feat: add campaign salvage report from parallel agent analysis

Analyzed all 10 historical loyalty campaigns using parallel agents.
Key finding: every campaign was purely transactional — none created
emotional connection or identity for members.

analysis/campaign-salvage-report.md

@@ -0,0 +1,78 @@
+# Campaign Salvage Report
+*Basecamp Coffee Loyalty Program — Analysis of 10 Historical Campaigns*
+
+---
+
+## The Pattern That Runs Through Everything
+
+Every single campaign — whether it ran, flopped, or never launched — shared one fatal flaw:
+
+**They were all about getting more transactions. None of them were about making people feel something.**
+
+Points multipliers, bonus rewards, streak counters, download incentives. Pure math. No narrative, no identity, no reason for a customer to feel differently about Basecamp Coffee after participating.
+
+---
+
+## Campaign-by-Campaign Findings
+
+### Campaign 01 — Double Points Weekend
+- **Result:** 23% spike during promo, back to baseline within 3 days, zero 30-day retention impact
+- **Core problem:** Rewarded existing behavior, changed nothing. ~$2,400 in points liability for rented engagement.
+- **Worth salvaging?** The urgency mechanic has latent value — only if paired with an identity hook.
+
+### Campaign 02 — Happy Hour Points Boost
+- **Result:** 8% traffic lift that evaporated. Push notifications generated complaints. 34% dismissal rate.
+- **Core problem:** Treated a mindshare problem as a rewards problem.
+- **Worth salvaging?** Time-targeting concept is valid. The intrusive execution is not.
+
+### Campaign 03 — Birthday Bonus
+- **Result:** 73% of redeemers were already regular customers. Did not reactivate lapsed members.
+- **Core problem:** Transactional thinking applied to a relational moment.
+- **Worth salvaging?** Birthday timing is a legitimate trigger. A free drink is forgettable; something personal could work.
+
+### Campaign 04 — Refer a Friend (Never launched)
+- **Core problem:** "Sign up and get points" is a weak motivator. No identity hook for the referrer.
+- **Worth salvaging?** Yes — but only paired with a shareable identity artifact (e.g., a personality quiz result).
+
+### Campaign 05 — Early Bird Bonus (Never ran)
+- **Core problem:** Would have subsidized behavior already happening. Morning regulars didn't need incentive.
+- **Worth salvaging?** Time-window mechanic could work aimed at *inactive* members instead.
+
+### Campaign 06 — Seasonal Drink Launch Points
+- **Result:** 31% trial lift, but only 12% returned. The seasonal drink's novelty did the work — points were redundant.
+- **Worth salvaging?** The real insight: personalized drink discovery based on known preferences.
+
+### Campaign 07 — App Download Bonus (Always-on)
+- **Result:** ~400 downloads/month, 69% one-and-done dropout. Incentivized installation, not engagement.
+- **Core problem:** The app has no pull after the first visit.
+- **Worth salvaging?** Fix the product first. Then rerun acquisition.
+
+### Campaign 08 — Survey Completion Bonus
+- **Result:** 23% completion rate, feedback never acted on. Circular loop.
+- **Two quotes worth more than all the data:**
+  - *"What if you helped people discover their coffee personality instead of just tracking points?"*
+  - *"I'd engage more if it felt like Basecamp, not like generic corporate rewards."*
+- **Worth salvaging?** Replace with identity-based discovery that generates feedback as a byproduct.
+
+### Campaign 09 — Social Share Bonus
+- **Result:** 12 redemptions. Baristas hated it. Customers felt instrumentalized.
+- **Core insight:** Social sharing works when it reflects identity, not when it's purchased. "I'm an Espresso Explorer!" is shareable. A points transaction is not.
+
+### Campaign 10 — Visit Streak Bonus (Never launched)
+- **Core problem:** Borrowed Duolingo's mechanic without asking if it transfers. Coffee 7 days in a row isn't an identity statement.
+- **Worth salvaging?** No. But the author's own conclusion is the most valuable line in any file:
+  *"All of these campaigns are about getting people to DO MORE TRANSACTIONS. None of them are about making people FEEL SOMETHING."*
+
+---
+
+## The Strategic Gap
+
+Every campaign treated engagement as a transaction frequency problem.
+
+The real problem is **identity**. Members don't engage because there's nothing to engage *with*. No personality, no story, no sense of belonging. Points aren't enough — people need a reason to care.
+
+One idea surfaces repeatedly across the files, sometimes explicitly, sometimes as a throwaway note in the margins:
+
+**Coffee personality.** What if the program helped people discover who they are as coffee drinkers — and built everything else around that?
+
+That's not a campaign. That's a platform.