-prompts-and-models-o.../dealix/docs/ops/customer_success_playbook.md
2026-05-01 14:03:52 +03:00

6.5 KiB
Raw Blame History

Dealix — Customer Success Playbook

For first 10 customers: manual-first delivery. After that, convert manual steps into product. Paired with FIRST_CUSTOMER_DELIVERY_TEMPLATE.md (tactical) and first_customer_delivery_template.md (intake form).


Principles

  1. Delight beats feature count. First customer's referral is worth more than 10 features.
  2. Manual acceleration > broken automation. If an automation isn't solid, do it by hand.
  3. Every ticket is product research. Log objection/question/pain → tomorrow's prompt upgrade.
  4. One reply thread = one success metric. Never context-switch mid-thread.
  5. Arabic default, English on request. Even with English-only customers, offer bilingual.

Support modes (what Dealix actually helps with)

Mode What Dealix does Customer expectation
Product question Answer via FAQ + route complex to Sami within 2h Answered accurately within hours
Pricing question Route to demo OR quote within 4h Clear quote same day
Pilot explanation 1 SAR × 7 days, unlimited refund — explain upfront Full clarity, no surprise
Privacy objection Send PDPL compliance sheet + answer data-residency Sheet within 1h
Arabic quality doubt Send 3 live sample threads + offer 20-min Zoom demo Samples within 30min
Integration limit Confirm HubSpot/Salesforce/Zoho/Bitrix/webhook Same-day answer
Route to demo Calendly → 20 min → Sami personal Zero friction
Route to payment Manual SOP (bank/STC Pay/invoice) Invoice within 15 min of verbal yes
Route to onboarding kickoff call within 4 hours of payment Onboarded Day 1

First customer onboarding (step-by-step)

Day 0 — Sale closed

  • Payment received (manual or Moyasar)
  • Confirmation email sent within 30 min
  • Calendly kickoff call booked for Day 1

Day 1 — Kickoff call (45 min)

Agenda (hard time-boxed):

  • 0-5 min: Welcome, confirm scope of pilot
  • 5-20 min: Intake form (see first_customer_delivery_template.md)
  • 20-35 min: Walk through how Dealix will work for their specific business
  • 35-40 min: Set success metric (e.g., "respond to 20 leads within 45 sec, 5 demos booked")
  • 40-45 min: Daily check-in schedule + next steps

Day 2-6 — Manual delivery

Sami does this manually for first 10 customers:

  • Every incoming lead → Sami reads → crafts Arabic reply (use Dealix prompt as draft) → sends → logs
  • Daily 10-minute check-in message to customer: "اليوم: رددنا على X leads، حجزنا Y demos"
  • Capture:
    • Which objections appeared?
    • Which messages converted best?
    • Did the customer express delight or frustration?

Day 7 — Pilot review

  • 30-min review call
  • Share metrics + learnings
  • Upgrade path: Pilot → Starter (999 SAR/mo)
  • Ask for testimonial (even if they don't upgrade)
  • Ask for referral

Day 8+ — If they upgrade

  • Switch to Starter billing
  • Weekly check-ins for first month
  • Monthly optimization review
  • Move to automation as trust builds

Success metrics per customer

Track in docs/ops/customer_success_tracker.csv (create on first customer):

| customer_id | start_date | pilot_end | leads_handled | demos_booked | replies_arabic | avg_response_time | nps_score | churn_risk | upgrade_status |


Red flags → escalate to Sami within 1h

  • Customer threatens to churn
  • Customer says "your Arabic is wrong" (go immediately to Zoom call)
  • Customer asks for feature not built (don't lie — offer roadmap ETA or alternative)
  • Customer misses 3 consecutive daily check-ins
  • Customer says "this is not what I expected"
  • Integration breaks or Dealix misses a lead (apologize + manual recovery within 1 hour)

Conversion from pilot → Starter

Day 5 soft ask:

"أيام قليلة باقية في pilot. بناءً على النتائج حتى الآن، هل تتوقع الاستمرار؟"

Day 7 firm ask:

"الـ pilot انتهى اليوم. النتائج كانت [X leads, Y demos, Z response time]. تبغى تكمل Starter (999 ريال/شهر) من الاثنين؟"

If yes → continue. If no → ask: "ما اللي منع الاستمرار؟" → log → make them feel great about saying no (it's valuable data).


Escalation path

  1. Customer ticket / question → Dealix answers via template or Sami within 4h
  2. Objection → Use reply_playbooks_ar.md → resolve OR escalate
  3. Complaint / churn signal → Sami personal call within 1 business day
  4. Technical issue / integration failure → manual recovery + root-cause within 48h
  5. Legal / compliance question → Sami + legal consult (if retained) within 3 business days

Manual-to-automation playbook

After 10 paying customers:

Manual today Auto in v2 Auto in v3
Sami crafts every Arabic reply LLM drafts, human approves Fully auto for "confirmed" message types
Manual daily check-in message Templated daily digest Auto-fire on low-activity day
Manual invoice generation Moyasar auto-invoice on verbal yes Zero-touch checkout flow
Sami reads every objection Classified by AI + logged Router handles 80%+ auto
Onboarding kickoff call Same (calls stay human) Same
Success review call Same Same

Principle: automate low-risk high-volume, keep human on high-context high-stakes.


What Dealix will NOT do

  • Do not make customer commitments beyond written scope
  • Do not answer legal or financial advice questions (route to Sami + disclaimer)
  • Do not send anything outside pilot scope without customer request
  • Do not share one customer's data with another customer
  • Do not offer refund beyond written policy
  • Do not modify Dealix product prompts for one customer without written request
  • Do not auto-send without approval during first 30 days per customer

Weekly customer success review (Fridays, 17:00)

  • Review all active customers (dashboard)
  • Identify top 1 at-risk → proactive outreach before Monday
  • Identify top 1 delighted → ask for testimonial + referral
  • Log top 3 recurring objections → update objection_library_ar.md
  • Log top 3 positive threads → capture message templates → update today_send_queue.md

The sovereign asset isn't code. It's the log of every conversation. After 10 customers, the Arabic-specific message library + objection responses + timing patterns become the moat.