Diagnosis
Diagnosis Mode (UC2)
Diagnosis mode is designed for existing products. If you already have users, metrics, and data — AI CPO will analyze the current state, find problems, and propose a growth strategy.
UC2 = Switch Formula "In Reverse"
UC1 (new product) analyzes why users will switch TO your product.
UC2 (diagnosis) analyzes why users "fire" your product —
what creates Push AWAY from you and Pull TOWARD competitors. Same formula, opposite direction.
UC2 Data Sources
For maximum diagnostic accuracy, connect external sources:
| Source | What It Provides | Artifacts Improved |
|---|---|---|
| PostHog | Retention, funnels, DAU, trends | Audit, Churn Diagnosis, Benchmarks |
| Google Analytics 4 | Traffic sources, conversions, demographics | GTM Strategy, Benchmarks, Unit Economics |
| Telegram | Posts and discussions in target audience channels | Pain Map, Job Statements |
| Yandex.Metrica | Visits, behavior, sources | Audit, GTM |
| Reviews (files) | Text reviews from App Store, Google Play | Pain Map, Job Scorecard |
| Support transcripts | Tickets, support chats | Tax Jobs, Firing Criteria |
Product Audit
Methodology: 5-Dimension Health Audit
A comprehensive analysis of the product's current state across 5 dimensions. Each dimension is scored from 0 to 100. AI generates an overall health score and identifies risk zones.
| Dimension | What It Evaluates | Key Metrics | Target Values |
|---|---|---|---|
| PMF Score | Degree of Product-Market Fit | Sean Ellis survey, retention D7/D30, engagement rate | Sean Ellis > 40%, D30 > 25% |
| Monetization Health | Financial health | LTV/CAC, ARPU trend, free→paid conversion, churn | LTV/CAC > 3, churn < 5% |
| Tech Debt | Technical sustainability | Feature delivery speed, incident frequency, coverage | Depends on stage |
| Team Capacity | Team-task alignment | Role coverage, overload, skill gaps | All critical roles filled |
| Market Position | Competitive position | Market share, niche trend, entry barriers | Growth ≥ market growth |
PMF Score: Sean Ellis Test
The most reliable Product-Market Fit indicator — Sean Ellis Test: "How disappointed would you be if our product ceased to exist?" If > 40% answer "very disappointed" — you have PMF.
Additional PMF Signals
- Customers use the product without reminders (daily active)
- Cancellation churn < involuntary churn (expired cards > voluntary cancellation)
- Organic referrals without a referral program
- Outrage when threatened with shutdown or changes
- Users invent workarounds for missing features
Audit Example
Email marketing SaaS:PMF Score: 65/100 (moderate — loyal core exists, but D30 retention is declining)
Monetization: 45/100 (LTV/CAC = 2.1x — below normal, churn 8%)
Tech Debt: 70/100 (manageable, API needs refactoring)
Team: 55/100 (no marketer, frontend overloaded)
Market: 60/100 (market growing, but competition intensifying)
Priority: Reduce churn (monetization) → Improve onboarding (PMF) → Hire marketer
Tips
- Connect PostHog/GA4 — AI will auto-extract retention, conversions, DAU
- Describe your team, stack, and metrics in chat — the audit will be more accurate
- The audit is a starting point: use it to prioritize subsequent artifacts
Churn Diagnosis
Methodology: Switch Formula in Reverse
Churn is a Switch "away from you." The user "fires" your product. The same formula applies:
Churn Switch Formula
CHURN happens when:PUSH (from your product) + PULL (to competitor/non-consumption) > INERTIA (habit with yours) + ANXIETY (fear of switching)
Push from your product: bugs, friction, unsolved Jobs, poor support
Pull to competitor: new product is better, cheaper, simpler
Inertia (yours): habit, data in the product, switching cost
Anxiety (about leaving): "what if the competitor is worse", "I'll lose my data"
Churn Reduction Strategy via Switch Formula
- Reduce Push: Fix bugs, kill Tax Jobs, improve support
- Weaken competitor Pull: Solve the same Jobs better, Fire Competitors in communications
- Increase Inertia (healthy): More data in the product, integrations, habits
- Increase exit Anxiety: "You'll lose X months of data", "Your configured processes will disappear"
Churn Types and Strategies
| Type | When They Leave | Switch Formula Cause | Strategy |
|---|---|---|---|
| Early churn | First 7 days | Weak Pull — didn't reach AHA moment (DP10) | Improve onboarding, reduce Time to Value |
| Mid-term churn | 1-3 months | Growing Push — unsolved Jobs, friction (Tax Jobs) | Engagement loops, habit formation, solve Tax Jobs |
| Late churn | 6+ months | Competitor Pull — better product appeared, or needs changed | Expansion revenue, new use cases, upsell |
| Involuntary churn | Expired payments | Not a Switch — technical payment failure | Dunning emails, retry logic, warnings |
Switch Cost Builder
A healthy way to reduce churn is to increase positive switching cost (not lock-in, but real value from accumulated data):
- Data accumulation: the longer they use it, the more valuable the historical data
- Integrations: connected to 5 services = hard to migrate
- Configured processes: templates, automations, workflows
- Team usage: the whole team is used to it = high inertia
- Internal social proof: "already 200 reports sent through us"
Churn Diagnosis Example
Email marketing SaaS:Early churn (35% of all losses): users don't send their first email within 3 days
Switch analysis:
— Push from product: complex domain setup (SPF, DKIM) = Tax Job
— Weak Pull: didn't reach AHA moment (first sent email)
— Inertia = 0 (just started, no habit)
— Exit Anxiety = 0 (nothing to lose)
Solution: setup wizard + first email template + day 2 push
Expected effect: early churn −40% → overall churn from 8% to 5.5%
Job Scorecard
Methodology: Job Completion Rate
Job Scorecard is a jobs-based alternative to NPS. Instead of abstract "recommendation," you see which specific jobs the product solves well and which it doesn't. Scoring is done for each job per segment.
In Product DNA this is called Job Completion Rate (JCR) — how fully the product performs the job. Value = positive prediction error in the brain (when the result exceeds expectations).
Rating Scale
| Score | Level | Description | Action |
|---|---|---|---|
| 5 | Delighted | Solves better than expected — AHA moment | Maintain, use in marketing |
| 4 | Satisfied | Solves well — meets expectations | Maintain, don't break |
| 3 | Neutral | Works, but not impressive | Improve if it's a Core Job for A-segment |
| 2 | Frustrated | Partially solved, with difficulty | Critical gap — improve urgently |
| 1 | Failed | Doesn't solve or worse than alternatives | Either fix or abandon this Job |
Opportunity Score
The most valuable opportunities are where Job Importance is high and Satisfaction is low:
Opportunity Score Formula
Opportunity = Job Importance + (Job Importance − Satisfaction)Opportunity > 15 → critical gap, improve urgently
Opportunity 10-15 → significant opportunity
Opportunity < 10 → satisfactory or unimportant
Scorecard Example
| Job | Importance | Seg. A | Seg. B | Opportunity | Action |
|---|---|---|---|---|---|
| Time tracking | 9 | 4.2 | 3.8 | 13.8 | Core — maintain |
| Client report | 8 | 2.5 | 3.1 | 13.5 | Critical gap |
| Invoicing | 7 | 1.8 | 2.0 | 12.2 | Integration or partnership |
| Project management | 5 | 3.5 | 4.0 | 6.5 | Good — don't touch |
Benchmarks
Methodology: Industry Comparison
AI compares your metrics with industry benchmarks. Benchmarks are selected based on niche, stage, business model, and audience (B2B/B2C). Without benchmarks, it's impossible to tell whether your metrics are good — "5% churn" can be excellent for B2C and terrible for Enterprise B2B.
Benchmark Categories
| Category | Metrics | Typical Benchmarks (B2B SaaS) |
|---|---|---|
| Retention | D1, D7, D30, D90 | D7: 40-60%, D30: 25-45%, D90: 15-30% |
| Monetization | Free→paid conv., ARPU, LTV | Conv: 3-7%, ARPU: $30-150/month |
| Growth | MoM growth, viral coefficient, CAC | MoM > 10%, K < 0.5 (typical), CAC: $50-500 |
| Engagement | DAU/MAU ratio, session length | DAU/MAU > 20%, session 3-10 min |
| Unit Economics | LTV/CAC, gross margin, payback | LTV/CAC > 3, margin > 70%, payback < 12 months |
Don't Celebrate Growth Without Checking Seasonality
Metrics growing in spring? Check if it's seasonality before celebrating "found PMF."
Classic case: "PMF found!" in spring → everything crashed in fall. For a new product: compare growth rate
week-over-week, exclude seasonal patterns in your category.
Benchmark Comparison Example
SaaS B2B, PMF stage:Retention D30: 28% vs Benchmark: 35-45% → Below normal
LTV/CAC: 4.3x vs Benchmark: 3-5x → Normal
Churn: 8% vs Benchmark: 3-5% → Above normal
Free→Paid conv.: 5% vs Benchmark: 3-7% → Normal
Growth Strategy
Methodology: Job Chain Map + AARRR
The growth strategy is built at the intersection of Product DNA (which jobs to expand or add) and the AARRR framework (which funnel stage to optimize). AI creates a 30/60/90-day plan.
5 Growth Levers
| Lever | What to Optimize | Product DNA Link | Metrics |
|---|---|---|---|
| Acquisition | New channels, funnel | Orientational Jobs = SEO capture. Activating Knowledge = content | CAC, conversion by channel |
| Activation | Onboarding, Time to Value | Micro Jobs = onboarding steps. DP10 = AHA moment | % reaching AHA, Time to Value |
| Retention | Engagement, habit loops | Frequency = daily/weekly Jobs. Tax Jobs = friction → churn | D7/D30/D90, DAU/MAU |
| Revenue | Upsell, cross-sell, pricing | Related Jobs = upsell opportunity. Big Job coverage = expansion | ARPU, NRR, expansion revenue |
| Referral | Viral mechanics, referrals | Virality = social/shared Jobs. DP11 = what they told others | K-factor, referral rate |
Job Chain Map: Growth Through Related Jobs
Growth strategy via Job Chain Map: identify which Related Jobs arise before and after the Core Job, and solve them too. This creates lock-in (positive switching cost) and increases LTV.
Job Chain Expansion Example
Core Job: Time tracking per projectPre-Job: Planning tasks for the day → (to-do integration)
Post-Job: Invoicing the client → (billing integration)
Related Job: Monthly productivity analysis → (analytics dashboard)
Strategy: Each new Job in the chain = new revenue stream + increased switching cost.
Priority: Post-Job "invoicing" — closer to end-value, higher margin.
90-Day Plan Format
Example
Days 1-30 (Quick wins):— Add onboarding wizard (40% early churn reduction) [Activation]
— Set up dunning emails (60% involuntary churn reduction) [Revenue]
— Kill Top-3 Tax Jobs from Job Scorecard [Retention]
Days 31-60 (Core improvements):
— Launch referral program (target: 15% of new users) [Referral]
— Improve client reports (Job Scorecard gap) [Retention → Revenue]
— Content on orientational jobs (3 blog articles) [Acquisition]
Days 61-90 (Scale):
— Add 2 acquisition channels (YouTube, partnerships) [Acquisition]
— Launch annual plan with 20% discount (churn reduction) [Revenue]
— Billing integration (Post-Job from Job Chain) [Revenue]
Metrics by Stage
| Stage | North Star Metric | Supporting Metrics | Validation Method |
|---|---|---|---|
| 0 → $1K MRR | Paying customers (count) | Interview → sale conversion | Direct sales, not funnel |
| $1K → $10K MRR | Call conversion rate | M1 churn, CAC per channel | CRM, manual analysis |
| $10K → $50K MRR | LTV/CAC ratio | LTV cohort M3/M6, NPS, NRR | Spreadsheets, ABCDX segmentation |
| $50K+ MRR | Revenue / Gross Profit | Cohort unit economics, NRR | Strict A/B + feature flags |
Growth Red Flags
- Conversion rate < 5% → PMF problem or wrong ICP (reconsider segment)
- 80% of support from C/D customers → remove them, focus on A/B
- Metrics growing = check seasonality before celebrating
- K-factor < 0.5 → viral growth is impossible, build on paid