focus

What this skill does

# 80/20 Focus

You suspect something you're doing isn't worth your time. This skill helps you decide: kill it, refine it, or pivot to something with more leverage.

**This skill is for evaluating whether an activity is worth your time — time allocation, not product decisions.** For deciding which features to build next, use **prioritize**. For finding your best acquisition channel without a specific activity to evaluate, use **growth**.

## How This Works

1. You describe the activity and what's (not) happening
2. I diagnose: wrong activity or wrong execution?
3. I compare it against higher-leverage alternatives at your stage
4. You get a verdict and a concrete plan for this week

---

## Step 1: Understand

Before diagnosing, establish the basics. Ask the founder:

- What exactly are you doing? (Be specific — not "marketing" but "sending 50 cold emails per week to SaaS founders")
- How much time per week?
- What results so far? (Replies, demos, signups, revenue — whatever the activity is supposed to produce)

2-3 questions. Not an interrogation.

---

## Step 2: Diagnose

Apply two filters:

### Ceiling Test

Even if you executed this perfectly, would it move the needle at your stage?

Signs of a low ceiling:
- Cold outreach to enterprise when you have no case studies or social proof
- SEO content strategy when you have 0 customers and no product-market fit signal
- Building integrations when your core product doesn't retain users
- Paid ads when you don't know your conversion rate or LTV
- Perfecting onboarding when you don't have enough signups to measure

A low ceiling means the activity can't work yet — not at this stage, not with these prerequisites missing.

### Execution Test

Are you doing the high-leverage 20% of this activity, or spreading effort across the full 100%?

Signs of a 100% spread:
- Sending 100 generic emails instead of 10 deeply researched ones
- Writing 4 blog posts per week instead of 1 exceptional one
- Building 5 features at once instead of finishing 1
- Posting on 4 social platforms instead of dominating 1
- Attending 3 networking events per week instead of deeply following up with 5 warm contacts

The 80/20 version is almost always: do less, but do it with more depth and intention.

If both tests fail — low ceiling AND spread execution — the verdict is Kill. Wrong activity done the wrong way cannot be refined into something that works.

---

## Step 3: Leverage Comparison

Regardless of the diagnosis, compare the current activity against 2-3 alternatives in the same category. The founder needs to see: even if this activity is fine, is it the BEST use of these hours?

### Category: Customer Acquisition

| Tactic | Early Stage (<$1k MRR) | Growth ($1k-$10k MRR) | Scaling ($10k+ MRR) |
|--------|----------------------|----------------------|---------------------|
| Direct outreach | High leverage (if targeted) | Medium — shifts to partnerships | Low — doesn't scale |
| Content/SEO | Low — too slow, no domain authority | Medium — start building | High — compounds over time |
| Paid ads | Low — don't know LTV yet | Medium — test with small budget | High — scale what converts |
| Community/social | High — build relationships that convert | High — establishes authority | Medium — diminishing personal returns |
| Referrals | Low — not enough users | High — program pays for itself | High — lowest CAC channel |
| Partnerships | Low — nothing to offer yet | High — mutual amplification | High — channel partnerships |

**Common trap:** Founders at early stage invest in SEO or paid ads because they feel scalable. But without product-market fit signal, you're scaling something that doesn't work.

### Category: Product

| Tactic | Early Stage | Growth | Scaling |
|--------|------------|--------|---------|
| New features | Medium — only if validating PMF | Low — focus on what exists | Medium — expand for new segments |
| Bug fixes | High — broken product kills trust | High — reliability matters | High — always |
| Onboarding polish | High — activation is everything | High — biggest ROI per hour | Medium — diminishing returns |
| Technical debt | Low — premature optimization | Medium — only if blocking you | High — invest systematically |
| Design polish | Low — function over form | Medium — builds trust | High — competitive differentiator |

**Common trap:** Building features before fixing activation. New features don't help if users never experience the existing ones.

### Category: Retention

| Tactic | Early Stage | Growth | Scaling |
|--------|------------|--------|---------|
| Manual check-ins | High — learn why people stay/leave | Medium — can't scale | Low — automate |
| Email sequences | Medium — worth a basic welcome | High — lifecycle program | High — sophisticated segmentation |
| Feedback collection | High — talk to every user | High — systematize | Medium — diminishing signal-to-noise |
| Usage monitoring | Low — not enough data | High — catch churn signals | High — predictive models |

**Common trap:** Building retention mechanics before you have enough users to retain. At early stage, just talk to people.

### Category: Revenue

| Tactic | Early Stage | Growth | Scaling |
|--------|------------|--------|---------|
| Pricing changes | High — most founders underprice | High — test annually | Medium — optimize |
| Upsells/expansion | Low — get the first sale right | High — existing customers are cheapest | High — major growth lever |
| Payment optimization | Low — not enough volume | Medium — reduce failed payments | High — dunning is money |
| Annual plans | Medium — if anyone will commit | High — improves cash flow | High — reduces churn |

**Common trap:** Offering discounts instead of raising prices. Most early-stage SaaS is underpriced, not overpriced.

### Category: Operations

| Tactic | Early Stage | Growth | Scaling |
|--------|------------|--------|---------|
| Legal/compliance | Low — do the minimum | Medium — as revenue grows | High — real liability |
| Automation/tooling | Low — manual is fine | Medium — automate repetitive | High — systems thinking |
| Hiring/contracting | Low — do it yourself | Medium — first hire | High — build the team |
| Financial tracking | Low — spreadsheet | Medium — proper books | High — forecasting matters |

**Common trap:** Automating things you do once a week. Your time is better spent on customers until the manual work takes hours per day.

---

## Step 4: Recommend

Deliver one of three verdicts:

### Kill
The activity has a low ceiling at the founder's stage, or alternatives are dramatically higher leverage. Stop entirely. Redirect the freed-up hours to a specific alternative.

### Refine
The activity is right, but execution is spread too thin. Cut to the 20% version — specific instructions on what to stop doing within this activity, and what to double down on.

### Pivot
The category is right (e.g., customer acquisition) but the tactic is wrong for this stage. Switch to a specific alternative tactic with instructions for the 80/20 version.

---

## Output Format

Every recommendation must follow this format:

**Verdict:** Kill / Refine / Pivot

**Why:** 2-3 sentences. What's the effort-to-result ratio? What's the ceiling? How does it compare to alternatives?

**This week:**
- Stop doing [specific thing]
- Start doing [specific thing] — here's the 20% version: [concrete instructions]
- Expected time shift: [X hrs/week freed up → redirected to Y]

**The 80/20 version of [recommended activity]:**
A specific description of the minimum effective dose. Not "do content marketing" but "write one LinkedIn post per day sharing a lesson from building your product. No blog, no SEO, no content calendar. Just the post."

**Trap to avoid:** One common mistake founders make when switching to this activity.

---

## Related Skills

- **prioritize** — RICE scoring for which features to build (product decisions, not time allocation)
- **growth** — PLG strategy and activation funnels
- **validat