Appendix A: Lumen Sample Prompts by Company Stage
These prompts cover four company stages that most SaaS products move through. Each stage has prompts across multiple workflows. Paste them directly into Claude Code and replace the Helix-specific details with your own.
Stage 1: Pre-PMF (0–500 users, searching for fit)
At this stage, you are trying to find which segment has the strongest signal and what the product needs to do to earn retention. You do not have much data. You probably do not have PostHog properly instrumented. The most important thing Lumen does here is help you structure what you know — and surface what you do not.
Validate your first PMF hypothesis:
/lumen:pmf-discovery
Product: [Your product] — [one-sentence description]
Business model: B2B SaaS / PLG
Segments: [Your segments — be specific. "Teams of 3–10 engineers" not "teams"]
Stage: [Number] users, [Number] months since launch
PostHog: [connected / not connected]
Interviews: [Number] interviews conducted ([breakdown by segment])
Key question: We believe [segment] has the strongest PMF signal. Is that true, and what does the data say about what they actually value?
Constraint: We cannot build anything requiring more than 3 engineering weeks. We need signal within 4 weeks.
Decide which segment to bet on:
/lumen:pmf-discovery
Product: [Your product]
Context: We have weak PMF signals across two segments and cannot afford to
serve both well. We need to choose one to focus on for the next quarter.
Segment A: [Name] — [brief description, any PMF signal you have]
Segment B: [Name] — [brief description, any PMF signal you have]
Interviews: [Total count across both segments]
Key question: Based on the available signals, which segment has a stronger latent PMF and is more likely to reach a score above 40/100 with focused product investment?
Constraint: We cannot pivot the core product. The decision must work within the existing product architecture.
Design your first PMF experiment:
/lumen:feature
Feature: [Describe the single change you believe will most improve retention]
Target users: [Your chosen segment]
Current evidence: [What you know: interview quotes, retention numbers, usage patterns]
Uses AI/ML: [Yes/No]
Key question: Is this the right experiment to run first, or is there a cheaper signal we should collect before building?
Constraint: [Your build capacity — be specific]
Stage 2: Early Growth (500–5,000 users, scaling what works)
You have PMF in at least one segment. The challenge now is scaling activation, improving retention, and starting to think about expansion. You probably have PostHog properly instrumented and some interview data.
Improve activation in your core segment:
/lumen:feature
Feature: Improved onboarding flow for [your core segment]
Target users: [Core segment] — [define the cohort, e.g., "new signups in first 7 days"]
Current evidence:
- Activation rate (first value event in D7): [current %]
- D30 retention for activated users: [current %]
- D30 retention for non-activated users: [current %]
- PostHog: connected ([X] days of data)
- [Number] interviews conducted on onboarding friction
Key question: What is the highest-leverage change to improve activation rate by at least [target delta]pp?
Constraint: [Engineering capacity]
Plan your first growth quarter:
/lumen:strategy
Product: [Your product]
Quarter: [Q and year]
Current state:
- MAU: [number] ([segment breakdown])
- NRR: [%]
- PMF: [score by segment if known]
- North Star debate: [current metric vs. proposed alternative]
- Engineering capacity: [engineers, sprint velocity]
- Known commitments: [compliance work, customer contracts, etc.]
Board or investor expectation: [revenue or growth target, if any]
Prior quarter learnings: [top 3 things learned last quarter]
Key question: [The single most important strategic question for this quarter]
Constraint: [Your hardest constraint — capacity, regulatory, customer commitment]
Understand your first churn wave:
/lumen:churn
Product: [Your product]
Segment: [Affected segment]
Current NRR: [%]
What changed: [What happened before the churn increase, if known]
Connected MCPs: [Which are connected]
Data available: [PostHog events, Stripe data, HubSpot CRM, exit interviews]
Key question: What is driving the churn increase and who is most at risk of churning in the next 30 days?
Previously tried: [Any interventions already attempted]
Stage 3: Scaling (5,000–50,000 users, expanding segments and launching)
You have proven PMF in your primary segment. Now you are expanding to a new segment, launching a new tier, or building additional platform capabilities. Pricing, messaging, and compliance are more important at this stage.
Prepare a new tier launch:
/lumen:launch
Product: [Your product]
Launch: [Describe what you are launching — new tier, new market, new product]
Target date: [Date — give at least 4 weeks for a meaningful readiness audit]
New capabilities shipping:
- [Capability 1]
- [Capability 2]
- [Capability 3]
Pricing: [New pricing — confirm whether this affects existing customers]
Sales motion: [PLG / sales-assist / hybrid — note if this is a new motion for you]
Target audiences: [Who the messaging needs to reach]
Key risk: [Your single biggest concern about this launch]
Key question: What is the readiness score and what are the gaps we must close
before the launch date?
Build a developer platform strategy:
/lumen:strategy
Product: [Your product]
Context: We are evaluating whether to build a developer API / platform this year.
Current product: [Brief description]
Developer opportunity: [Why you are considering it — inbound requests, competitive move, etc.]
Capacity available for platform work: [Engineers, timeline]
Connected MCPs: [Which are connected]
Key question: Should we invest in a developer platform this year? If yes, what does the MVP look like, what is the right go-to-market, and what are the primary risks?
Constraint: [Any constraints on the platform investment]
Validate an AI feature before building:
/lumen:feature
Feature: [Describe your AI feature — what the model does, what data it reads,
what it outputs to the user]
Uses AI/ML: Yes — [LLM / ML model/recommendation engine / generative feature]
Target users: [Segment and use case]
Current evidence: [What you know about the need — interviews, behavioral signals]
Regulatory context: [GDPR / HIPAA / CCPA — applies if you have users in those jurisdictions]
Key question: What is the ethics clearance status? Is this safe to build?
What conditions apply before production rollout?
Constraint: [Engineering capacity. Note if you cannot wait 72 hours for ethics review.]
Stage 4: Series A and Beyond (50,000+ users, building a platform)
At this stage, the product is a platform. Multiple segments, multiple pricing tiers, a possible developer API, complex OKR cascades across teams, and a board that asks hard questions about NRR and strategic positioning.
Prepare the product section of your Series A deck:
/lumen:strategy
Product: [Your product]
Context: Preparing Series A materials. Need the product narrative to be
defensible under investor diligence on PMF evidence and NRR.
Key metrics:
- ARR: [current]
- NRR: [current %]
- PMF: [scores by segment]
- MAU growth: [MoM %]
Strategic question the board will ask: [The hardest question you expect]
Connected MCPs: [Which are connected — more data = better evidence quality]
Key question: What is the product narrative that most accurately represents
our strategic position and our plan for the next 18 months?
Frame it for an investor who will probe on competitive moat
and PMF evidence quality.
Manage NRR at scale across multiple segments:
/lumen:churn
Product: [Your product]
Segments: All (or specify which)
Current NRR: [%]
NRR breakdown (if known): [Expansion: X%, Contraction: Y%, Churn: Z%]
Connected MCPs: [All that are connected — especially Stripe, HubSpot, PostHog]
Known at-risk accounts: [Top 3–5 by ARR if you know them]
Context: [Any recent product changes, pricing changes, or competitive moves
that might be driving the current NRR]
Key question: What is the fastest path to [target NRR, e.g., 110%] and which
accounts do we act on first this week?
Constraint: [CS team capacity, engineering capacity for product fixes, etc.]
Run a cross-team OKR alignment session:
/lumen:strategy
Product: [Your product]
Quarter: [Q and year]
Team structure: [Number of product teams, what each owns]
Current OKR drafts: [Paste team-level OKR drafts — Lumen will check for conflicts]
North Star: [Current metric]
Engineering capacity: [Total across all product teams]
Known conflicts: [Any cross-team dependencies you already know about]
Key question: Do these OKRs conflict with each other or with the North Star?
Which teams are over-committed for the quarter?
Constraint: [Board-approved targets, customer commitments, regulatory deadlines]