✦

Workspace · 14 tasks reviewed · 4 owners

Marketing task list — Plexus compatibility analysis

Rick, Sheeva, Mark, and Adam submitted 14 task ideas for Plexus to take on. Below: which are compatible today, which need one new connector, which need infrastructure Plexus doesn't have yet, and where to start.

TIER 1 · ship this week with foundation docs

Tier 2 · one connector unlocks all of them

Tier 3 · need scheduling + research primitives

~140

HOURS / MONTH SAVED · IF ALL SHIP

Compatible today

Tier 1 — Foundation docs are the only missing feature

Pure template + tone + ICP application. No external connectors needed for v1.

◐ Ships this week

Task	Owner	Foundation reads	Hrs/mo
Event content engine One event brief → pre-event LinkedIn posts, session summaries, follow-up email, post-event blog.	Rick · B2B	Pion ToV ICP List Content guide Personas Asset library	40
B2B campaign email nurture journey Campaign messaging + sectors → multi-email nurture tailored to each segment.	Example	ICP List Pion ToV Sector positioning Product 101	12
Lead Magnet curation Brief + research data → hero report chapters + 2 mini-report angles + social/email copy.	Rick · B2B	Pion ToV ICP List Competitive positioning Trend monitor	20
Campaign creative skeleton brief Marketing calendar → structured brief with core hook, objectives, visual direction, content pillars.	Sheeva · B2C	Brief template SB ToV Audience profile OKRs	8–10
Subtotal — all 4 ship together	2 owners	Reuses same foundation docs	~82

THE ONE NEW PLEXUS FEATURE NEEDED

A workspaceDocument table — upload Pion ToV / ICP / content guide / templates once, flag which ones to always inject into every skill's system prompt. Skill UI gets a "Reads from:" chip picker. Roughly a one-day build, unlocks all 4 Tier-1 tasks.

One connector unlocks 7 tasks

Tier 2 — Lightdash is the killer integration

Read structured data → AI synthesizes against ToV + KPI definitions → structured report or table.

▦ Ships next sprint

Task	Lightdash	Iterable	Other	Hrs/mo
Case Studies Rick	✓	—	SF · Marketo · LI	3–5
Post-campaign performance report Sheeva	✓	✓	Agorapulse · GA	10
Identify high value users Mark	✓	✓	—	—
Identify churn risk factors Mark	✓	✓	—	—
Core database health reporting Mark	—	✓	—	2
Weekly trading marketing report Adam	✓	—	GA · Count	4
Merch placement performance Adam	✓	—	Count	4
Kevel merch fillers Adam	✓	—	—	2
Subtotal — Lightdash + Iterable	6 tasks	3 tasks	—	~25

ROI MATH

Lightdash connector: 2–3 days to build (auth, query templates, schema discovery, rate limiting). Unlocks ~25 hours/month of recurring saved time. Pays back inside week 1. Iterable connector second, ~1 day, picks up the remaining cohort/database tasks.

Needs new primitives

Tier 3 — Real-time signal + scheduling

Quality depends on data freshness and recurring execution Plexus doesn't have yet.

⚠ Needs platform work first

Social trend content ideator

Quality depends entirely on whether the Student trend monitor + Agorapulse give fresh signal. Without that, it's brainstorming from training data — not a moat. Build the trend monitor first, then this skill becomes powerful.

Sheeva · weekly · 20 hrs/mo

LTO visibility

Multi-source (Asana + Lightdash) + daily scheduling + priority ranking with business signals. Plexus has no cron-skill primitive yet. Needs skillSchedule table + a Cloudflare Cron Trigger Worker. ~1-week build.

Adam · daily · 20 hrs/mo

Search optimisation

Algolia API + daily anomaly detection. Same shape as LTO visibility — needs cron + diff/anomaly logic. Once the scheduling primitive exists, this is a 1-day skill on top.

Adam · daily

PUSHBACK

Skip churn-risk-factors for v1

Surfacing causal factors over churn cohorts is data science, not skill-shaped. AI can describe correlations but won't reliably surface causal drivers from observational data. Defer until you have a real BI/analyst pairing.

Three honest reads

What this list tells us about Plexus

The list is well-scoped.

Every task has a clear input, output, frequency, and effort estimate. This is much better signal than "build me an AI marketing thing." The team has thought about what they want — that's gold.

Tier 1 is criminally underpriced.

Event content engine is 40 hours/month saved for what is essentially "template + ToV + ICP applied to a brief." That's the easiest thing for AI to do well. One feature (foundation docs) unlocks 82 hrs/month.

Research skills are the hidden asset.

"Industry/sector trend monitor" + "Student trend monitor" appear as inputs to 5 tasks. These aren't connectors — they're Plexus skills themselves. Owning research-as-a-skill differentiates Plexus from "ChatGPT with a brand kit."

The build order

Five phases, each shippable on its own

Every phase delivers user-visible value. No "platform-only" phases the team has to wait through.

This week Phase 1

Foundation docs + ship one Tier-1 skill

Add the workspaceDocument table. Upload Pion ToV, ICP, content guide, brief template, OKRs. Build Campaign creative skeleton brief (Sheeva's task — smallest scope, fastest validator). Ship to Sheeva in 48 hours; get her feedback.

+8–10 hrs/mo Validates foundation-docs pattern

Week 2 Phase 2

Roll out the other 3 Tier-1 skills

Event content engine, B2B nurture journey, Lead Magnet curation. All use the same foundation-docs mechanism — they're essentially different system prompts on the same plumbing.

+72 hrs/mo 2 teammates validated

Week 3–4 Phase 3

Lightdash connector + Case Studies skill

Build the Lightdash read connector properly (auth, query templates, schema discovery, rate limits). Ship Case Studies first — Rick's task has high visibility because the team currently can't agree on the data. The 6 other Lightdash-dependent tasks follow as 1-day skills each.

+25 hrs/mo at the end of the sprint 7 tasks unlocked

Week 5 Phase 4

Skill scheduling primitive

skillSchedule table + a Cloudflare Cron Trigger Worker. Daily/weekly/monthly skill runs become first-class. Unlocks LTO visibility, Search optimisation, and the weekly trading report as scheduled instead of manual.

+24 hrs/mo scheduled Skills run themselves

Week 6+ Phase 5

Research-skill primitive

Build "trend monitor" as a Plexus-native skill — RSS + web scrape + RAG over recent industry sources, output a structured weekly trend digest. This is the strategic moat: Plexus owns the research that 5 other skills consume, not a generic AI assistant.

Unblocks social trend ideator Strategic differentiator

MY VOTE FOR FIRST SHIP

Campaign creative skeleton brief — Sheeva's task

Smallest scope, fastest validator of the foundation-docs pattern, fastest user feedback loop. Sheeva runs it 2–4 times a month and will know in week 1 whether it's better than the manual brief.

TL;DR

The list collectively justifies ~140 hrs/month back to the team for ~3 weeks of focused build.

4 tasks ship this week with one new feature (foundation docs). ~82 hrs/mo saved.
7 tasks ship next month with one new connector (Lightdash). ~25 hrs/mo saved.
3 tasks need scheduling + the trend-monitor primitive — that's the strategic moat.
Skip churn-risk-factors for v1 (analytical, not skill-shaped) and re-think social-trend-ideator's data source first.
The team is asking for skills, not chat. Only one task really benefits from an interactive thread; the rest are "I want a button that produces my deliverable." Validates the AutomationView design over a chat-first model.
The quality gain column is the real pitch: consistency, governance, grounded-in-our-data. Plexus's strongest pitch isn't "faster" — it's "every output is grounded in our ToV, ICP, KPI definitions, business priorities."